Would CUDA be a better path than Gaudi 3 128GB for Pixtral Large 124B?

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.

Can Pixtral Large 124B run on Gaudi 3 128GB?

Q: What should I upgrade first if Pixtral Large 124B feels slow on Gaudi 3 128GB?

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.

YES — Runs Great

S93Excellent

Estimated from fit model

Pixtral Large 124B needs ~94.7 GB VRAM. Gaudi 3 128GB has 128.0 GB. With Q4_K_M quantization, expect ~37 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.

Capabilities:

Select quantization to explore

Q4_K_M (Medium quality) — 94.7 GB, 37.2 tok/s, Runs well

94.7 GB required128.0 GB available

74% VRAM used

Fit status

Runs well

Decode

37.2 tok/s

TTFT

5199 ms

Safe context

115K

Memory

94.7 GB / 128.0 GB

Memory breakdown

Weights75.6 GB

KV Cache5.4 GB

Runtime0.9 GB

Headroom12.8 GB

See how fast it feels

See how fast it feelsPixtral Large 124B on Gaudi 3 128GB
1st promptCold start — includes initialization
>Solve: If a train travels at 60 mph for 2.5 hours, how far does it go?
Let me work through this step by step.

Given information:
  - Speed = 60 mph (miles per hour)
  - Time  = 2.5 hours

Formula: Distance = Speed x Time

Calculation:
  Distance = 60 mph x 2.5 h
           = 150 miles

The train travels 150 miles.

Double-check: at 60 mph the train covers 60 miles each hour. In half an hour it covers 30 miles. So in 2.5 hours: 60 + 60 + 30 = 150 miles. Confirmed.
2nd promptWarm — model ready, no init overhead
>What if the train stops for 20 minutes halfway?
The stop doesn't change the total distance covered while moving, but it affects total trip time.

Breakdown:
  - First half: 1.25 hours at 60 mph = 75 miles
  - Stop: 20 minutes (no distance)
  - Second half: 1.25 hours at 60 mph = 75 miles

Total distance: 150 miles (same)
Total time: 2.5 h + 0.33 h = 2.83 hours
Average speed: 150 / 2.83 = 53 mph

The distance stays the same but average speed drops to 53 mph because of the stop.
Estimated: 37.2 tok/s decode · 5.2s TTFT (warm) · 93 tok/s prefill

What limits this setup

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.

Best improvement path

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.

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	S	Runs well	37.2 tok/s	2836 ms	115K
Coding	S	Runs well	37.2 tok/s	5199 ms	115K
Agentic Coding	S	Runs well	37.2 tok/s	7562 ms	115K
Reasoning	S	Runs well	37.2 tok/s	6144 ms	115K
RAG	S	Runs well	37.2 tok/s	9453 ms	115K

Inference speed

Pixtral Large 124B inference speed — tokens per second by GPU & Mac

Estimated decode speed (tokens/sec) for Pixtral Large 124B at Q4_K_M across popular GPUs and Apple Silicon, including multi-GPU rigs, using the fastest local runtime per device. Fastest is MacBook Pro M4 Max 128GB at ~8 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?
MacBook Pro M4 Max 128GB	128 GB	Q4_K_M	8.0	Offloads

Quantization options

How Pixtral Large 124B (124B params) fits at each quantization level on Gaudi 3 128GB (128.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	48.4 GB	Low	A84
Q3_K_S	3	60.8 GB	Low	S86
NVFP4	4

Get started

Copy-paste commands to run Pixtral Large 124B on your machine.

Run

lms load Pixtral-Large-Instruct-2411 && lms server start

Frequently asked questions

Can Gaudi 3 128GB run Pixtral Large 124B?

Yes, Gaudi 3 128GB can run Pixtral Large 124B with a S grade (Runs well). Expected decode speed: 37.2 tok/s.

How much VRAM does Pixtral Large 124B need?

Pixtral Large 124B (124B parameters) requires approximately 94.7 GB of memory with Q4_K_M quantization.

What is the best quantization for Pixtral Large 124B?

The recommended quantization for Pixtral Large 124B is Q4_K_M, which balances quality and memory efficiency.

What speed will Pixtral Large 124B run at on Gaudi 3 128GB?

On Gaudi 3 128GB, Pixtral Large 124B achieves approximately 37.2 tokens per second decode speed with a time-to-first-token of 5199ms using Q4_K_M quantization.

Can Gaudi 3 128GB run Pixtral Large 124B for coding?

For coding workloads, Pixtral Large 124B on Gaudi 3 128GB receives a S grade with 37.2 tok/s and 115K context.

What context window can Pixtral Large 124B use on Gaudi 3 128GB?

On Gaudi 3 128GB, Pixtral Large 124B can safely use up to 115K tokens of context. The model's official context limit is 131K, but available memory constrains the safe maximum.

What should I upgrade first if Pixtral Large 124B feels slow on Gaudi 3 128GB?

See all results for Gaudi 3 128GB See all hardware for Pixtral Large 124B

Embed this result▼

Paste this snippet into any page to show a live fit card.

<iframe src="https://willitrunai.com/embed/pixtral-large-124b-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: