Can I run Qwen2.5 32B on NVIDIA RTX 3090 24GB?
Yes: Qwen2.5 32B runs on NVIDIA RTX 3090 24GB with IQ4_XS at roughly 31 tokens/s. Estimated memory 19.5 GiB at 8,192 tokens of context.
With IQ4_XS you can push the context to about 11,008 tokens before memory runs out (F16 KV cache).
Memory by quantisation (8,192-token context)
| Quantisation | Memory | Verdict | Est. speed |
|---|---|---|---|
| F16Unquantized reference | 64.3 GiB | Does not fit | — |
| Q8_0Near-lossless | 35.7 GiB | Needs CPU offload | — |
| Q6_KVery close to F16 | 28.3 GiB | Needs CPU offload | — |
| Q5_K_MGreat quality/size | 25.0 GiB | Needs CPU offload | — |
| Q4_K_MBest default for most | 21.8 GiB | Tight | ~28 tok/s |
| IQ4_XSSmaller 4-bit (imatrix) | 19.5 GiB | Fits | ~31 tok/s |
| Q3_K_MNoticeable loss | 18.2 GiB | Fits | ~34 tok/s |
| Q2_KLast resort; big loss | 16.1 GiB | Fits | ~38 tok/s |
NVIDIA RTX 3090 24GB offers about 24 GiB of memory (94% usable for a model) at 936 GB/s. Speed is a bandwidth-bound estimate of single-stream generation at 65% efficiency; real results vary by runtime, drivers and settings, so measure on your machine.
How this is calculated
Weights ≈ parameters × bits per weight ÷ 8. KV cache ≈ 2 × 64 layers × 8 KV heads × 128 head dim × 2 bytes per token. Decode speed ≈ efficiency × memory bandwidth ÷ bytes read per token. Learn the maths in VRAM math and memory bandwidth, or try the VRAM calculator.
Qwen2.5 32B on other hardware
Similar models on NVIDIA RTX 3090 24GB
Running slowly or out of memory? See CUDA out of memory and the quantisation lesson.