Back to search results
Llama 5 405B · IQ2_XXS
NVIDIA DGX Spark128GBvLLM 0.7.3L0 Self-reported
This page aggregates 1 real-world runs of Llama 5 405B (IQ2_XXS) on NVIDIA DGX Spark with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.
2 tok/s
Decode
Decode speed
15 tok/s
Prefill
Prefill speed
6.67 s
TTFT
Time to first token
115.3 GB
VRAM
VRAM usage
L0 Self-reported
1
Measured runs
1
Independent sources
Other
Source platforms
Today
Last verified
Performance
- vLLM · IQ3_S Decode 1.4 · Prefill 10.5 ·
- vLLM · IQ2_XS Decode 1.7 · Prefill 12.8 ·
- vLLM · IQ2_XXS (current)Decode 2 · Prefill 15 ·
Core figures
- Decode (avg)
- 2 tok/s
- Prefill (avg)
- 15 tok/s
- TTFT (avg)
- 6.67 s
- VRAM (avg)
- 115.3 GB
- MTP acceptance rate
- —
- TTFB
- — GB
- Power draw
- — W
Configuration
Member-level fields are taken from the most recent run
- Model
- Llama 5 405B
- Quantization
- IQ2_XXS
- Framework
- vLLM
- Version
- 0.7.3
- Context length
- 4096 tokens
- Batch size
- 512
- 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 DGX Spark
- Nominal VRAM
- 128 GB
- Measured VRAM (avg)
- 115.3 GB
- OS
- Linux
- Driver
- 570.65
- CUDA
- 12.6
Sources & evidence
1 measured records in total, each traceable to its original source
0.7.3 · Linux · CUDA 12.6 · 4096 ctx
2 tok/s
Decode
15 tok/s
Prefill
6.67 s
TTFT
115.3 GB
VRAM
—
MTP
— W
Power draw
L0 理论估算（add-500-records-batch2）