← AI Hashrate 中文

Can Arc A770 16GB run gemma-4-26B-A4B-it?

Intel · 16 GB GDDR6 · 560 GB/s bandwidth · Google · 25.2B (active 3.8B) · MoE · model ctx up to 256K

✅ Yes — Arc A770 16GB (16 GB) can run gemma-4-26B-A4B-it at Q4 (8K)
VRAM: 15.42 GB needed of 16 GB
Speed: ~136.7 tok/s (estimated)
TT · Tight FitComposite score: VRAM headroom × 0.5 + tok/s speed × 0.5. Q4 @ 8K context.
ONEAPI
Tight fit — try smaller model or shorter contextTight fit — try smaller model or shorter contextTight fit — try smaller model or shorter context
Estimates from memory-bandwidth formula; rows marked "measured" override. Methodology.

Fit & speed by quant and context

QuantContextVRAM neededFits?tok/s (decode)
Q44K15.27 GBTight136.7 ± est.
Q48K default15.42 GBTight136.7 ± est.
Q432K16.36 GBNo236.3 ± est.
FP164K51.56 GBNo6.9 ± est.
FP168K default51.71 GBNo6.9 ± est.
FP1632K52.65 GBNo6.7 ± est.

Fits = weights + KV(ctx) + 1.5 GB overhead ≤ 95% of VRAM. Measured anchors are context-agnostic; the fit verdict is recomputed per context. A missing tok/s means the model is far beyond this card (offload-only territory).

VRAM breakdown at 8K context

QuantWeightsKV cacheOverheadTotal neededArc A770 16GB VRAM
Q414.11 GB0.31 GB1.5 GB15.42 GB16 GB
FP1650.4 GB0.31 GB1.5 GB51.71 GB16 GB

Other GPUs that run gemma-4-26B-A4B-it

All GPUs for gemma-4-26B-A4B-it →

Other models for the Arc A770 16GB

All models on the Arc A770 16GB →