diff --git a/.claude/agents/algo-implementer.md b/.claude/agents/algo-implementer.md
index 959dcf76..ffec7786 100644
--- a/.claude/agents/algo-implementer.md
+++ b/.claude/agents/algo-implementer.md
@@ -28,5 +28,5 @@ from its `roadmap/<domain>/<Name>/` spec into the codebase, following both exact
 - **No comments** (except license/`NOLINT`). Allman braces, brace-init.
 - **Tests**: `TEST_F` on `float`, `StrictMock` only, anonymous-namespace fixture; implement EXACTLY
   the spec's cases — no redundant or extra tests.
-- **Embedded**: scoped `#pragma GCC push_options`/`optimize`/`pop_options` + `OPTIMIZE_FOR_SPEED` on hot paths.
+- **Embedded**: no `#pragma GCC optimize`/`optimize` attribute; `OPTIMIZE_FOR_SPEED` (forced inlining) on hot paths.
 - **Terse**: no preamble/postamble, no plan restatement, no narration; don't re-read files; batch reads.
diff --git a/.claude/agents/executor.md b/.claude/agents/executor.md
index d12784ca..cea92319 100644
--- a/.claude/agents/executor.md
+++ b/.claude/agents/executor.md
@@ -1,6 +1,6 @@
 ---
 name: executor
-description: Implement code changes in numerical-toolbox — float-only templates, no heap, embedded pragmas, TEST_F on float, CMake wiring, docs. Needs a clear task or plan.
+description: Implement code changes in numerical-toolbox — float-only templates, no heap, forced inlining on hot paths, TEST_F on float, CMake wiring, docs. Needs a clear task or plan.
 model: claude-sonnet-4-6
 tools: [Read, Write, Edit, Bash, TodoWrite]
 ---
diff --git a/.claude/agents/planner.md b/.claude/agents/planner.md
index 30fd10e0..3f98085f 100644
--- a/.claude/agents/planner.md
+++ b/.claude/agents/planner.md
@@ -25,7 +25,7 @@ Canonical rules: `AGENTS.md`. Produce plans only — no code edits.
    - [ ] No heap; no recursion; tests too
    - [ ] `template<typename T>` + `static_assert(std::is_floating_point_v<T>)`; `float` only
    - [ ] `TEST_F` on `float` — no `TYPED_TEST`; `StrictMock` only; never plain `TEST()`
-   - [ ] scoped `#pragma GCC push_options`/`optimize`/`pop_options` + `OPTIMIZE_FOR_SPEED` on hot paths
+   - [ ] no `#pragma GCC optimize`/`optimize` attribute; `OPTIMIZE_FOR_SPEED` on hot paths
    - [ ] `doc/` update planned
 
 **Terse**: no preamble/postamble, no plan restatement; don't re-read files; batch reads.
diff --git a/.claude/agents/reviewer.md b/.claude/agents/reviewer.md
index 7a64c59f..8179420e 100644
--- a/.claude/agents/reviewer.md
+++ b/.claude/agents/reviewer.md
@@ -1,6 +1,6 @@
 ---
 name: reviewer
-description: Review code changes against numerical-toolbox standards — no heap, float-only templates, embedded pragmas, TEST_F on float, SOLID, docs. Does NOT modify files.
+description: Review code changes against numerical-toolbox standards — no heap, float-only templates, forced inlining on hot paths, TEST_F on float, SOLID, docs. Does NOT modify files.
 model: claude-sonnet-4-6
 tools: [Read, Bash]
 ---
@@ -35,7 +35,7 @@ End with totals + verdict: APPROVE / REQUEST CHANGES.
 - [ ] `extern template` guarded by `#ifdef NUMERICAL_TOOLBOX_COVERAGE_BUILD`.
 
 **Embedded optimizations (WARNING)**
-- [ ] `#pragma GCC push_options` + `optimize("O3","fast-math")` after `#pragma once`, matching `pop_options` at end of algorithm headers (no TU leak; none in test `.cpp`).
+- [ ] No `#pragma GCC optimize` and no `optimize` attribute anywhere (they stop GCC inlining across the boundary).
 - [ ] `OPTIMIZE_FOR_SPEED` on `Filter/Compute/Update/Solve/Step`.
 
 **Namespaces (WARNING)**
diff --git a/.github/agents/algo-implementer.agent.md b/.github/agents/algo-implementer.agent.md
index 52855b94..51c68874 100644
--- a/.github/agents/algo-implementer.agent.md
+++ b/.github/agents/algo-implementer.agent.md
@@ -31,5 +31,5 @@ Authoritative rules: `AGENTS.md`. Recipe: `roadmap/DEPLOYMENT.md`. Follow both e
 - **No comments** (except license/`NOLINT`). Allman braces, brace-init.
 - **Tests**: `TEST_F` on `float`, `StrictMock` only, anonymous-namespace fixture; implement EXACTLY
   the spec's cases — no redundant or extra tests.
-- **Embedded**: scoped `#pragma GCC push_options`/`optimize`/`pop_options` + `OPTIMIZE_FOR_SPEED` on hot paths.
+- **Embedded**: no `#pragma GCC optimize`/`optimize` attribute; `OPTIMIZE_FOR_SPEED` (forced inlining) on hot paths.
 - **Terse**: no preamble/postamble, no plan restatement, no narration; don't re-read files; batch reads.
diff --git a/.github/agents/executor.agent.md b/.github/agents/executor.agent.md
index 3f16e93d..07f50f7b 100644
--- a/.github/agents/executor.agent.md
+++ b/.github/agents/executor.agent.md
@@ -1,5 +1,5 @@
 ---
-description: "Implement code changes in numerical-toolbox — float-only templates, no heap, embedded pragmas, TEST_F on float, CMake wiring, docs. Needs a clear task or plan."
+description: "Implement code changes in numerical-toolbox — float-only templates, no heap, forced inlining on hot paths, TEST_F on float, CMake wiring, docs. Needs a clear task or plan."
 tools: [read, edit, search, execute, todo]
 model: "Claude Sonnet 4.6"
 handoffs:
diff --git a/.github/agents/planner.agent.md b/.github/agents/planner.agent.md
index 85c5c7ca..e6746308 100644
--- a/.github/agents/planner.agent.md
+++ b/.github/agents/planner.agent.md
@@ -28,7 +28,7 @@ Canonical rules: `AGENTS.md`. Produce plans only — no code edits.
    - [ ] No heap; no recursion; tests too
    - [ ] `template<typename T>` + `static_assert(std::is_floating_point_v<T>)`; `float` only
    - [ ] `TEST_F` on `float` — no `TYPED_TEST`; `StrictMock` only; never plain `TEST()`
-   - [ ] scoped `#pragma GCC push_options`/`optimize`/`pop_options` + `OPTIMIZE_FOR_SPEED` on hot paths
+   - [ ] no `#pragma GCC optimize`/`optimize` attribute; `OPTIMIZE_FOR_SPEED` on hot paths
    - [ ] `doc/` update planned
 
 **Terse**: no preamble/postamble, no plan restatement; don't re-read files; batch reads.
diff --git a/.github/agents/reviewer.agent.md b/.github/agents/reviewer.agent.md
index 64b6be26..5b7ac011 100644
--- a/.github/agents/reviewer.agent.md
+++ b/.github/agents/reviewer.agent.md
@@ -1,5 +1,5 @@
 ---
-description: "Review code changes against numerical-toolbox standards: no heap, float-only templates, embedded pragmas, TEST_F on float, SOLID, docs. Does NOT modify files."
+description: "Review code changes against numerical-toolbox standards: no heap, float-only templates, forced inlining on hot paths, TEST_F on float, SOLID, docs. Does NOT modify files."
 tools: [read, search]
 model: "claude-sonnet-4-6"
 handoffs:
@@ -41,7 +41,7 @@ End with totals + verdict: APPROVE / REQUEST CHANGES.
 - [ ] `extern template` guarded by `#ifdef NUMERICAL_TOOLBOX_COVERAGE_BUILD`.
 
 **Embedded optimizations (WARNING)**
-- [ ] `#pragma GCC push_options` + `optimize("O3","fast-math")` after `#pragma once`, matching `pop_options` at end of algorithm headers (no TU leak; none in test `.cpp`).
+- [ ] No `#pragma GCC optimize` and no `optimize` attribute anywhere (they stop GCC inlining across the boundary).
 - [ ] `OPTIMIZE_FOR_SPEED` on `Filter/Compute/Update/Solve/Step`.
 
 **Namespaces (WARNING)**
diff --git a/.github/copilot-instructions.md b/.github/copilot-instructions.md
index fbe7cdfb..f2a71873 100644
--- a/.github/copilot-instructions.md
+++ b/.github/copilot-instructions.md
@@ -6,7 +6,7 @@ Code-file specifics: `.github/instructions/` (applyTo-scoped). Deployment recipe
 Essentials (full detail in AGENTS.md):
 - **No heap** — bounded containers / `std::array` / `std::optional`; no recursion; tests too.
 - **Float-only** — generic `template<typename T>` + `static_assert(std::is_floating_point_v<T>)`; instantiate/test `float`; no Q15/Q31.
-- **Embedded** — scoped `#pragma GCC push_options` / `optimize("O3","fast-math")` … `pop_options` (never leak to the TU) + `OPTIMIZE_FOR_SPEED` on hot paths.
+- **Embedded** — no `#pragma GCC optimize` / `optimize` attribute (they block inlining); the consumer sets optimization flags per TU; `OPTIMIZE_FOR_SPEED` (forced inlining) on hot paths.
 - **No comments** (except license/NOLINT). Allman braces, brace-init, PascalCase/camelCase.
 - **Tests** — `TEST_F` on `float`, `StrictMock` only, never plain `TEST()`, no redundant cases.
 - **No exceptions** — `std::optional`/status enums; interfaces `virtual ~I() = default`.
diff --git a/.github/instructions/numerical-cpp.instructions.md b/.github/instructions/numerical-cpp.instructions.md
index 442cf557..7bd08028 100644
--- a/.github/instructions/numerical-cpp.instructions.md
+++ b/.github/instructions/numerical-cpp.instructions.md
@@ -1,5 +1,5 @@
 ---
-description: "Numerical C++ rules (float-only): no heap, bounded containers, generic template<typename T> instantiated for float, embedded pragmas, Allman/brace-init, SOLID, const-correct. Canonical: AGENTS.md."
+description: "Numerical C++ rules (float-only): no heap, bounded containers, generic template<typename T> instantiated for float, forced inlining on hot paths, Allman/brace-init, SOLID, const-correct. Canonical: AGENTS.md."
 applyTo: "**/*.{hpp,cpp,h}"
 ---
 
@@ -41,26 +41,11 @@ To replace a function with a platform-specific implementation, define the corres
 
 ## Embedded Optimizations
 
-Every algorithm header MUST bracket its body with a scoped pragma, so the options never leak into the including translation unit:
+Never use `#pragma GCC optimize` or the `optimize` attribute. GCC does not inline a callee whose optimization options differ from its caller's, so options set per header or per function turn every small helper on a hot path into an out-of-line call. Optimization level and floating-point flags are the consumer's, applied to whole translation units (embedded: `-O2`/`-O3` with `-ffast-math -fno-finite-math-only`).
 
-```cpp
-#pragma once
+Apply `OPTIMIZE_FOR_SPEED` (from `numerical/math/CompilerOptimizations.hpp`) on hot-path methods: `Compute()`, `Filter()`, `Calculate()`, `Solve()`, `Update()`, `Step()`. It forces inlining when `NumericalToolbox_ENABLE_OPTIMIZATIONS` is defined and expands to nothing otherwise.
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
-// ... includes and header body ...
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
-```
-
-The `pop_options` block is the last thing in the header. Never use the pragma in test `.cpp` files.
-
-Apply `OPTIMIZE_FOR_SPEED` (from `numerical/math/CompilerOptimizations.hpp`) on hot-path methods: `Compute()`, `Filter()`, `Calculate()`, `Solve()`, `Update()`, `Step()`.
+Use `math::IsFinite` for finiteness checks: a consumer may still build with `-ffinite-math-only`.
 
 ## Naming
 
diff --git a/AGENTS.md b/AGENTS.md
index d094f987..0a13925e 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -22,29 +22,18 @@ resource-constrained embedded systems. Real-time, deterministic, no heap.
   `infra::BoundedDeque<T>::WithMaxSize<N>`, `infra::BoundedList<T>::WithMaxSize<N>`,
   `std::array<T,N>`, `std::optional<T>`. Stack/static only. No recursion. **Tests too.**
 
-## Embedded optimizations (algorithm headers)
-
-```cpp
-#pragma once
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-#include "numerical/math/CompilerOptimizations.hpp"
-
-// ... header body ...
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
-```
-
-The `pop_options` block is the **last thing in the header**: the pragma must never leak into the
-including translation unit (it would silently apply fast-math, e.g. folded NaN checks and
-reassociation, to unrelated code). Use `math::IsFinite` for finiteness checks. Test `.cpp` files
-never use the pragma.
-
-`OPTIMIZE_FOR_SPEED` on hot paths (`Filter/Compute/Update/Solve/Step`). Pure interfaces exempt.
+## Embedded optimizations
+
+- **No `#pragma GCC optimize` and no `optimize` attribute**, in headers or sources. GCC does not
+  inline a callee whose optimization options differ from its caller's, so options set per header or
+  per function turn every small helper on a hot path (element access, accessors, `math::` wrappers)
+  into an out-of-line call.
+- The optimization level and floating-point model belong to the consumer and apply to whole
+  translation units. For embedded targets: `-O2` or `-O3` with `-ffast-math -fno-finite-math-only`.
+- `OPTIMIZE_FOR_SPEED` on hot paths (`Filter/Compute/Update/Solve/Step`). Pure interfaces exempt.
+  With `NumericalToolbox_ENABLE_OPTIMIZATIONS` it forces inlining (`always_inline`, `hot`, `inline`);
+  without it, it expands to nothing.
+- Use `math::IsFinite` for finiteness checks: a consumer may still build with `-ffinite-math-only`.
 
 ## Style
 
diff --git a/CLAUDE.md b/CLAUDE.md
index 1e464645..994b533b 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -6,7 +6,7 @@ Code-file specifics: `.github/instructions/`. Deployment recipe: `roadmap/DEPLOY
 Essentials (full detail in AGENTS.md):
 - **No heap** — bounded containers / `std::array` / `std::optional`; no recursion; tests too.
 - **Float-only** — generic `template<typename T>` + `static_assert(std::is_floating_point_v<T>)`; instantiate/test `float`; no Q15/Q31.
-- **Embedded** — scoped `#pragma GCC push_options` / `optimize("O3","fast-math")` … `pop_options` (never leak to the TU) + `OPTIMIZE_FOR_SPEED` on hot paths.
+- **Embedded** — no `#pragma GCC optimize` / `optimize` attribute (they block inlining); the consumer sets optimization flags per TU; `OPTIMIZE_FOR_SPEED` (forced inlining) on hot paths.
 - **No comments** (except license/NOLINT). Allman braces, brace-init, PascalCase/camelCase.
 - **Tests** — `TEST_F` on `float`, `StrictMock` only, never plain `TEST()`, no redundant cases.
 - **No exceptions** — `std::optional`/status enums; interfaces `virtual ~I() = default`.
diff --git a/ROADMAP.md b/ROADMAP.md
index 69b4bd9b..9d25061e 100644
--- a/ROADMAP.md
+++ b/ROADMAP.md
@@ -3,7 +3,7 @@
 Prioritized backlog of reusable numerical components for generic embedded applications,
 ordered **easiest → hardest to implement** within the constraints of this library
 (templated on `float` / `math::Q15` / `math::Q31`, no heap, bounded containers,
-`#pragma GCC optimize` + `OPTIMIZE_FOR_SPEED` hot paths, typed tests, `doc/` page).
+`OPTIMIZE_FOR_SPEED` hot paths, typed tests, `doc/` page).
 
 Difficulty legend:
 
@@ -52,7 +52,7 @@ on bounded `math::Vector`/`math::Matrix` inputs; tests are `TEST_F` on `float`.
 Every new component should follow the established repository conventions:
 
 - [ ] Header-only template supporting `float` / `math::Q15` / `math::Q31` (or *float-first* where noted)
-- [ ] Scoped `#pragma GCC push_options` / `optimize("O3", "fast-math")` after `#pragma once`, `pop_options` at end of file; `OPTIMIZE_FOR_SPEED` on hot paths
+- [ ] No `#pragma GCC optimize` / `optimize` attribute; `OPTIMIZE_FOR_SPEED` on hot paths
 - [ ] No heap, no recursion, bounded containers (`infra::BoundedVector`, `std::array`)
 - [ ] `static_assert` on supported types and dimensions
 - [ ] Typed tests (`TYPED_TEST`) for multi-type components; `TEST_F` for single-type; `StrictMock` only
diff --git a/cmake/QemuHelpers.cmake b/cmake/QemuHelpers.cmake
index 84480bcb..23e679a2 100644
--- a/cmake/QemuHelpers.cmake
+++ b/cmake/QemuHelpers.cmake
@@ -11,4 +11,8 @@ function(numerical_link_qemu_runtime target)
         hal.qemu.cortex
         gmock_main
     )
+    # newlib-nano's printf has no float conversions unless _printf_float is linked. Without them the
+    # vsnprintf behind std::ostream << float returns a bogus length, libstdc++ allocas that many bytes,
+    # and the first failure message that prints a float wraps the stack pointer and locks up the core.
+    target_link_options(${target} PRIVATE "LINKER:--undefined=_printf_float")
 endfunction()
diff --git a/doc/analysis/DiscreteWaveletTransform.md b/doc/analysis/DiscreteWaveletTransform.md
index 8e2f517e..b7e59a89 100644
--- a/doc/analysis/DiscreteWaveletTransform.md
+++ b/doc/analysis/DiscreteWaveletTransform.md
@@ -115,7 +115,7 @@ $$\hat{x} = [1.0,\; 2.0,\; 3.0,\; 4.0] \checkmark$$
 - **Filter length vs. signal length.** At each level the signal halves; once it equals $P$ the
   periodic convolution wraps completely. Stop decomposition before the signal shorter than the
   filter length to avoid artefacts.
-- **Fast-math reordering.** With `#pragma GCC optimize("fast-math")` floating-point associativity
+- **Fast-math reordering.** With `-ffast-math` floating-point associativity
   relaxes; reconstruction residuals may reach $10^{-5}$ rather than $10^{-7}$ for 32-bit floats.
 
 ## Variants & Generalizations
diff --git a/doc/filters/passive/BiquadCascade.md b/doc/filters/passive/BiquadCascade.md
index 2920bae1..42745256 100644
--- a/doc/filters/passive/BiquadCascade.md
+++ b/doc/filters/passive/BiquadCascade.md
@@ -94,8 +94,8 @@ Input $[1, 0, 0, \ldots]$ → output $[1, 0, 0, \ldots]$ — identity passthroug
   $|z| = 1$. Finite-precision rounding can move a pole just outside, causing instability.
   Use $Q \leq 30$ in single precision; double precision or lattice realizations for higher $Q$.
 - **Denormal floats**: small state values approaching the denormal range stall the FPU pipeline
-  on many embedded cores. Enabling flush-to-zero (FTZ) or the fast-math pragma prevents this
-  at the cost of negligible numerical error.
+  on many embedded cores. Enabling flush-to-zero (FTZ) prevents this at the cost of negligible
+  numerical error.
 - **DC gain normalization**: the RBJ low-pass has unity DC gain by construction. Gain-staging
   between sections is not required; each section's output is well-scaled relative to its input.
 
diff --git a/doc/math/MatrixNorms.md b/doc/math/MatrixNorms.md
index 9df1653b..4756a025 100644
--- a/doc/math/MatrixNorms.md
+++ b/doc/math/MatrixNorms.md
@@ -59,7 +59,7 @@ Vector $\mathbf{v} = [3,\, 4]^\top$: $\|\mathbf{v}\|_2 = 5$, and $\hat{\mathbf{v
 
 **Zero vector normalisation** — dividing by $\|\mathbf{v}\|_2 = 0$ is undefined. The implementation returns an empty optional for near-zero norms.
 
-**Fast-math semantics** — `#pragma GCC optimize("fast-math")` may reorder floating-point operations. The norms are sums of non-negative values, so reordering does not change the sign of the result, but catastrophic cancellation can still occur for near-zero off-diagonal entries.
+**Fast-math semantics** — `-ffast-math` may reorder floating-point operations. The norms are sums of non-negative values, so reordering does not change the sign of the result, but catastrophic cancellation can still occur for near-zero off-diagonal entries.
 
 **Non-square matrices** — FrobeniusNorm, OneNorm, and InfinityNorm apply to any $m \times n$ matrix.
 
diff --git a/doc/performance-optimization/README.md b/doc/performance-optimization/README.md
index adc2ef85..31699fa3 100644
--- a/doc/performance-optimization/README.md
+++ b/doc/performance-optimization/README.md
@@ -44,19 +44,23 @@ set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -Wl,--gc-sections")
 
 ### Fast-Math Considerations
 
-```cpp
-// Enables aggressive floating-point optimizations
-// WARNING: May change numerical behavior slightly
-#pragma GCC optimize("fast-math")
+```cmake
+# Aggressive floating-point optimizations for whole translation units, keeping NaN/Inf checks
+add_compile_options(-ffast-math -fno-finite-math-only)
 ```
 
 Effects of `-ffast-math`:
-- Assumes no NaN or Infinity
-- Allows reordering of operations
+- No `errno` from math functions, so `sqrtf` becomes a single `vsqrt.f32`
+- Allows reordering of operations and reciprocal multiplication instead of division
 - Enables FMA (Fused Multiply-Add) instructions
+- `-ffinite-math-only` (part of `-ffast-math`) assumes no NaN or Infinity and folds `std::isnan`/`std::isfinite`
 - May break IEEE 754 compliance
 
-**Use only when**: You control all inputs and don't need strict IEEE behavior.
+**Use only when**: You control all inputs and don't need strict IEEE behavior. Keep
+`-fno-finite-math-only` unless no code relies on NaN or Infinity checks.
+
+Set these options for whole translation units, never per function: see
+[Optimization Options Must Match Across Calls](#optimization-options-must-match-across-calls).
 
 ---
 
@@ -135,10 +139,17 @@ inline constexpr std::array<float, 512> sineLUT = []() {
 #define HOT_FUNCTION __attribute__((hot))
 
 // Combined macro for critical functions
-#define OPTIMIZE_FOR_SPEED \
-    __attribute__((always_inline, hot, optimize("-O3"), optimize("-ffast-math"))) inline
+#define OPTIMIZE_FOR_SPEED __attribute__((always_inline, hot)) inline
 ```
 
+#### Optimization Options Must Match Across Calls
+
+GCC does not inline a function into a caller whose optimization options differ from its own.
+`#pragma GCC optimize` and `__attribute__((optimize(...)))` give the functions they cover options of
+their own, so every small helper such a function calls — element access, accessors, unit wrappers —
+stays an out-of-line call, and the functions themselves are no longer inlined into their callers.
+Choose the optimization level and floating-point flags for whole translation units instead.
+
 ### 5. Prefer Fixed-Size Types
 
 ```cpp
@@ -204,32 +215,16 @@ set(CMAKE_CXX_FLAGS_DEBUG "-Og -g" CACHE STRING "Debug flags" FORCE)
 - Register allocation
 - Still debuggable (variable inspection works)
 
-### Solution 2: Per-File Optimization Pragmas
-
-```cpp
-// Bracket the performance-critical code; never leave the options active for the rest of the TU
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
-// Performance-critical implementation...
+### Solution 2: Per-Translation-Unit Options
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
-```
-
-### Solution 3: Per-Function Attributes
-
-```cpp
-__attribute__((optimize("-O3")))
-void CriticalFunction() {
-    // This function is always optimized
-}
+```cmake
+# Build the performance-critical sources optimized while the rest stays debuggable
+set_source_files_properties(Controller.cpp PROPERTIES COMPILE_OPTIONS "$<$<CONFIG:Debug>:-O2>")
 ```
 
-**Note**: Function-level attributes don't propagate to callees. Use file-level pragmas for better results.
+Every function in a translation unit shares its options, so inlining is unaffected. Do not use
+`#pragma GCC optimize` or `__attribute__((optimize(...)))` instead: see
+[Optimization Options Must Match Across Calls](#optimization-options-must-match-across-calls).
 
