Move benchmark runners to c8id and prebuilt images

[robert3005]

Summary

• replace c6id.metal benchmark runners with spot-capable c8id.32xlarge
• use c8id.8xlarge for benchmark build jobs
• use the prebuilt AMI for benchmark compilation
• remove Rust setup from SQL benchmark execution jobs that only consume downloaded binaries
• right-size the CodSpeed runner through the new CI runner class

Merge vortex-data/.github-private#29 before this PR. Merging #9734 first is also recommended so prebuilt dependency reconciliation is active.

Validation

• strict yamllint on all five changed workflows
• git diff --check

[codspeed-hq]

Merging this PR will degrade performance by 1.98%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 7 improved benchmarks
❌ 9 regressed benchmarks
✅ 2120 untouched benchmarks
⏩ 164 skipped benchmarks¹
🗄️ 1 archived benchmark run²

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	decompress[u8, (10000, 256)]	41.7 µs	66 µs	-36.83%
❌	Simulation	decompress[u32, (4000, 1024)]	55.5 µs	84.5 µs	-34.33%
❌	Simulation	random_i16[0.8]	75.8 µs	95.1 µs	-20.32%
❌	Simulation	random_i32[0.5]	71.3 µs	89 µs	-19.93%
❌	Simulation	take[duplicates/repeated/primitive/nonnull/chunks=16/indices=1000]	198.7 µs	233.3 µs	-14.82%
❌	Simulation	decompress[u8, (4000, 1024)]	46.9 µs	54.9 µs	-14.59%
❌	Simulation	primitive_i64_arrow_row	68.6 µs	80.3 µs	-14.55%
❌	Simulation	mul_decimal_i64_nonnull	90.2 µs	104.5 µs	-13.67%
❌	WallTime	words_gather_scalar_avx2[65536]	8.3 µs	9.4 µs	-11.6%
⚡	Simulation	filter_run_end[len=4096_run=16_density=0.1]	138 µs	95.3 µs	+44.81%
⚡	Simulation	cached_indices_i32[0.01]	50.9 µs	37.5 µs	+35.66%
⚡	Simulation	optimize_lookup_predicate[ids=1, shape=in_list]	132.7 µs	102.3 µs	+29.74%
⚡	Simulation	decompress[u64, (4000, 1024)]	90.1 µs	71.8 µs	+25.46%
⚡	Simulation	cached_indices_i32[0.1]	66 µs	53.7 µs	+22.77%
⚡	Simulation	decompress[u16, (4000, 1024)]	60.7 µs	49.6 µs	+22.36%
⚡	WallTime	arrow_checked_add_u32_avx2[16384]	21.3 µs	17.7 µs	+20.38%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing rk/benchmark-runners-c8id (b867b3f) with develop (f80b83c)}

Footnotes

1. 164 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

2. 1 benchmark was run, but is now archived. If it was deleted in another branch, consider rebasing to remove it from the report. Instead if it was added back, click here to restore it. ↩