AMD · 20 GB GDDR6 · 800 GB/s bandwidth ·
Alibaba · 34.7B (active 3.0B) · MoE · model ctx up to 256K
No — Qwen3.5-35B-A3B at Q4 needs ≈ 20.39 GB even at 4K context, beyond the RX 7900 XT 20GB's 20 GB GDDR6. It would only run with heavy CPU/disk offload. Decode at Q4/8K: ≈ 453.0 tok/s (estimated, batch 1). FP16 needs ≈ 71.0 GB — does not fit on this card. Estimates come from memory-bandwidth math; rows tagged measured override estimates. Relative ranking is more reliable than absolute tok/s. Methodology.
Fit & speed by quant and context
Quant
Context
VRAM needed
Fits?
tok/s (decode)
Q4
4K
20.39 GB
No
466.5est.
Q4
8K default
20.69 GB
No
453.0est.
Q4
32K
22.49 GB
No
383.4est.
FP16
4K
70.7 GB
No
10.7est.
FP16
8K default
71.0 GB
No
10.6est.
FP16
32K
72.8 GB
No
10.1est.
Fits = weights + KV(ctx) + 1 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).