 ---
 
@@ -430,13 +425,11 @@ struct GoodStruct {
 
 ## Quick Reference Card
 
-### GCC Optimization Pragmas
-```cpp
-#pragma GCC optimize("O3")           // Maximum speed
-#pragma GCC optimize("Os")           // Minimum size  
-#pragma GCC optimize("fast-math")    // Aggressive FP
-#pragma GCC push_options             // Save current options
-#pragma GCC pop_options              // Restore options
+### Optimization Flags (whole translation units)
+```cmake
+add_compile_options(-O3)                                 # Maximum speed
+add_compile_options(-Os)                                 # Minimum size
+add_compile_options(-ffast-math -fno-finite-math-only)   # Aggressive FP, NaN/Inf checks kept
 ```
 
 ### Function Attributes
diff --git a/numerical/analysis/ConvolutionCorrelation.hpp b/numerical/analysis/ConvolutionCorrelation.hpp
index 9cc3e388..c4e4aefe 100644
--- a/numerical/analysis/ConvolutionCorrelation.hpp
+++ b/numerical/analysis/ConvolutionCorrelation.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/analysis/FastFourierTransform.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/ComplexNumber.hpp"
@@ -159,7 +154,3 @@ namespace analysis
         FastFourierTransform<float>&);
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/Decibels.hpp b/numerical/analysis/Decibels.hpp
index 6574d284..23157de6 100644
--- a/numerical/analysis/Decibels.hpp
+++ b/numerical/analysis/Decibels.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include <algorithm>
@@ -49,7 +44,3 @@ namespace analysis
         return ToDecibels(maxRatio) - ToDecibels(minRatio);
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/DiscreteCosineTransform.hpp b/numerical/analysis/DiscreteCosineTransform.hpp
index 9dbc51b4..224ac1c6 100644
--- a/numerical/analysis/DiscreteCosineTransform.hpp
+++ b/numerical/analysis/DiscreteCosineTransform.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "numerical/analysis/FastFourierTransform.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -113,7 +108,3 @@ namespace analysis
     extern template class DiscreteCosineTransform<float, 8>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/DiscreteWaveletTransform.hpp b/numerical/analysis/DiscreteWaveletTransform.hpp
index c54287e5..1ffdfb54 100644
--- a/numerical/analysis/DiscreteWaveletTransform.hpp
+++ b/numerical/analysis/DiscreteWaveletTransform.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -261,7 +256,3 @@ namespace analysis
     extern template class DiscreteWaveletTransform<float, 16, 3, 4>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/FastFourierTransformRadix2Impl.hpp b/numerical/analysis/FastFourierTransformRadix2Impl.hpp
index acc3392d..8d56c546 100644
--- a/numerical/analysis/FastFourierTransformRadix2Impl.hpp
+++ b/numerical/analysis/FastFourierTransformRadix2Impl.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "numerical/analysis/FastFourierTransform.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -144,7 +139,3 @@ namespace analysis
     extern template class FastFourierTransformRadix2Impl<math::Q31, 8>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/GoertzelAlgorithm.hpp b/numerical/analysis/GoertzelAlgorithm.hpp
index ad08c6e9..cea30756 100644
--- a/numerical/analysis/GoertzelAlgorithm.hpp
+++ b/numerical/analysis/GoertzelAlgorithm.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/ComplexNumber.hpp"
@@ -124,7 +119,3 @@ namespace analysis
     extern template class GoertzelAlgorithm<float>;
 }
 #endif
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/HilbertTransform.hpp b/numerical/analysis/HilbertTransform.hpp
index 8496b5bc..ccf63a76 100644
--- a/numerical/analysis/HilbertTransform.hpp
+++ b/numerical/analysis/HilbertTransform.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "numerical/analysis/FastFourierTransform.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -165,7 +160,3 @@ namespace analysis
     extern template class HilbertFir<float, 31>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/MelFilterbank.hpp b/numerical/analysis/MelFilterbank.hpp
index 1d48d8e3..23897aab 100644
--- a/numerical/analysis/MelFilterbank.hpp
+++ b/numerical/analysis/MelFilterbank.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -174,7 +169,3 @@ namespace analysis
     extern template class MelFilterbank<float, 256, 20>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/Mfcc.hpp b/numerical/analysis/Mfcc.hpp
index 5c9ea755..779187c7 100644
--- a/numerical/analysis/Mfcc.hpp
+++ b/numerical/analysis/Mfcc.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/analysis/MelFilterbank.hpp"
@@ -125,7 +120,3 @@ namespace analysis
     extern template class Mfcc<float, 256, 20, 13>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/PowerDensitySpectrum.hpp b/numerical/analysis/PowerDensitySpectrum.hpp
index f215c429..848d6764 100644
--- a/numerical/analysis/PowerDensitySpectrum.hpp
+++ b/numerical/analysis/PowerDensitySpectrum.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/analysis/FastFourierTransform.hpp"
 #include "numerical/analysis/windowing/Windowing.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -129,7 +124,3 @@ namespace analysis
     extern template class PowerSpectralDensity<float, 512, test::FftStub<float, 512>, test::TwiddleFactorsStub<float, 256>, 0>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/RealFastFourierTransform.hpp b/numerical/analysis/RealFastFourierTransform.hpp
index d4d45254..79140969 100644
--- a/numerical/analysis/RealFastFourierTransform.hpp
+++ b/numerical/analysis/RealFastFourierTransform.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "numerical/analysis/FastFourierTransform.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -124,7 +119,3 @@ namespace analysis
     extern template class RealFastFourierTransform<float, 16>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/SignalDetectors.hpp b/numerical/analysis/SignalDetectors.hpp
index b2422eec..b60c9666 100644
--- a/numerical/analysis/SignalDetectors.hpp
+++ b/numerical/analysis/SignalDetectors.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include <cstdint>
@@ -135,7 +130,3 @@ namespace analysis
     extern template class RmsEnvelope<float>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/TwiddleFactorsTable.hpp b/numerical/analysis/TwiddleFactorsTable.hpp
index 5a76d473..0909fcb2 100644
--- a/numerical/analysis/TwiddleFactorsTable.hpp
+++ b/numerical/analysis/TwiddleFactorsTable.hpp
@@ -5,11 +5,6 @@
 #include <cmath>
 #include <numbers>
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace analysis
 {
     template<typename QNumberType, std::size_t Length>
@@ -42,7 +37,3 @@ namespace analysis
     extern template class TwiddleFactorsTable<float, 8>;
 }
 #endif
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/analysis/windowing/Windowing.hpp b/numerical/analysis/windowing/Windowing.hpp
index 3e915c11..ef825db9 100644
--- a/numerical/analysis/windowing/Windowing.hpp
+++ b/numerical/analysis/windowing/Windowing.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/QNumber.hpp"
@@ -127,7 +122,3 @@ namespace windowing
     extern template class RectangularWindow<math::Q31>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/control_analysis/ContinuousToDiscrete.hpp b/numerical/control_analysis/ContinuousToDiscrete.hpp
index bfc94833..f0908255 100644
--- a/numerical/control_analysis/ContinuousToDiscrete.hpp
+++ b/numerical/control_analysis/ContinuousToDiscrete.hpp
@@ -1,8 +1,4 @@
 #pragma once
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
 #include "numerical/math/Math.hpp"
@@ -179,7 +175,3 @@ namespace control_analysis
     extern template class ContinuousToDiscrete<float, 2, 1, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/control_analysis/ControllabilityObservability.hpp b/numerical/control_analysis/ControllabilityObservability.hpp
index 9cc3ba06..2934ed5c 100644
--- a/numerical/control_analysis/ControllabilityObservability.hpp
+++ b/numerical/control_analysis/ControllabilityObservability.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
 #include "numerical/math/Math.hpp"
@@ -243,7 +238,3 @@ namespace control_analysis
         math::Matrix<float, 2, 2>&, std::size_t, std::size_t, float);
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/control_analysis/FrequencyResponse.hpp b/numerical/control_analysis/FrequencyResponse.hpp
index 69bd9e17..0e39258d 100644
--- a/numerical/control_analysis/FrequencyResponse.hpp
+++ b/numerical/control_analysis/FrequencyResponse.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
@@ -110,7 +105,3 @@ namespace control_analysis
     extern template class FrequencyResponse<float, 128>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/control_analysis/RootLocus.hpp b/numerical/control_analysis/RootLocus.hpp
index f55f07cc..005f3649 100644
--- a/numerical/control_analysis/RootLocus.hpp
+++ b/numerical/control_analysis/RootLocus.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -140,7 +135,3 @@ namespace control_analysis
     extern template class RootLocus<float, 5, 100>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/control_analysis/TransferFunctionStateSpace.hpp b/numerical/control_analysis/TransferFunctionStateSpace.hpp
index b2c3d6ae..d32ae95b 100644
--- a/numerical/control_analysis/TransferFunctionStateSpace.hpp
+++ b/numerical/control_analysis/TransferFunctionStateSpace.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -167,7 +162,3 @@ namespace control_analysis
     extern template class TransferFunctionStateSpace<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/BangBangHysteresis.hpp b/numerical/controllers/implementations/BangBangHysteresis.hpp
index 0f306b62..84a254f3 100644
--- a/numerical/controllers/implementations/BangBangHysteresis.hpp
+++ b/numerical/controllers/implementations/BangBangHysteresis.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <type_traits>
@@ -74,7 +69,3 @@ namespace controllers
     extern template class BangBangHysteresis<float>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/DeadbeatControl.hpp b/numerical/controllers/implementations/DeadbeatControl.hpp
index 5d312ebb..8934459c 100644
--- a/numerical/controllers/implementations/DeadbeatControl.hpp
+++ b/numerical/controllers/implementations/DeadbeatControl.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/controllers/interfaces/StateFeedbackController.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
@@ -148,7 +143,3 @@ namespace controllers
     extern template class DeadbeatControl<float, 2, 2, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/Feedforward2Dof.hpp b/numerical/controllers/implementations/Feedforward2Dof.hpp
index 05ecd276..365001b0 100644
--- a/numerical/controllers/implementations/Feedforward2Dof.hpp
+++ b/numerical/controllers/implementations/Feedforward2Dof.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/controllers/implementations/SaturationRateLimiter.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <type_traits>
@@ -78,7 +73,3 @@ namespace controllers
     extern template class Feedforward2Dof<float>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/GainScheduledController.hpp b/numerical/controllers/implementations/GainScheduledController.hpp
index 0b2b74d3..ffe00039 100644
--- a/numerical/controllers/implementations/GainScheduledController.hpp
+++ b/numerical/controllers/implementations/GainScheduledController.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <array>
 #include <cassert>
@@ -94,7 +89,3 @@ namespace controllers
     extern template class GainScheduledController<float, 3, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/IntegralStateFeedbackLqi.hpp b/numerical/controllers/implementations/IntegralStateFeedbackLqi.hpp
index 03a2c66e..dfc899ce 100644
--- a/numerical/controllers/implementations/IntegralStateFeedbackLqi.hpp
+++ b/numerical/controllers/implementations/IntegralStateFeedbackLqi.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/controllers/implementations/Lqr.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
@@ -172,7 +167,3 @@ namespace controllers
     extern template class IntegralStateFeedbackLqi<float, 2, 1, 1, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/LeadLagCompensator.hpp b/numerical/controllers/implementations/LeadLagCompensator.hpp
index 6379cccc..4f670381 100644
--- a/numerical/controllers/implementations/LeadLagCompensator.hpp
+++ b/numerical/controllers/implementations/LeadLagCompensator.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <type_traits>
 
@@ -74,7 +69,3 @@ namespace controllers
