feat[gpu]: export as Arrow device array binary

[0ax1]

Turns out whilst Arrow device array supports binary, cuDF does not. Therefore, we don't currently exercise binary as part of the cuDF harness.

[codspeed-hq]

Merging this PR will not alter performance

⚠️ 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.

⚡ 4 improved benchmarks
❌ 4 regressed benchmarks
✅ 1519 untouched benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	bitwise_not_vortex_buffer_mut[128]	216.9 ns	275.3 ns	-21.19%
❌	Simulation	bitwise_not_vortex_buffer_mut[1024]	278.6 ns	336.9 ns	-17.31%
❌	WallTime	cuda/bitpacked_u8/unpack/3bw[100M]	300 µs	354 µs	-15.25%
❌	Simulation	bitwise_not_vortex_buffer_mut[2048]	342.2 ns	400.6 ns	-14.56%
⚡	Simulation	chunked_bool_canonical_into[(1000, 10)]	46.8 µs	31.9 µs	+46.83%
⚡	Simulation	chunked_varbinview_canonical_into[(1000, 10)]	198.2 µs	162 µs	+22.31%
⚡	Simulation	chunked_varbinview_into_canonical[(1000, 10)]	213.4 µs	177.2 µs	+20.41%
⚡	Simulation	baseline_lt[16, 65536]	246.7 µs	219.1 µs	+12.62%

Tip

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

---

_{Comparing ad/gpu-arrow-binary-export (7af4940) with develop (f46621d)}