Back to search results
Llama-3.2-3B-Instruct · Q5_K_M
NVIDIA H100 80GB HBM380GBllama.cpp v0.21L2 Cross-framework verified
This page aggregates 1 real-world runs of Llama-3.2-3B-Instruct (Q5_K_M) on NVIDIA H100 80GB HBM3 with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
313.79 tok/s
Decode
Decode speed
1,508.71 tok/s
Prefill
Prefill speed
1.01 s
TTFT
Time to first token
7.96 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
16 days ago
Last verified
Performance
- llama.cpp · FP16 Decode 379.95 · Prefill 1874.2 ·
- llama.cpp · Q5_K_M (current)Decode 313.79 · Prefill 1508.71 ·
- Ollama · FP16 Decode 308.75 · Prefill 1127.72 ·
- Ollama · IQ2_XS Decode 295.44 · Prefill 1300.71 ·
Core figures
- Decode (avg)
- 313.79 tok/s
- Prefill (avg)
- 1,508.71 tok/s
- TTFT (avg)
- 1.01 s
- VRAM (avg)
- 7.96 GB
- MTP acceptance rate
- —
- TTFB
- 2.51 GB
- Power draw
- 602.6 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Llama-3.2-3B-Instruct
- Quantization
- Q5_K_M
- Framework
- llama.cpp
- Version
- v0.21
- Context length
- 2048 tokens
- Batch size
- 512
- GPU layers
- 28
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- NVIDIA H100 80GB HBM3
- Nominal VRAM
- 80 GB
- Measured VRAM (avg)
- 7.96 GB
- OS
- Ubuntu 24.04
- Driver
- 572.83
- CUDA
- 13.0
- Power draw
- 602.6 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.21 · Ubuntu 24.04 · CUDA 13.0 · 2048 ctx
313.79 tok/s
Decode
1,508.71 tok/s
Prefill
1.01 s
TTFT
7.96 GB
VRAM
—
MTP
602.6 W
Power draw
L2 batch seed 2026-10-19: llama.cpp on hw_id=55 running model_id=70 at Q5_K_M