     extern template class LeadLagCompensator<float>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/Lqg.hpp b/numerical/controllers/implementations/Lqg.hpp
index 3189e9d8..b350d59b 100644
--- a/numerical/controllers/implementations/Lqg.hpp
+++ b/numerical/controllers/implementations/Lqg.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/controllers/implementations/Lqr.hpp"
 #include "numerical/controllers/interfaces/OutputFeedbackController.hpp"
 #include "numerical/filters/active/KalmanFilter.hpp"
@@ -103,7 +98,3 @@ namespace controllers
     extern template class Lqg<float, 4, 1, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/Lqr.hpp b/numerical/controllers/implementations/Lqr.hpp
index 516f9288..3aca891e 100644
--- a/numerical/controllers/implementations/Lqr.hpp
+++ b/numerical/controllers/implementations/Lqr.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/controllers/interfaces/StateFeedbackController.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -145,7 +140,3 @@ namespace controllers
     extern template class Lqr<float, 2, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/LuenbergerObserver.hpp b/numerical/controllers/implementations/LuenbergerObserver.hpp
index 8fdf914a..33c15f20 100644
--- a/numerical/controllers/implementations/LuenbergerObserver.hpp
+++ b/numerical/controllers/implementations/LuenbergerObserver.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
 #include "numerical/solvers/GaussianElimination.hpp"
@@ -143,7 +138,3 @@ namespace controllers
     extern template class LuenbergerObserver<float, 2, 2, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/Mpc.hpp b/numerical/controllers/implementations/Mpc.hpp
index 0ca8b159..8004dd6a 100644
--- a/numerical/controllers/implementations/Mpc.hpp
+++ b/numerical/controllers/implementations/Mpc.hpp
@@ -8,11 +8,6 @@
 #include <algorithm>
 #include <optional>
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace controllers
 {
     template<typename T, std::size_t StateSize, std::size_t InputSize>
@@ -309,7 +304,3 @@ namespace controllers
     extern template class Mpc<float, 2, 1, 10, 5>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/PidIncremental.hpp b/numerical/controllers/implementations/PidIncremental.hpp
index 23189204..e822bc0b 100644
--- a/numerical/controllers/implementations/PidIncremental.hpp
+++ b/numerical/controllers/implementations/PidIncremental.hpp
@@ -4,11 +4,6 @@
 #include "numerical/controllers/interfaces/PidDriver.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace controllers
 {
     template<typename QNumberType>
@@ -235,7 +230,3 @@ namespace controllers
     extern template class PidIncrementalSynchronous<math::Q31>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/controllers/implementations/SaturationRateLimiter.hpp b/numerical/controllers/implementations/SaturationRateLimiter.hpp
index bf24b81b..c28abc9d 100644
--- a/numerical/controllers/implementations/SaturationRateLimiter.hpp
+++ b/numerical/controllers/implementations/SaturationRateLimiter.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <algorithm>
 #include <type_traits>
@@ -101,7 +96,3 @@ namespace controllers
     extern template class SlewLimitedSaturation<float>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/estimators/offline/ExpectationMaximization.hpp b/numerical/estimators/offline/ExpectationMaximization.hpp
index bb0c89aa..79e59430 100644
--- a/numerical/estimators/offline/ExpectationMaximization.hpp
+++ b/numerical/estimators/offline/ExpectationMaximization.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/filters/active/KalmanSmoother.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -227,7 +222,3 @@ namespace estimators
     extern template class ExpectationMaximization<4, 2, 20>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/estimators/offline/LinearRegression.hpp b/numerical/estimators/offline/LinearRegression.hpp
index d624a34d..158381c2 100644
--- a/numerical/estimators/offline/LinearRegression.hpp
+++ b/numerical/estimators/offline/LinearRegression.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/estimators/Estimator.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/solvers/QrDecomposition.hpp"
@@ -77,7 +72,3 @@ namespace estimators
     extern template class LinearRegression<float, 4, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/estimators/offline/PolynomialFitting.hpp b/numerical/estimators/offline/PolynomialFitting.hpp
index d9f630fd..41b0f7d7 100644
--- a/numerical/estimators/offline/PolynomialFitting.hpp
+++ b/numerical/estimators/offline/PolynomialFitting.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
 #include "numerical/solvers/QrDecomposition.hpp"
@@ -73,7 +68,3 @@ namespace estimators
     extern template class PolynomialFitting<float, 8, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/estimators/offline/TotalLeastSquares.hpp b/numerical/estimators/offline/TotalLeastSquares.hpp
index 655d42e0..70cf0728 100644
--- a/numerical/estimators/offline/TotalLeastSquares.hpp
+++ b/numerical/estimators/offline/TotalLeastSquares.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -84,7 +79,3 @@ namespace estimators
     extern template class TotalLeastSquares<float, 10, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/estimators/offline/YuleWalker.hpp b/numerical/estimators/offline/YuleWalker.hpp
index 7fe8411b..887c6f63 100644
--- a/numerical/estimators/offline/YuleWalker.hpp
+++ b/numerical/estimators/offline/YuleWalker.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
 #include "numerical/math/Statistics.hpp"
@@ -125,7 +120,3 @@ namespace estimators
     extern template class YuleWalker<float, 128, 4>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/estimators/online/LmsAdaptiveFilter.hpp b/numerical/estimators/online/LmsAdaptiveFilter.hpp
index 49bffdc0..c16090d0 100644
--- a/numerical/estimators/online/LmsAdaptiveFilter.hpp
+++ b/numerical/estimators/online/LmsAdaptiveFilter.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <array>
 #include <cstddef>
@@ -121,7 +116,3 @@ namespace estimators
     extern template class LmsAdaptiveFilter<float, 4>;
 }
 #endif
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/estimators/online/RecursiveLeastSquares.hpp b/numerical/estimators/online/RecursiveLeastSquares.hpp
index 1e8a38cc..f9977d66 100644
--- a/numerical/estimators/online/RecursiveLeastSquares.hpp
+++ b/numerical/estimators/online/RecursiveLeastSquares.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/estimators/Estimator.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -147,7 +142,3 @@ namespace estimators
     extern template class RecursiveLeastSquares<float, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/AhrsMadgwickMahony.hpp b/numerical/filters/active/AhrsMadgwickMahony.hpp
index bffa68df..e3cab510 100644
--- a/numerical/filters/active/AhrsMadgwickMahony.hpp
+++ b/numerical/filters/active/AhrsMadgwickMahony.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Geometry3D.hpp"
 #include "numerical/math/Math.hpp"
@@ -300,7 +295,3 @@ namespace filters
     extern template class AhrsFilter<float, AhrsMode::Mahony>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/AlphaBetaFilter.hpp b/numerical/filters/active/AlphaBetaFilter.hpp
index 7b1de441..e801b26d 100644
--- a/numerical/filters/active/AlphaBetaFilter.hpp
+++ b/numerical/filters/active/AlphaBetaFilter.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
@@ -153,7 +148,3 @@ namespace filters
     extern template class AlphaBetaFilter<float, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/ComplementaryFilter.hpp b/numerical/filters/active/ComplementaryFilter.hpp
index a45fdcca..8d8f0946 100644
--- a/numerical/filters/active/ComplementaryFilter.hpp
+++ b/numerical/filters/active/ComplementaryFilter.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include <cassert>
@@ -108,7 +103,3 @@ namespace filters
     extern template class ComplementaryFilter<float>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/ExtendedKalmanFilter.hpp b/numerical/filters/active/ExtendedKalmanFilter.hpp
index 5698a204..1261fe32 100644
--- a/numerical/filters/active/ExtendedKalmanFilter.hpp
+++ b/numerical/filters/active/ExtendedKalmanFilter.hpp
@@ -5,11 +5,6 @@
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/MatrixOperations.hpp"
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace filters
 {
     template<typename QNumberType, std::size_t StateSize, std::size_t MeasurementSize, std::size_t ControlSize = 0>
@@ -126,7 +121,3 @@ namespace filters
     extern template class ExtendedKalmanFilter<float, 3, 1, 0>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/KalmanFilter.hpp b/numerical/filters/active/KalmanFilter.hpp
index bf1e45a2..4c5f6578 100644
--- a/numerical/filters/active/KalmanFilter.hpp
+++ b/numerical/filters/active/KalmanFilter.hpp
@@ -5,11 +5,6 @@
 #include "numerical/math/LinearTimeInvariant.hpp"
 #include "numerical/math/MatrixOperations.hpp"
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace filters
 {
     template<typename QNumberType, std::size_t StateSize, std::size_t MeasurementSize, std::size_t ControlSize = 0>
@@ -136,7 +131,3 @@ namespace filters
     extern template class KalmanFilter<float, 2, 1, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/KalmanFilterBase.hpp b/numerical/filters/active/KalmanFilterBase.hpp
index 1b2db1db..a8b25bce 100644
--- a/numerical/filters/active/KalmanFilterBase.hpp
+++ b/numerical/filters/active/KalmanFilterBase.hpp
@@ -5,11 +5,6 @@
 #include "numerical/math/MatrixOperations.hpp"
 #include "numerical/solvers/GaussianElimination.hpp"
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace filters
 {
     namespace detail
@@ -176,7 +171,3 @@ namespace filters
     extern template class KalmanFilterBase<float, 4, 2, 0>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/KalmanSmoother.hpp b/numerical/filters/active/KalmanSmoother.hpp
index 416513c1..70d07d04 100644
--- a/numerical/filters/active/KalmanSmoother.hpp
+++ b/numerical/filters/active/KalmanSmoother.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CholeskyDecomposition.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
@@ -249,7 +244,3 @@ namespace filters
     extern template class KalmanSmoother<4, 2, 20>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/SquareRootKalmanFilter.hpp b/numerical/filters/active/SquareRootKalmanFilter.hpp
index a1c7ae18..6fd0e47b 100644
--- a/numerical/filters/active/SquareRootKalmanFilter.hpp
+++ b/numerical/filters/active/SquareRootKalmanFilter.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/filters/active/KalmanFilterBase.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/GivensRotation.hpp"
@@ -224,7 +219,3 @@ namespace filters
     extern template class SquareRootKalmanFilter<float, 3, 1, 0>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/active/UnscentedKalmanFilter.hpp b/numerical/filters/active/UnscentedKalmanFilter.hpp
index 71a87b5f..78ab6082 100644
--- a/numerical/filters/active/UnscentedKalmanFilter.hpp
+++ b/numerical/filters/active/UnscentedKalmanFilter.hpp
@@ -10,11 +10,6 @@
 #include <algorithm>
 #include <array>
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace filters
 {
     struct UkfParameters
@@ -284,7 +279,3 @@ namespace filters
     extern template class UnscentedKalmanFilter<float, 3, 1, 0>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/BiquadCascade.hpp b/numerical/filters/passive/BiquadCascade.hpp
index 0b0c4993..de6da899 100644
--- a/numerical/filters/passive/BiquadCascade.hpp
+++ b/numerical/filters/passive/BiquadCascade.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include <array>
@@ -204,7 +199,3 @@ namespace filters::passive
     extern template class BiquadCascade<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/CicFilter.hpp b/numerical/filters/passive/CicFilter.hpp
index c456a57b..f2a6b3ea 100644
--- a/numerical/filters/passive/CicFilter.hpp
+++ b/numerical/filters/passive/CicFilter.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <array>
 #include <cstddef>
@@ -121,7 +116,3 @@ namespace filters::passive
     extern template class CicDecimator<float, 2, 4, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/ExponentialMovingAverage.hpp b/numerical/filters/passive/ExponentialMovingAverage.hpp
index a167dd94..e426ede4 100644
--- a/numerical/filters/passive/ExponentialMovingAverage.hpp
+++ b/numerical/filters/passive/ExponentialMovingAverage.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <cassert>
 #include <numbers>
@@ -73,7 +68,3 @@ namespace filters::passive
