MacBook Pro M4 Max 96GB经济之选
96 GB Unified (+48)
B
Adds memory headroom for longer context windows and future model growth.23.8 tok/s 解码
Adds memory headroom for longer context windows and future model growth.
~$2,499 MSRP
Can Qwen 3 1.7B run on NVIDIA L40S 48GB?
YES — Runs Great
B61Good
Estimated from fit model
Qwen 3 1.7B needs ~8.4 GB VRAM. NVIDIA L40S 48GB has 48.0 GB. With Q4_K_M quantization, expect ~26 tok/s.
Runtime: llama.cppCapacity: RoomyBandwidth: HighStack: StandardBottleneck: Balanced
Operating mode
Choose the run profile you care about
Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.
Current mode
Balanced
Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.
Select quantization to explore
Q4_K_M (Medium quality) — 8.4 GB, 25.5 tok/s, Runs well
8.4 GB required48.0 GB available
18% VRAM used
Fit status
Runs well
Decode
25.5 tok/s
TTFT
7592 ms
Safe context
33K
Memory
8.4 GB / 48.0 GB
Memory breakdown
Weights1.0 GB
KV Cache1.7 GB
Runtime0.9 GB
Headroom4.8 GB
See how fast it feels
See how fast it feelsQwen 3 1.7B on NVIDIA L40S 48GB
1st promptCold start — includes initialization
>What is local AI inference?
Local AI inference means running an AI language model directly on your own hardware — your laptop, desktop, or server — instead of sending requests to a remote cloud API.
When you run inference locally the model weights are loaded into your GPU or unified memory. Each token you generate requires reading those weights from memory, so memory bandwidth is the main bottleneck for decode speed.
Key benefits of running locally:
- Full privacy: your prompts never leave your machine
- No per-token cost or rate limits
- Works offline once the model is downloaded
- Latency depends only on your hardware
2nd promptWarm — model ready, no init overhead
>How much VRAM do I need?
It depends on the model size and quantization level. A rough rule of thumb:
Model size Q4 (4-bit) Q8 (8-bit) FP16
7B params ~4.3 GB ~7.5 GB ~14 GB
13B params ~7.9 GB ~13.9 GB ~26 GB
70B params ~42.7 GB ~74.9 GB ~140 GB
Most people use 4-bit quantization (Q4_K_M) which gives 90-95% of full quality at a fraction of the memory. A 24 GB GPU can comfortably run most 7B-13B models.
Estimated: 25.5 tok/s decode · 7.6s TTFT (warm) · 64 tok/s prefill
What limits this setup
This setup is broadly balanced for this model.
No major red flags
This recommendation has enough memory headroom and acceptable estimated speed for the selected workload.
Best improvement path
Performance by workload
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | B | Runs well | 25.5 tok/s | 4141 ms | 33K |
| Coding | B | Runs well | 25.5 tok/s | 7592 ms | 33K |
| Agentic Coding | B | Runs well | 25.5 tok/s | 11043 ms | 33K |
| Reasoning | B | Runs well | 25.5 tok/s | 8973 ms | 33K |
| RAG | B | Runs well | 25.5 tok/s | 13804 ms | 33K |
Inference speed
Qwen 3 1.7B inference speed — tokens per second by GPU & Mac
Estimated decode speed (tokens/sec) for Qwen 3 1.7B at Q4_K_M across popular GPUs and Apple Silicon, using the fastest local runtime per device. Fastest is RTX 5090 32GB at ~32 tok/s. Speed is memory-bandwidth bound, so cards that fit the whole model in VRAM run far faster than ones that offload to system RAM.
