Back to search results
Qwen2.5-0.5B-Instruct · Q8_0
NVIDIA RTX 509032GBllama.cpp v0.24L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen2.5-0.5B-Instruct (Q8_0) on NVIDIA RTX 5090 with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
358.97 tok/s
Decode
Decode speed
1,824.25 tok/s
Prefill
Prefill speed
1.68 s
TTFT
Time to first token
3.31 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
2 days ago
Last verified
Performance
- llama.cpp · IQ4_XS Decode 319.07 · Prefill 1265.71 ·
- llama.cpp · Q8_0 (current)Decode 358.97 · Prefill 1824.25 ·
- vLLM · IQ3_XS Decode 285.37 · Prefill 1069.59 ·
Core figures
- Decode (avg)
- 358.97 tok/s
- Prefill (avg)
- 1,824.25 tok/s
- TTFT (avg)
- 1.68 s
- VRAM (avg)
- 3.31 GB
- MTP acceptance rate
- —
- TTFB
- 2.87 GB
- Power draw
- 340.5 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen2.5-0.5B-Instruct
- Quantization
- Q8_0
- Framework
- llama.cpp
- Version
- v0.24
- Context length
- 8192 tokens
- Batch size
- 256
- GPU layers
- 99
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- NVIDIA RTX 5090
- Nominal VRAM
- 32 GB
- Measured VRAM (avg)
- 3.31 GB
- OS
- Ubuntu 24.04
- Driver
- 570.86
- CUDA
- 12.4
- Power draw
- 340.5 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.24 · Ubuntu 24.04 · CUDA 12.4 · 8192 ctx
358.97 tok/s
Decode
1,824.25 tok/s
Prefill
1.68 s
TTFT
3.31 GB
VRAM
—
MTP
340.5 W
Power draw
L2 batch seed 2026-10-16: llama.cpp on hw_id=13 running model_id=59 at Q8_0