Llama 4 Scout 17B 16E needs ~83.1 GB VRAM. Gaudi 3 128GB has 128.0 GB. With Q4_K_M quantization, expect ~99 tok/s.
Operating mode
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
Fit status
Runs well
Decode
99.0 tok/s
TTFT
1955 ms
Safe context
261K
Memory
83.1 GB / 128.0 GB
The raw memory story may look fine, but the software ecosystem is still a constraint here.
Runtime ecosystem is narrower than CUDA
Intel GPUs can look attractive on memory per dollar, but local AI tooling, kernels, and model coverage are still broader and easier on CUDA today.
Prefer CUDA if you want the path of least resistance
If your goal is maximum runtime coverage, easier troubleshooting, and better support for new local AI releases, CUDA is usually still the safer upgrade path.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | A | Runs well | 99.0 tok/s | 1066 ms | 261K |
| Coding | A | Runs well | 99.0 tok/s | 1955 ms | 261K |
| Agentic Coding | A | Runs well | 99.0 tok/s | 2843 ms | 261K |
| Reasoning | A | Runs well | 99.0 tok/s | 2310 ms | 261K |
| RAG | A | Runs well | 99.0 tok/s | 3554 ms | 261K |
How Llama 4 Scout 17B 16E (109B params) fits at each quantization level on Gaudi 3 128GB (128.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 42.5 GB | Low | A71 |
Q3_K_S | 3 | 53.4 GB | Low | A73 |
NVFP4 | 4 |
Copy-paste commands to run Llama 4 Scout 17B 16E on your machine.
Run
lms load Llama-4-Scout-17B-16E-Instruct && lms server startYour hardware
| Model | Params | Grade | Decode | Capabilities |
|---|---|---|---|---|
 Devstral 2 123B Instruct | 123B | S | 37.5 tok/s | |
 Qwen 3.5 122B A10B | 122B | S |
Yes, Gaudi 3 128GB can run Llama 4 Scout 17B 16E with a A grade (Runs well). Expected decode speed: 99.0 tok/s.
Llama 4 Scout 17B 16E (109B parameters) requires approximately 83.1 GB of memory with Q4_K_M quantization.
The recommended quantization for Llama 4 Scout 17B 16E is Q4_K_M, which balances quality and memory efficiency.
On Gaudi 3 128GB, Llama 4 Scout 17B 16E achieves approximately 99.0 tokens per second decode speed with a time-to-first-token of 1955ms using Q4_K_M quantization.
For coding workloads, Llama 4 Scout 17B 16E on Gaudi 3 128GB receives a A grade with 99.0 tok/s and 261K context.
On Gaudi 3 128GB, Llama 4 Scout 17B 16E can safely use up to 261K tokens of context. The model's official context limit is 10.5M, but available memory constrains the safe maximum.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/llama-4-scout-17b-16e-on-gaudi-3-128gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview:
61.0 GB |
| Medium |
| A74 |
Q4_K_M | 4 | 66.5 GB | Medium | A75 |
Q5_K_M | 5 | 78.5 GB | High | A76 |
Q6_KBest for your GPU | 6 | 89.4 GB | High | A76 |
Q8_0 | 8 | 116.6 GB | Very High | F0 |
F16 | 16 | 223.5 GB | Maximum | F0 |
| 104.1 tok/s |
| Mistral Small 4 119B | 119B | S | 112.9 tok/s |
 GPT-OSS 120B | 117B | S | 39.5 tok/s |
 Command A 111B | 111B | S | 41.8 tok/s |
Prefer CUDA if you want the path of least resistance. If your goal is maximum runtime coverage, easier troubleshooting, and better support for new local AI releases, CUDA is usually still the safer upgrade path.
Often yes, if your goal is the easiest setup and the widest runtime support. Intel can offer attractive memory capacity, but CUDA still tends to win on tooling maturity, guides, kernels, and model coverage for local AI.