| GPU / Mac | Memory | Quant | Speed (tok/s) | Fits? |
|---|---|---|---|---|
 RTX 5090 32GB | 32 GB | Q4_K_M | 32.3 | Fits |
| RTX 4080 Super 16GB | 16 GB | Q4_K_M | 27.2 | Fits |
| RTX 4090 24GB | 24 GB | Q4_K_M | 23.8 | Fits |
| RTX 3090 24GB | 24 GB | Q4_K_M | 23.8 | Fits |
| RTX 4070 12GB | 12 GB | Q4_K_M | 23.8 | Fits |
| RTX 3060 12GB | 12 GB | Q4_K_M | 23.8 | Fits |
| RTX 4060 8GB | 8 GB | Q4_K_M | 23.8 | Fits |
RX 7900 XTX 24GB | 24 GB | Q4_K_M | 23.8 | Fits |
MacBook Pro M4 Max 128GB | 128 GB | Q4_K_M | 23.8 | Fits |
Mac Studio M3 Ultra 256GB | 256 GB | Q4_K_M | 23.8 | Fits |
Mac Studio M2 Ultra 128GB | 128 GB | Q4_K_M | 23.8 | Fits |
Mac Studio M1 Ultra 128GB | 128 GB | Q4_K_M | 23.8 | Fits |
MacBook Pro M4 Max 64GB | 64 GB | Q4_K_M | 23.8 | Fits |
MacBook Pro M3 Max 64GB | 64 GB | Q4_K_M | 23.8 | Fits |
MacBook Pro M1 Max 64GB | 64 GB | Q4_K_M | 23.8 | Fits |
MacBook Pro M4 Pro 48GB | 48 GB | Q4_K_M | 23.8 | Fits |
Estimates for single-stream decoding at Q4_K_M; real tokens/sec varies with prompt length, context, batch size, and runtime build. Prompt processing (prefill) is faster than the decode figures shown here.
Quantization options
How Qwen 3 1.7B (1.7000000476837158B params) fits at each quantization level on NVIDIA L40S 48GB (48.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 0.7 GB | Low | B61 |
Q3_K_S | 3 | 0.8 GB | Low | B61 |
NVFP4 | 4 | 1.0 GB | Medium | B61 |
Q4_K_M | 4 | 1.0 GB | Medium | B61 |
Q5_K_M | 5 | 1.2 GB | High | B61 |
Q6_K | 6 | 1.4 GB | High | B61 |
Q8_0 | 8 | 1.8 GB | Very High | B61 |
F16Best for your GPU | 16 | 3.5 GB | Maximum | B61 |
Get started
Copy-paste commands to run Qwen 3 1.7B on your machine.
Run
ollama run qwen3:1.7b升级选项
能流畅运行 Qwen 3 1.7B 的硬件
Frequently asked questions
Can NVIDIA L40S 48GB run Qwen 3 1.7B?
Yes, NVIDIA L40S 48GB can run Qwen 3 1.7B with a B grade (Runs well). Expected decode speed: 25.5 tok/s.
How much VRAM does Qwen 3 1.7B need?
Qwen 3 1.7B (1.7000000476837158B parameters) requires approximately 8.4 GB of memory with Q4_K_M quantization.
What is the best quantization for Qwen 3 1.7B?
The recommended quantization for Qwen 3 1.7B is Q4_K_M, which balances quality and memory efficiency.
What speed will Qwen 3 1.7B run at on NVIDIA L40S 48GB?
On NVIDIA L40S 48GB, Qwen 3 1.7B achieves approximately 25.5 tokens per second decode speed with a time-to-first-token of 7592ms using Q4_K_M quantization.
Can NVIDIA L40S 48GB run Qwen 3 1.7B for coding?
For coding workloads, Qwen 3 1.7B on NVIDIA L40S 48GB receives a B grade with 25.5 tok/s and 33K context.
What context window can Qwen 3 1.7B use on NVIDIA L40S 48GB?
On NVIDIA L40S 48GB, Qwen 3 1.7B can safely use up to 33K tokens of context. The model's official context limit is 33K, but available memory constrains the safe maximum.
Embed this result▼
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/qwen-3-1.7b-on-l40s-48gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview: