Can K EXAONE 236B A23B run on Gaudi 3 128GB?

YES — With Q2_K

C48Usable

○Estimated from fit model

K EXAONE 236B A23B needs ~133.4 GB VRAM. Gaudi 3 128GB has 128.0 GB. With Q2_K quantization, expect ~17 tok/s.

Runtime: llama.cppCapacity: OffloadBandwidth: 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.

K EXAONE 236B A23B at Q4_K_M needs 185.3 GB — too much for Gaudi 3 128GB (128.0 GB). Runs at Q2_K (133.4 GB) with low quality.

Capabilities:

Select quantization to explore

Q4_K_M (Medium quality) — 185.3 GB, exceeds 128.0 GB available

185.3 GB required128.0 GB available

145% VRAM needed

57.3 GB over capacity — needs offload or smaller quantization

Fit status

Too heavy

Decode

6.5 tok/s

TTFT

29936 ms

Safe context

Memory

185.3 GB / 128.0 GB

Offload

30%

Memory breakdown

Weights144.0 GB

KV Cache27.7 GB

Runtime0.9 GB

Headroom12.8 GB

See how fast it feels

With memory offload — actual speed may be lower

See how fast it feelsK EXAONE 236B A23B on Gaudi 3 128GB
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: 6.5 tok/s decode · 29.9s TTFT (warm) · 16 tok/s prefill

What limits this setup

The raw memory story may look fine, but the software ecosystem is still a constraint here.

Very little memory headroom

You can run the model, but there is not much room left for longer context, bigger batches, extra apps, or future model updates.

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.

Buy headroom, not only minimum fit

A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	F	Too heavy	7.6 tok/s	13939 ms	4K
Coding	F	Too heavy	6.5 tok/s	29936 ms	4K
Agentic Coding	F	Too heavy	4.9 tok/s	57830 ms	4K
Reasoning	F	Too heavy	6.5 tok/s	35378 ms	4K
RAG	F	Too heavy	4.9 tok/s	72287 ms	4K

Quantization options

How K EXAONE 236B A23B (236B params) fits at each quantization level on Gaudi 3 128GB (128.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_KBest for your GPU	2	92.0 GB	Low	C48
Q3_K_S	3	115.6 GB	Low	F0

Get started

Copy-paste commands to run K EXAONE 236B A23B on your machine.

Run

lms load hf-lgai-exaone--k-exaone-236b-a23b-gguf && lms server start

Upgrade options

Hardware that runs K EXAONE 236B A23B well

Mac Studio M3 Ultra 256GBBest value

256 GB Unified (+128)

Makes the model fit on the accelerator instead of staying completely out of reach.3.4 tok/s decode

Makes the model fit on the accelerator instead of staying completely out of reach.

Adds memory headroom for longer context windows and future model growth.

~$6,999 MSRP

AMD Instinct MI350X 288GBBudget pick

288 GB VRAM (+160)8000 GB/s (+4300)

Makes the model fit on the accelerator instead of staying completely out of reach.40.6 tok/s decode