Skip to content

Can Tesla P100 16G run Hy-Memory-30B-A3B?

Comfortable: GGUF Q8_0-expsIQ2_XXS needs 9.9 GB at 4K, batch 1; headroom 6.1 GB.

Text input, FP16 KV cache, one card. Unified-memory capacity uses the GPU guide’s usable limit; negative headroom may require host offload. Speed source settings may differ from this 4K estimate.

Measured runs

No measured run for this pair. Sizing is an estimate, not a runtime test.

Quant memory at 4K

QuantTotal GBVerdictWeight sizing
GGUF Q3_K_M15.5TightVerified file
GGUF Q4_K_M19.5Does not fitVerified file
GGUF Q6_K26.2Does not fitVerified file
GGUF Q8_033.8Does not fitVerified file
GGUF Q8_0-expsIQ2_XXS9.9ComfortableVerified file
GGUF Q8_0-expsIQ1_M8.8ComfortableVerified file

Context capacity · GGUF Q8_0-expsIQ2_XXS

ContextTotal GBVerdict
4K9.9Comfortable
16K11.2Comfortable
32K12.8Comfortable
128K22.7Does not fit

Generation speed

GGUF Q8_0-expsIQ2_XXS: 846 tok/s · Bandwidth ceiling (not measured)

Other model / GPU pairs