ayo 2558fcbd39 Improve GPU compatibility, RDNA2 kernels and startup diagnostics ...

- Match render/compute adapters and gate RR models by GPU capabilities
- Add RDNA2 native texture kernels and fix FP16 rounding
- Harden model builds, cache validation and packaging
- Add startup tracing, timeouts, documentation and regression tests

2026-10-19 11:36:40 -04:00

..

build_native.sh

Improve GPU compatibility, RDNA2 kernels and startup diagnostics

2026-10-19 11:36:40 -04:00

ctile.py

Add MicroCUDA backend, adaptive output wait, wider native K coverage (improve perf)

2026-10-07 15:54:38 -04:00

ifconv.py

Port K kernel tuning and M gfx1101 flags

2026-10-05 14:48:37 -04:00

in_fuse.py

Port K kernel tuning and M gfx1101 flags

2026-10-05 14:48:37 -04:00

in_hoist.py

shim: adaptive marker polling, opt-in GPU-side input wait/completion and RGBA8 output

2026-10-13 22:30:41 -04:00

in_lds2reg.py

Add MicroCUDA backend, adaptive output wait, wider native K coverage (improve perf)

2026-10-07 15:54:38 -04:00

in_regs.hip

Port K kernel tuning and M gfx1101 flags

2026-10-05 14:48:37 -04:00

m_ptx_rt.h

Port K kernel tuning and M gfx1101 flags

2026-10-05 14:48:37 -04:00

native_rt.h

Port K kernel tuning and M gfx1101 flags

2026-10-05 14:48:37 -04:00

output_tile.h

Improve GPU compatibility, RDNA2 kernels and startup diagnostics

2026-10-19 11:36:40 -04:00

ptx2hip.py

Improve GPU compatibility, RDNA2 kernels and startup diagnostics

2026-10-19 11:36:40 -04:00

ptx_rt.h

Improve GPU compatibility, RDNA2 kernels and startup diagnostics

2026-10-19 11:36:40 -04:00

rewrite_output.py

Add MicroCUDA backend, adaptive output wait, wider native K coverage (improve perf)

2026-10-07 15:54:38 -04:00

split_output_paths.py

shim: adaptive marker polling, opt-in GPU-side input wait/completion and RGBA8 output

2026-10-13 22:30:41 -04:00