Q4_K_M: Smooth Fit 7.94 GB / 31.4 GB Usable
Qwen3-8B on NVIDIA GeForce RTX 5090 32GB: VRAM Requirement & Speed Estimate
Yes. At Q4_K_M 4-bit quantization with a 16K context window, Qwen3-8B requires 7.94 GB of VRAM (4.97 GB weights + 2.42 GB KV cache + 0.55 GB overhead), fitting within RTX 5090's 31.4 GB usable memory at ~208.8 tokens/s.
Concurrency (Batch):
Dense 8.2B·36L (d=4096)·32Q/8KV·Ctx 128K
Smooth Fit7.94 / 31.4 GB~208.8 tok/s· Q4_K_M · 16K ctx
3. Weight Quantization
4. Per-Request Context16K (16,384 tok)
Verified: 2026-11-20·Roofline bandwidth/compute estimate (Batch=1)
Smooth Fit
7.94/ 31.4 GB usable
Physical 32 GB · Load 25.3%
Theoretical Decode Speed
~208.8tokens/s
Actual speed varies ~±25% by runtime
Max Context at Batch=1
128K/ 128K
Enable Q8/Q4 KV quantization to scale batch concurrency
Weights: 4.97 GBKV Cache (1×16K): 2.42 GBOverhead: 0.55 GB
🚀 Blazing Fast (Instant CoT)~2.4s for a 500-token response