gpuasm.com seems down (Cloudflare 1033)

[florianmattana]

Body:

Hey, just a heads up that gpuasm.com looks unreachable right now. I was trying autosass on a SM120 cubin (RTX 5070 Ti) and the client fails on initialize with a Cloudflare 530 / error 1033, which usually means the origin behind the tunnel is down.

python3 optimize.py --cubin 01_vector_add.cubin --kernel _Z10vector_addPKfS0_Pfi --dry-run
Loaded 01_vector_add.cubin (5,840 bytes)
Listing kernels...
RuntimeError: Failed after 3 retries: HTTP 530: error code: 1033

While I have you, two quick questions:

• Any easy way to check service status before kicking off a run?
• Does gpuasm.com support SM120 disasm/reasm? Want to know if it's worth retrying once you're back up.

Thanks for the project, looking forward to trying it on a real kernel.