Add performance diagnostic to diskann-benchmark

[hildebrandmw]

Add a warning to diskann-benchmark if it looks like the binary was compiled for the x86-64 target CPU instead of x86-64-v3, x86-64-v4, or some other at-least AVX2 compatible CPU.

Also add a warning for aarch64 noting that full SIMD support is not available yet.

With diskann-benchmark being available now on crates.io, I just know that this will be compiled and run with the default x86-64 target CPU, when the preferred target is strongly at least x86-64-v3.

Implementation

Our architecture detection/selection is performed by diskann-wide. As such, we can detect if the whole application has been compiled for an x86-64-v3 compatible target by checking if the compile time architecture diskann_wide::arch::Current is at least diskann_wide::arch::x86_64::V3. To do this comparison, a light-weight, opaque diskann_wide::arch::Level is introduced to allow Architectures to be ordered without instantiating the associated architecture via the Architecture::level associated function.

Checks for compatibility can then be done using

use diskann_wide::{Architecture, arch};
if arch::Current::level() >= arch::x86_64::V3::level() {
    // We're at least AVX2 compatible
}

I think it's reasonable to add this to diskann-wide since diskann-benchmark is not the only crate that may need this functionality and diskann-wide is really the source of truth when it comes to micro-architecture specific functionality.

[Copilot]

Pull request overview

This PR adds a compile-time performance diagnostic to diskann-benchmark by introducing an orderable “capability level” abstraction in diskann-wide, then using that to warn when the benchmark binary appears to be compiled for an insufficient micro-architecture target (notably x86-64 instead of x86-64-v3+), plus an informational warning on aarch64.

Changes:

• Add arch::Level and Architecture::level() to enable ordering micro-architecture capability without instantiating an Architecture.
• Implement level() for Scalar, x86_64::V3, and x86_64::V4, plus add ordering tests.
• Extend diskann-benchmark CLI with --quiet and emit target-related performance warnings.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
diskann-wide/src/helpers.rs	Removes dead commented-out cfg_attr block.
diskann-wide/src/arch/x86_64/v3/mod.rs	Adds ordering support and exposes V3 capability via level().
diskann-wide/src/arch/x86_64/v4/mod.rs	Adds ordering support and exposes V4 capability via level().
diskann-wide/src/arch/x86_64/mod.rs	Introduces internal level ordering and tests for the ordering.
diskann-wide/src/arch/mod.rs	Adds Level type + Architecture::level() API and related docs.
diskann-wide/src/arch/emulated/mod.rs	Implements level() for Scalar.
diskann-benchmark/src/main.rs	Adds top-level Cli wrapper with --quiet and emits compile-target warnings; updates tests accordingly.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 78.84615% with 22 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.99%. Comparing base (f3fcaae) to head (cfc2259).
⚠️ Report is 5 commits behind head on main.

Files with missing lines	Patch %	Lines
diskann-benchmark/src/main.rs	68.11%	22 Missing ⚠️

❌ Your patch status has failed because the patch coverage (78.84%) is below the target coverage (90.00%). You can increase the patch coverage or adjust the target coverage.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #744      +/-   ##
==========================================
- Coverage   89.01%   88.99%   -0.03%     
==========================================
  Files         428      428              
  Lines       78294    78368      +74     
==========================================
+ Hits        69692    69742      +50     
- Misses       8602     8626      +24     

Flag	Coverage Δ
miri	88.99% <78.84%> (?)
unittests	88.99% <78.84%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
diskann-wide/src/arch/emulated/mod.rs	100.00% <100.00%> (ø)
diskann-wide/src/arch/mod.rs	83.79% <100.00%> (+0.85%)	⬆️
diskann-wide/src/arch/x86_64/mod.rs	91.62% <100.00%> (+0.76%)	⬆️
diskann-wide/src/arch/x86_64/v3/mod.rs	95.88% <100.00%> (-1.73%)	⬇️
diskann-wide/src/arch/x86_64/v4/mod.rs	23.88% <100.00%> (+1.29%)	⬆️
diskann-wide/src/helpers.rs	100.00% <ø> (ø)
diskann-benchmark/src/main.rs	91.12% <68.11%> (-4.37%)	⬇️

... and 6 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.