Can I Run Qwen 3.5 on NVIDIA L40S?
Autor: Jakub Rusinowski · Ostatnia aktualizacja: 21 lipca 2026
Yes, comfortably — Qwen 3.5 35B-A3B at Q4_K_M needs 23.8 GB at 8K context (21.1 GB weights + 2.7 GB KV cache/overhead), leaving ~24.2 GB of the NVIDIA L40S's 48 GB, at ~176 tok/s (est.), with room for up to 64K context.
See what else this hardware can run →
Jako partner Amazon zarabiamy na kwalifikujących się zakupach. Linki do chmurowych GPU są linkami polecającymi — możemy otrzymać prowizję bez dodatkowych kosztów dla Ciebie.
NVIDIA L40S Specs
| VRAM | 48 GB |
| Memory Bandwidth | 864 GB/s |
Qwen 3.5 35B-A3B on the NVIDIA L40S: VRAM by quantization
| Quant | VRAM needed | Fits 48 GB? | Max context | Whole PC |
|---|---|---|---|---|
| F16 | 72.7 GB | ✗ No | — | — |
| Q8_0 | 39.9 GB | ✓ Yes | 32K | Build for this |
| Q6_K | 31.4 GB | ✓ Yes | 64K | — |
| Q5_K_M | 27.5 GB | ✓ Yes | 64K | Build for this |
| Q4_K_M | 23.8 GB | ✓ Yes | 64K | Build for this |
| Q3_K_M | 17.6 GB | ✓ Yes | 128K | — |
| Q2_K | 14.2 GB | ✓ Yes | 128K | — |
Assumes an 8K-token context with an f16 KV cache, on an estimated architecture — this model publishes no config we can read. A longer window needs more; a quantized KV cache needs less. “Max context” is the largest window that still fits in 48 GB. Figures are estimates from parameter count, quantization and memory bandwidth — the analyzer lets you tune KV-cache quant and context.
Context costs VRAM too. At Q4_K_M on the NVIDIA L40S: 8K 23.8 GB · 32K 29.4 GB · 128K 52 GB ✗. Past 128K it no longer fits 48 GB — a q8_0 KV cache buys roughly half of that back, which the calculator will price for you.
No compatible GPU? Qwen 3.5 35B-A3B on 32 GB of system RAM, CPU only: runs, but it is tight.
Qwen 3.5 Sizes That Fit the NVIDIA L40S (at 8K context)
| Qwen 3.5 35B-A3B | Q4_K_M · 23.8 GB at 8K context · ~176 tok/s (est.) |
| Qwen 3.5 27B | Q4_K_M · 18.8 GB at 8K context · ~42 tok/s (est.) |
| Qwen 3.5 9B | Q4_K_M · 6.5 GB at 8K context · ~109 tok/s (est.) |
| Qwen 3.5 4B | Q4_K_M · 3.5 GB at 8K context · ~184 tok/s (est.) |
| Qwen 3.5 2B | Q4_K_M · 2.7 GB at 8K context · ~237 tok/s (est.) |
| Qwen 3.5 0.8B | Q4_K_M · 1.8 GB at 8K context · ~314 tok/s (est.) |
At 2 hrs/day, buying (~$7,499) beats renting at $0.34/hr after about 30.6 years.
Affiliate links — we may earn a commission if you sign up, at no extra cost to you.
Cloud rates verified 2026-07 — estimates, and marketplace prices vary. Buying price is GPU MSRP only, not a full PC.
FAQ
Can I run Qwen 3.5 on the NVIDIA L40S?
Yes, comfortably — Qwen 3.5 35B-A3B at Q4_K_M needs 23.8 GB at 8K context (21.1 GB weights + 2.7 GB KV cache/overhead), leaving ~24.2 GB of the NVIDIA L40S's 48 GB, at ~176 tok/s (est.), with room for up to 64K context.
What's the best Qwen 3.5 size for the NVIDIA L40S?
Qwen 3.5 35B-A3B at Q4_K_M quantization needs 23.8 GB at 8K context (21.1 GB weights + 2.7 GB KV cache/overhead), estimated ~176 tokens/sec, up to 64K context.
Every Qwen 3.5 size on the NVIDIA L40S
| Size | VRAM at 8K | Verdict | Speed |
|---|---|---|---|
| Qwen 3.5 397B-A17B | 139 GB | Does not fit | — |
| Qwen 3.5 122B-A10B | 77.3 GB | Runs at Q2_K | — |
| Qwen 3.5 35B-A3B | 23.8 GB | Runs | ~176 tok/s |
| Qwen 3.5 27B | 18.8 GB | Runs | ~42 tok/s |
| Qwen 3.5 9B | 6.5 GB | Runs | ~109 tok/s |
| Qwen 3.5 4B | 3.5 GB | Runs | ~184 tok/s |
| Qwen 3.5 2B | 2.7 GB | Runs | ~237 tok/s |
| Qwen 3.5 0.8B | 1.8 GB | Runs | ~314 tok/s |
Qwen 3.5 on Other GPUs
- Qwen 3.5 on NVIDIA RTX 6000 Ada Generation
- Qwen 3.5 on Apple M3 Pro
- Qwen 3.5 on NVIDIA GeForce RTX 5090
- Qwen 3.5 on Apple M4
- Qwen 3.5 on Apple M2 Pro
Popular Models on the NVIDIA L40S
VRAM Tier
Troubleshooting
Buying Guide
← Can I Run It? | Qwen 3.5 Model Page | NVIDIA L40S GPU Page | VRAM calculator | Check Your Hardware