Build benchmarks on VMs instead of metal instances which we use for runs

[AdamGS]

Rationale for this change

Metal capacity will always be limited, better to use VMs for the build step to make better use of it.

[codspeed-hq]

Merging this PR will degrade performance by 19.98%

⚡ 1 improved benchmark
❌ 3 regressed benchmarks
✅ 1933 untouched benchmarks
⏩ 85 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	decompress[u64, (1000, 16)]	65.1 µs	133.7 µs	-51.31%
❌	Simulation	decode_varbin[(1000, 2)]	62.1 µs	78.6 µs	-21.02%
❌	Simulation	compress_fsst[(1000, 64, 8)]	1 ms	1.2 ms	-11.38%
⚡	Simulation	decompress[u32, (10000, 256)]	100.4 µs	83.5 µs	+20.27%

Tip

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

---

_{Comparing adamg/build-vm (a0a7e66) with develop (fed7038)}

Footnotes

1. 85 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. ↩