Local model performance on your hardware

reported speed:

32.6 tokens/s generation · 493.0 tokens/s prompt processing

quant:

Q6_K (GGUF)

kv:

q4_0

Reported by the source; GPU count, offloading and concurrent requests can change this figure. Check the full setup before comparing.

User reports Qwen3.8-27B Q6_K at 32.6 t/s decode (tg256, no MTP) on 2x Tesla P100 16GB with tensor split, versus 17.51 t/s upstream at the fork point. Setup is a llama.cpp fork with CUDA work for Pascal (sm_60), q4_0 KV cache, and fp16 math with fp32 accumulation. Prefill pp2048 at 0 context is 493 t/s versus ~250 t/s upstream. With MTP speculative decoding the fork reaches 54 t/s at 2k context and 29-35 t/s at 260k context. Prefill at 260k context is 123 t/s filling and 153 t/s for a question on a loaded context. Perplexity on the gate corpus at -c 4096 is 2.6101.

reported speed:

110.0 tokens/s generation

quant:

Q4_0 (GGUF)

kv:

F16

Reported by the source; GPU count, offloading and concurrent requests can change this figure. Check the full setup before comparing.

User reports Nemotron-3.5-Lightning-30B-A3B at 110 tok/s writing code on two Tesla P100 16GB cards. Setup is llama.cpp b10970 with Q4_0 weights, F16 KV cache, tensor split across both cards, and the model's built-in MTP draft head at n-max 2. The same model reaches 95 tok/s on prose, 50 tok/s at 128k context, 39 tok/s at 256k, and 16 tok/s at 1M tokens. The study covers 545 speed measurements of 29 models from 2B to 122B parameters, all weights and KV cache in VRAM with no system RAM offload. A 119B MoE model runs 39 tok/s against 4.3 tok/s for a 70B dense model on the same cards. Tensor split makes dense models from 8B up 21-44% faster. A single P100 throttles to 906 MHz and loses 25% under sustained load, while two cards share the heat and lose 5.5%.

reported speed:

24.5 tokens/s generation · 448.7 tokens/s prompt processing

quant:

Q4_K_XL (GGUF)

kv:

F16

Reported by the source; GPU count, offloading and concurrent requests can change this figure. Check the full setup before comparing.

User reports Qwen3.8-27B Q4_K_XL at 24.5 tok/s generation and 448.7 tok/s prompt processing on two Tesla P100 16GB cards with tensor split. Setup is a patched llama.cpp (upstream b10660 plus eleven patches) with F16 KV cache and full GPU offload across two cards. The figures are the after-patch numbers from a patch series that improves decode and prefill; the same run measured 22.3 tok/s generation and 427.0 tok/s prompt processing before the patches. A real 7,655-token request reached 398.5 tok/s prompt processing, and four concurrent agents reached 20.3 tok/s each (72.9 aggregate).

reported speed:

50-60 tokens/s generation · 350.0 tokens/s prompt processing

quant:

Q6_K (GGUF)

Reported by the source; GPU count, offloading and concurrent requests can change this figure. Check the full setup before comparing.

User reports Qwen3.8 27B at 50-60 t/s generation and 350 t/s prefill on 2x Tesla P100 16GB with a custom llama.cpp fork. Setup is llama.cpp with Q6_K quant, 262144 context, batch 32768, ubatch 1024, and MTP speculative decoding. At 260k context, generation is 30-35 t/s and prefill is 110 t/s. GPUs are capped at 175W/250W each and run at 79C with minor thermal throttling; user estimates 5-10% higher numbers with better cooling.

reported speed:

54-60 tokens/s generation · 440-500 tokens/s prompt processing

quant:

UD_Q4_K_XL (GGUF)

kv:

16bit

Reported by the source; GPU count, offloading and concurrent requests can change this figure. Check the full setup before comparing.

User reports Qwen3.6 35B A3B at 54-60 t/s generation (66-72 t/s on code) on a single Tesla P100 16GB, up from 30-35 t/s prose on an RX 6600 XT. Setup is llama.cpp with shinbunbun patches, UD_Q4_K_XL quant, 16-bit KV cache, MTP speculative decoding, and --n-cpu-moe 22, running 32k context. Prefill is 600 t/s at 0 ctx dropping to 440-500 t/s by 10k. The user notes the P100 is underrated for the price and has a second card coming for full offload.

reported speed:

85.0 tokens/s generation · 300.0 tokens/s prompt processing

quant:

Q4_K_XL (GGUF)

Reported by the source; GPU count, offloading and concurrent requests can change this figure. Check the full setup before comparing.

User reports Qwen3.6-35B-A3B at ~85 t/s generation and ~300 t/s prompt processing on 2x Tesla P100. Setup is llama.cpp with the Q4_K_XL GGUF, MTP speculative decoding, and community P100 patches that added about 50% decode; a single P100 on Q2_K_XL reached ~76 t/s generation and ~250 t/s prompt processing. Card count barely affects single-stream speed, and PCIe lane width (x16/x16, x16/x8, x8/x8) made no difference. The user corrects an earlier 3-card result that was invalidated by stuck 405 MHz core clocks.

reported speed:

70.0 tokens/s generation

quant:

Q4_K_XL (GGUF)

Reported by the source; GPU count, offloading and concurrent requests can change this figure. Check the full setup before comparing.

User reports almost 70 t/s with Qwen3.6-35B-A3B Q4_K_XL on a budget build using three Nvidia Tesla P100 16GB GPUs. The GPUs cost $80 each and are split across three nodes to manage thermals; the motherboard required a patched BIOS to enable Above 4G Decoding. User is still testing and optimizing, and has published a GitHub repository for the build.

reported speed:

40.0 tokens/s generation

quant:

Q6_K (GGUF)

Reported by the source; GPU count, offloading and concurrent requests can change this figure. Check the full setup before comparing.

User reports Qwen3.8 27B at 40+ t/s on two Tesla P100s, with a roughly 30 t/s average across long contexts and up to 55 t/s at 0 context. Setup is a custom llama.cpp fork with P100 kernel optimizations, Q6_K quant, both cards capped at 175W and communicating over PCIe gen 3. Speeds vary by about ±2 t/s; the user notes fp16 math saves 40% or more on prefill with negligible accuracy loss.