Skip to content

Can Tesla P40 24G run Nemotron 3.5 Bolt E4B?

Comfortable: GGUF Q8_0 needs 8.5 GB at 4K, batch 1; headroom 15.5 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 Q4_K_M6.6ComfortableVerified file
GGUF Q8_08.5ComfortableVerified file

Context capacity · GGUF Q8_0

ContextTotal GBVerdict
4K8.5Comfortable
16K8.7Comfortable
32K9.0Comfortable

Generation speed

GGUF Q8_0: 26–34 tok/s · Derived range

Other model / GPU pairs