     extern template class ExponentialMovingAverage<float>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/Fir.hpp b/numerical/filters/passive/Fir.hpp
index 7dfbeaf7..40d7dcaf 100644
--- a/numerical/filters/passive/Fir.hpp
+++ b/numerical/filters/passive/Fir.hpp
@@ -3,11 +3,6 @@
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/RecursiveBuffer.hpp"
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace filters::passive
 {
     template<typename QNumberType, std::size_t N>
@@ -60,7 +55,3 @@ namespace filters::passive
     extern template class Fir<float, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/Iir.hpp b/numerical/filters/passive/Iir.hpp
index 700f0125..34887df5 100644
--- a/numerical/filters/passive/Iir.hpp
+++ b/numerical/filters/passive/Iir.hpp
@@ -3,11 +3,6 @@
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/RecursiveBuffer.hpp"
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace filters::passive
 {
     template<typename QNumberType, std::size_t P, std::size_t Q>
@@ -71,7 +66,3 @@ namespace filters::passive
     extern template class Iir<float, 3, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/IirFilterDesign.hpp b/numerical/filters/passive/IirFilterDesign.hpp
index 9cb59047..e675cabf 100644
--- a/numerical/filters/passive/IirFilterDesign.hpp
+++ b/numerical/filters/passive/IirFilterDesign.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/filters/passive/BiquadCascade.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/ComplexNumber.hpp"
@@ -512,7 +507,3 @@ namespace filters::passive
     extern template class IirFilterDesign<float, 8>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/MedianFilter.hpp b/numerical/filters/passive/MedianFilter.hpp
index cc4c51b7..d9232182 100644
--- a/numerical/filters/passive/MedianFilter.hpp
+++ b/numerical/filters/passive/MedianFilter.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <array>
 #include <cstddef>
@@ -73,7 +68,3 @@ namespace filters::passive
     extern template class MedianFilter<float, 5>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/MovingAverage.hpp b/numerical/filters/passive/MovingAverage.hpp
index 437b16e2..35eaf5ae 100644
--- a/numerical/filters/passive/MovingAverage.hpp
+++ b/numerical/filters/passive/MovingAverage.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/RecursiveBuffer.hpp"
 
@@ -65,7 +60,3 @@ namespace filters::passive
     extern template class MovingAverage<float, 4>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/NotchCombFilter.hpp b/numerical/filters/passive/NotchCombFilter.hpp
index f1cec4eb..b1e42a76 100644
--- a/numerical/filters/passive/NotchCombFilter.hpp
+++ b/numerical/filters/passive/NotchCombFilter.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include <array>
@@ -127,7 +122,3 @@ namespace filters::passive
     extern template class CombFilter<float, 20, true>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/filters/passive/SavitzkyGolayFilter.hpp b/numerical/filters/passive/SavitzkyGolayFilter.hpp
index 68ca7306..af91bd98 100644
--- a/numerical/filters/passive/SavitzkyGolayFilter.hpp
+++ b/numerical/filters/passive/SavitzkyGolayFilter.hpp
@@ -1,11 +1,6 @@
 // Copyright 2024 Numerical Toolbox Contributors. All rights reserved.
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CholeskyDecomposition.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include <array>
@@ -128,7 +123,3 @@ namespace filters::passive
     extern template class SavitzkyGolayFilter<float, 5, 2, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/CholeskyDecomposition.hpp b/numerical/math/CholeskyDecomposition.hpp
index 4424f4d3..3c2dcae4 100644
--- a/numerical/math/CholeskyDecomposition.hpp
+++ b/numerical/math/CholeskyDecomposition.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -118,7 +113,3 @@ namespace math
     extern template class CholeskyDecomposition<float, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/CompilerOptimizations.hpp b/numerical/math/CompilerOptimizations.hpp
index b9d1fbaa..a0c7fc00 100644
--- a/numerical/math/CompilerOptimizations.hpp
+++ b/numerical/math/CompilerOptimizations.hpp
@@ -1,11 +1,7 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__) && defined(NumericalToolbox_ENABLE_OPTIMIZATIONS)
-#define OPTIMIZE_FOR_SPEED __attribute__((always_inline, hot, optimize("-O3"), optimize("-ffast-math"))) inline
-#elif defined(__clang__) && defined(NumericalToolbox_ENABLE_OPTIMIZATIONS)
+#if (defined(__GNUC__) || defined(__clang__)) && defined(NumericalToolbox_ENABLE_OPTIMIZATIONS)
 #define OPTIMIZE_FOR_SPEED __attribute__((always_inline, hot)) inline
-#elif defined(_MSC_VER)
-#define OPTIMIZE_FOR_SPEED
 #else
 #define OPTIMIZE_FOR_SPEED
 #endif
diff --git a/numerical/math/ComplexNumber.hpp b/numerical/math/ComplexNumber.hpp
index 3a22cab6..2638b537 100644
--- a/numerical/math/ComplexNumber.hpp
+++ b/numerical/math/ComplexNumber.hpp
@@ -3,11 +3,6 @@
 #include "numerical/math/QNumber.hpp"
 #include <type_traits>
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace math
 {
     template<typename QNumberType>
@@ -140,7 +135,3 @@ namespace math
     extern template class Complex<Q31>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/ConsistencyMetrics.hpp b/numerical/math/ConsistencyMetrics.hpp
index 06c369d7..1014e578 100644
--- a/numerical/math/ConsistencyMetrics.hpp
+++ b/numerical/math/ConsistencyMetrics.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CholeskyDecomposition.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -108,7 +103,3 @@ namespace math
     extern template class ConsistencyMetrics<float, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/Cordic.hpp b/numerical/math/Cordic.hpp
index 86b08735..78509355 100644
--- a/numerical/math/Cordic.hpp
+++ b/numerical/math/Cordic.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include <array>
@@ -183,7 +178,3 @@ namespace math
     extern template class Cordic<float, 8>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/Geometry3D.hpp b/numerical/math/Geometry3D.hpp
index a02be99a..340a62d5 100644
--- a/numerical/math/Geometry3D.hpp
+++ b/numerical/math/Geometry3D.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -90,7 +85,3 @@ namespace math
         return math::Sqrt(DotProduct(v, v));
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/GivensRotation.hpp b/numerical/math/GivensRotation.hpp
index 114f6985..91e5d21e 100644
--- a/numerical/math/GivensRotation.hpp
+++ b/numerical/math/GivensRotation.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include <type_traits>
@@ -54,7 +49,3 @@ namespace math
         y = newY;
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/HouseholderTransform.hpp b/numerical/math/HouseholderTransform.hpp
index e9b906c1..735a4efa 100644
--- a/numerical/math/HouseholderTransform.hpp
+++ b/numerical/math/HouseholderTransform.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -82,7 +77,3 @@ namespace math
         }
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/LinearTimeInvariant.hpp b/numerical/math/LinearTimeInvariant.hpp
index c89a325e..f04c564f 100644
--- a/numerical/math/LinearTimeInvariant.hpp
+++ b/numerical/math/LinearTimeInvariant.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
 
@@ -70,7 +65,3 @@ namespace math
     };
 
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/Math.hpp b/numerical/math/Math.hpp
index 89628483..eed45f71 100644
--- a/numerical/math/Math.hpp
+++ b/numerical/math/Math.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include <bit>
 #include <cmath>
 #include <cstdint>
@@ -331,7 +326,3 @@ namespace math
 #endif
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/Matrix.hpp b/numerical/math/Matrix.hpp
index f0b0bf15..18bca0e3 100644
--- a/numerical/math/Matrix.hpp
+++ b/numerical/math/Matrix.hpp
@@ -5,11 +5,6 @@
 #include <array>
 #include <type_traits>
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace math
 {
     template<typename T, size_t Rows, size_t Cols>
@@ -368,7 +363,3 @@ namespace math
     extern template class Matrix<float, 3, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/MatrixExponential.hpp b/numerical/math/MatrixExponential.hpp
index ae68792f..784d9d72 100644
--- a/numerical/math/MatrixExponential.hpp
+++ b/numerical/math/MatrixExponential.hpp
@@ -1,8 +1,4 @@
 #pragma once
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -187,7 +183,3 @@ namespace math
     extern template class MatrixExponential<float, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/MatrixNorms.hpp b/numerical/math/MatrixNorms.hpp
index 738a9478..0c8fb225 100644
--- a/numerical/math/MatrixNorms.hpp
+++ b/numerical/math/MatrixNorms.hpp
@@ -1,8 +1,4 @@
 #pragma once
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -96,7 +92,3 @@ namespace math
         return result;
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/MatrixOperations.hpp b/numerical/math/MatrixOperations.hpp
index a38d185e..2fbe1ed8 100644
--- a/numerical/math/MatrixOperations.hpp
+++ b/numerical/math/MatrixOperations.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
 #include <cstddef>
@@ -26,7 +21,3 @@ namespace math
         return a * m * a.Transpose();
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/QNumber.hpp b/numerical/math/QNumber.hpp
index a94f5f30..e42923b3 100644
--- a/numerical/math/QNumber.hpp
+++ b/numerical/math/QNumber.hpp
@@ -7,11 +7,6 @@
 #include <limits>
 #include <numbers>
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace math
 {
     inline constexpr double pi = std::numbers::pi;
@@ -266,7 +261,3 @@ namespace math
     extern template class QNumber<int16_t, 15>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/Quaternion.hpp b/numerical/math/Quaternion.hpp
index d6e03f70..1903efce 100644
--- a/numerical/math/Quaternion.hpp
+++ b/numerical/math/Quaternion.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Geometry3D.hpp"
 #include "numerical/math/Math.hpp"
@@ -292,7 +287,3 @@ namespace math
     extern template class Quaternion<float>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/RecursiveBuffer.hpp b/numerical/math/RecursiveBuffer.hpp
index c602884b..5a0cb4e6 100644
--- a/numerical/math/RecursiveBuffer.hpp
+++ b/numerical/math/RecursiveBuffer.hpp
@@ -4,11 +4,6 @@
 #include <algorithm>
 #include <array>
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace math
 {
     struct IndexRelative
@@ -102,7 +97,3 @@ namespace math
     extern template class RecursiveBuffer<Q15, 4>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/StepResponseMetrics.hpp b/numerical/math/StepResponseMetrics.hpp
index ee404210..bf0fdb7a 100644
--- a/numerical/math/StepResponseMetrics.hpp
+++ b/numerical/math/StepResponseMetrics.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -115,7 +110,3 @@ namespace math
         return reference - tailMean;
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/Toeplitz.hpp b/numerical/math/Toeplitz.hpp
index a6d304f5..3eec6bd5 100644
--- a/numerical/math/Toeplitz.hpp
+++ b/numerical/math/Toeplitz.hpp
@@ -4,11 +4,6 @@
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace math
 {
     template<typename T, std::size_t N>
@@ -169,7 +164,3 @@ namespace math
     extern template class ToeplitzMatrix<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/Tolerance.hpp b/numerical/math/Tolerance.hpp
index d0d71684..959bbbf6 100644
--- a/numerical/math/Tolerance.hpp
+++ b/numerical/math/Tolerance.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 namespace math
 {
     template<typename T>
@@ -13,7 +8,3 @@ namespace math
         return 1e-3f;
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/TriangularSolve.hpp b/numerical/math/TriangularSolve.hpp
index f2a3461e..bbd5e365 100644
--- a/numerical/math/TriangularSolve.hpp
+++ b/numerical/math/TriangularSolve.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -85,7 +80,3 @@ namespace math
         return x;
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/math/platform/MathArm.hpp b/numerical/math/platform/MathArm.hpp
index 195436a7..8c874338 100644
--- a/numerical/math/platform/MathArm.hpp
+++ b/numerical/math/platform/MathArm.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include <bit>
 #include <cmath>
 #include <cstdint>
@@ -757,7 +752,3 @@ namespace math
 #endif
 
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/nonlinear_control/BacksteppingControl.hpp b/numerical/nonlinear_control/BacksteppingControl.hpp
index 65dc0abd..3c33f5b9 100644
--- a/numerical/nonlinear_control/BacksteppingControl.hpp
+++ b/numerical/nonlinear_control/BacksteppingControl.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
