Back to search results
gemma-4-E2B-it · IQ3_XS
2x AMD Strix HaloStrata v0.50L2 Cross-framework verified
This page aggregates 1 real-world runs of gemma-4-E2B-it (IQ3_XS) on 2x AMD Strix Halo with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.
321.59 tok/s
Decode
Decode speed
1,311.64 tok/s
Prefill
Prefill speed
2.09 s
TTFT
Time to first token
4.7 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
2 days ago
Last verified
Performance
- vLLM · Q3_K_S Decode 277.77 · Prefill 1024.78 ·
- Strata · IQ3_XS (current)Decode 321.59 · Prefill 1311.64 ·
Core figures
- Decode (avg)
- 321.59 tok/s
- Prefill (avg)
- 1,311.64 tok/s
- TTFT (avg)
- 2.09 s
- VRAM (avg)
- 4.7 GB
- MTP acceptance rate
- 0.48
- TTFB
- 2.57 GB
- Power draw
- 69.4 W
Configuration
Member-level fields are taken from the most recent run
- Model
- gemma-4-E2B-it
- Quantization
- IQ3_XS
- Framework
- Strata
- Version
- v0.50
- Context length
- 4096 tokens
- Batch size
- 512
- GPU layers
- 8
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- 2x AMD Strix Halo
- Measured VRAM (avg)
- 4.7 GB
- OS
- Ubuntu 24.04
- Driver
- amdgpu-pro 24.20
- CUDA
- N/A
- Power draw
- 69.4 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.50 · Ubuntu 24.04 · CUDA N/A · 4096 ctx
321.59 tok/s
Decode
1,311.64 tok/s
Prefill
2.09 s
TTFT
4.7 GB
VRAM
0.48
MTP
69.4 W
Power draw
L2 batch seed 2026-10-16: Strata on hw_id=88 running model_id=66 at IQ3_XS