@@ -106,7 +101,3 @@ namespace nonlinear_control
     extern template class BacksteppingControl<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/nonlinear_control/FeedbackLinearization.hpp b/numerical/nonlinear_control/FeedbackLinearization.hpp
index 20028745..44785605 100644
--- a/numerical/nonlinear_control/FeedbackLinearization.hpp
+++ b/numerical/nonlinear_control/FeedbackLinearization.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
 #include "numerical/solvers/GaussianElimination.hpp"
@@ -75,7 +70,3 @@ namespace nonlinear_control
     extern template class FeedbackLinearization<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/nonlinear_control/ModelReferenceAdaptiveControl.hpp b/numerical/nonlinear_control/ModelReferenceAdaptiveControl.hpp
index c39a2673..52e18cc4 100644
--- a/numerical/nonlinear_control/ModelReferenceAdaptiveControl.hpp
+++ b/numerical/nonlinear_control/ModelReferenceAdaptiveControl.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
@@ -127,7 +122,3 @@ namespace nonlinear_control
     extern template class ModelReferenceAdaptiveControl<float, 1, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/optimization/Adam.hpp b/numerical/optimization/Adam.hpp
index 5a676a35..7d792b39 100644
--- a/numerical/optimization/Adam.hpp
+++ b/numerical/optimization/Adam.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/optimization/StepOptimizer.hpp"
@@ -86,7 +81,3 @@ namespace optimization
     extern template class Adam<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/optimization/BayesianOptimization.hpp b/numerical/optimization/BayesianOptimization.hpp
index b1b5cd30..b54aa85e 100644
--- a/numerical/optimization/BayesianOptimization.hpp
+++ b/numerical/optimization/BayesianOptimization.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -314,7 +309,3 @@ namespace optimization
     extern template class BayesianOptimization<2, 20, 100>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/optimization/GradientDescent.hpp b/numerical/optimization/GradientDescent.hpp
index 75c32d0d..fd5c8405 100644
--- a/numerical/optimization/GradientDescent.hpp
+++ b/numerical/optimization/GradientDescent.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/optimization/Optimizer.hpp"
 #include <optional>
@@ -74,7 +69,3 @@ namespace optimization
     extern template class GradientDescent<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/optimization/Sgd.hpp b/numerical/optimization/Sgd.hpp
index 9ff9df85..cb8db07c 100644
--- a/numerical/optimization/Sgd.hpp
+++ b/numerical/optimization/Sgd.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/optimization/StepOptimizer.hpp"
 
@@ -64,7 +59,3 @@ namespace optimization
     extern template class Sgd<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/regularization/L1.hpp b/numerical/regularization/L1.hpp
index 3e6c350a..af1787fb 100644
--- a/numerical/regularization/L1.hpp
+++ b/numerical/regularization/L1.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/regularization/Regularization.hpp"
 
@@ -65,7 +60,3 @@ namespace regularization
     extern template class L1<float, 4>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/regularization/L2.hpp b/numerical/regularization/L2.hpp
index 8a363fa4..b36ce7ab 100644
--- a/numerical/regularization/L2.hpp
+++ b/numerical/regularization/L2.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/regularization/Regularization.hpp"
 
@@ -58,7 +53,3 @@ namespace regularization
     extern template class L2<float, 4>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/robust_control/ActiveDisturbanceRejection.hpp b/numerical/robust_control/ActiveDisturbanceRejection.hpp
index a2ba72b3..114e40fa 100644
--- a/numerical/robust_control/ActiveDisturbanceRejection.hpp
+++ b/numerical/robust_control/ActiveDisturbanceRejection.hpp
@@ -3,11 +3,6 @@
 
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -159,7 +154,3 @@ namespace robust_control
     extern template class ActiveDisturbanceRejectionControl<float, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/robust_control/DisturbanceObserver.hpp b/numerical/robust_control/DisturbanceObserver.hpp
index c122907b..2ed0102c 100644
--- a/numerical/robust_control/DisturbanceObserver.hpp
+++ b/numerical/robust_control/DisturbanceObserver.hpp
@@ -1,11 +1,6 @@
 // Copyright (c) 2024, Numerical Toolbox Contributors. All rights reserved.
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/filters/passive/BiquadCascade.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
@@ -163,7 +158,3 @@ namespace robust_control
     extern template class DisturbanceObserver<float, 1, 1, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/robust_control/HInfinityStateFeedback.hpp b/numerical/robust_control/HInfinityStateFeedback.hpp
index cc5ba3d1..8a3b0a8e 100644
--- a/numerical/robust_control/HInfinityStateFeedback.hpp
+++ b/numerical/robust_control/HInfinityStateFeedback.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CholeskyDecomposition.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -278,7 +273,3 @@ namespace robust_control
     extern template class HInfinityStateFeedback<float, 2, 1, 1, 2>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/robust_control/SlidingModeControl.hpp b/numerical/robust_control/SlidingModeControl.hpp
index 35148388..e2bb3fce 100644
--- a/numerical/robust_control/SlidingModeControl.hpp
+++ b/numerical/robust_control/SlidingModeControl.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/LinearTimeInvariant.hpp"
@@ -128,7 +123,3 @@ namespace robust_control
     extern template class SlidingModeControl<float, 2, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/ConditionNumber.hpp b/numerical/solvers/ConditionNumber.hpp
index d6f057d8..efff5bfd 100644
--- a/numerical/solvers/ConditionNumber.hpp
+++ b/numerical/solvers/ConditionNumber.hpp
@@ -1,8 +1,4 @@
 #pragma once
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -37,7 +33,3 @@ namespace solvers
         return normA * normInv;
     }
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/DiscreteAlgebraicRiccatiEquation.hpp b/numerical/solvers/DiscreteAlgebraicRiccatiEquation.hpp
index 895834bb..2875bbb4 100644
--- a/numerical/solvers/DiscreteAlgebraicRiccatiEquation.hpp
+++ b/numerical/solvers/DiscreteAlgebraicRiccatiEquation.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
 #include "numerical/math/QNumber.hpp"
@@ -124,7 +119,3 @@ namespace solvers
     extern template class DiscreteAlgebraicRiccatiEquation<float, 4, 1>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/DormandPrince45.hpp b/numerical/solvers/DormandPrince45.hpp
index 96a8361e..d0a1ba79 100644
--- a/numerical/solvers/DormandPrince45.hpp
+++ b/numerical/solvers/DormandPrince45.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -254,7 +249,3 @@ namespace solvers
     extern template class DormandPrince45<float, 1, 0>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/DurandKerner.hpp b/numerical/solvers/DurandKerner.hpp
index 2402f454..d178c6fb 100644
--- a/numerical/solvers/DurandKerner.hpp
+++ b/numerical/solvers/DurandKerner.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/BoundedVector.hpp"
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
@@ -173,7 +168,3 @@ namespace solvers
     extern template class DurandKerner<float, 10>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/GaussianElimination.hpp b/numerical/solvers/GaussianElimination.hpp
index 71c9e5a0..974da0d6 100644
--- a/numerical/solvers/GaussianElimination.hpp
+++ b/numerical/solvers/GaussianElimination.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -185,7 +180,3 @@ namespace solvers
     extern template class GaussianElimination<math::Q31, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/JacobiEigenSolver.hpp b/numerical/solvers/JacobiEigenSolver.hpp
index 0cf89d5b..f4fb8e93 100644
--- a/numerical/solvers/JacobiEigenSolver.hpp
+++ b/numerical/solvers/JacobiEigenSolver.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/GivensRotation.hpp"
 #include "numerical/math/Math.hpp"
@@ -169,7 +164,3 @@ namespace solvers
     extern template class JacobiEigenSolver<float, 4>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/LevinsonDurbin.hpp b/numerical/solvers/LevinsonDurbin.hpp
index c1cc0ee9..d507b1ad 100644
--- a/numerical/solvers/LevinsonDurbin.hpp
+++ b/numerical/solvers/LevinsonDurbin.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Toeplitz.hpp"
 #include "numerical/solvers/Solver.hpp"
@@ -106,7 +101,3 @@ namespace solvers
     extern template class LevinsonDurbin<float, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/LuDecomposition.hpp b/numerical/solvers/LuDecomposition.hpp
index b8a3bfce..23a54998 100644
--- a/numerical/solvers/LuDecomposition.hpp
+++ b/numerical/solvers/LuDecomposition.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -176,7 +171,3 @@ namespace solvers
     extern template class LuDecomposition<float, 4>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/LyapunovSylvester.hpp b/numerical/solvers/LyapunovSylvester.hpp
index 94ce15ce..f75b562f 100644
--- a/numerical/solvers/LyapunovSylvester.hpp
+++ b/numerical/solvers/LyapunovSylvester.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
 #include "numerical/solvers/LuDecomposition.hpp"
@@ -147,7 +142,3 @@ namespace solvers
     extern template class SylvesterSolver<float, 3, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/QrDecomposition.hpp b/numerical/solvers/QrDecomposition.hpp
index 1215913e..112c8df9 100644
--- a/numerical/solvers/QrDecomposition.hpp
+++ b/numerical/solvers/QrDecomposition.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/GivensRotation.hpp"
 #include "numerical/math/HouseholderTransform.hpp"
@@ -173,7 +168,3 @@ namespace solvers
     extern template class QrDecomposition<float, 3, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/RungeKuttaIntegrators.hpp b/numerical/solvers/RungeKuttaIntegrators.hpp
index a1b6fb73..63206eb1 100644
--- a/numerical/solvers/RungeKuttaIntegrators.hpp
+++ b/numerical/solvers/RungeKuttaIntegrators.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Matrix.hpp"
 #include "numerical/solvers/OdeSystem.hpp"
@@ -90,7 +85,3 @@ namespace solvers
     extern template class RungeKutta4<float, 1, 0>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/SingularValueDecomposition.hpp b/numerical/solvers/SingularValueDecomposition.hpp
index 50a5debc..dd31a8aa 100644
--- a/numerical/solvers/SingularValueDecomposition.hpp
+++ b/numerical/solvers/SingularValueDecomposition.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "infra/util/ReallyAssert.hpp"
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/GivensRotation.hpp"
@@ -533,7 +528,3 @@ namespace solvers
     extern template class SingularValueDecomposition<float, 3, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/numerical/solvers/SpectralRadius.hpp b/numerical/solvers/SpectralRadius.hpp
index 3755bc36..c1fbb0d2 100644
--- a/numerical/solvers/SpectralRadius.hpp
+++ b/numerical/solvers/SpectralRadius.hpp
@@ -1,10 +1,5 @@
 #pragma once
 
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
-
 #include "numerical/math/CompilerOptimizations.hpp"
 #include "numerical/math/Math.hpp"
 #include "numerical/math/Matrix.hpp"
@@ -98,7 +93,3 @@ namespace solvers
     extern template class SpectralRadius<float, 3>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
diff --git a/roadmap/DEPLOYMENT.md b/roadmap/DEPLOYMENT.md
index 6933ebbe..63ab214f 100644
--- a/roadmap/DEPLOYMENT.md
+++ b/roadmap/DEPLOYMENT.md
@@ -4,8 +4,7 @@ Turn one `roadmap/<domain>/<Name>/` spec into shipped code. Rules: `../AGENTS.md
 Read the spec's three files first (`implementation.md`, `tests.md`, `explanation.md`), then:
 
 1. **Header** `numerical/<domain>/<Name>.hpp`
-   - `#pragma once` → `#pragma GCC push_options` + `#pragma GCC optimize("O3","fast-math")` (GCC-only guard) →
-     `#include "numerical/math/CompilerOptimizations.hpp"`; end the file with `#pragma GCC pop_options`.
+   - `#pragma once` → `#include "numerical/math/CompilerOptimizations.hpp"`; no `#pragma GCC optimize`.
    - `template<typename T[, std::size_t sizes...]>` with
      `static_assert(std::is_floating_point_v<T>, "<Name> supports floating-point types");`.
    - Implement per `implementation.md`; `OPTIMIZE_FOR_SPEED` on the hot path(s).
diff --git a/roadmap/README.md b/roadmap/README.md
index 65f94079..461d7164 100644
--- a/roadmap/README.md
+++ b/roadmap/README.md
@@ -37,10 +37,6 @@ Header shape (mirroring existing components such as `Fir.hpp`):
 
 ```cpp
 #pragma once
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC push_options
-#pragma GCC optimize("O3", "fast-math")
-#endif
 #include "numerical/math/CompilerOptimizations.hpp"
 
 namespace <ns>
@@ -57,10 +53,6 @@ namespace <ns>
     extern template class <Name><float /*, sizes... */>;
 #endif
 }
-
-#if defined(__GNUC__) && !defined(__clang__)
-#pragma GCC pop_options
-#endif
 ```
 
 Coverage `.cpp`:
diff --git a/roadmap/math/SE3Transform/implementation.md b/roadmap/math/SE3Transform/implementation.md
index 4e90173f..942b4983 100644
--- a/roadmap/math/SE3Transform/implementation.md
+++ b/roadmap/math/SE3Transform/implementation.md
@@ -77,8 +77,8 @@ function Log():                                 # inverse of Exp
 
 ## Deployment
 
-- Header: `numerical/math/SE3Transform.hpp` — `#pragma once` →
-  `#pragma GCC optimize("O3","fast-math")`, `OPTIMIZE_FOR_SPEED` on `operator*`/`Apply`, and
+- Header: `numerical/math/SE3Transform.hpp` — `#pragma once`, `OPTIMIZE_FOR_SPEED` on
+  `operator*`/`Apply`, and
   `extern template class SE3<float>;` under `#ifdef NUMERICAL_TOOLBOX_COVERAGE_BUILD`.
 - Coverage: `numerical/math/SE3Transform.cpp` → `template class SE3<float>;`
 - Test: `numerical/math/test/TestSE3Transform.cpp`