VOOZH about

URL: https://willitrunai.com/can-run/codestral-22b-on-arc-a770-16gb

⇱ Codestral 22B on Intel Arc A770 16GB? YES

Can Codestral 22B run on Intel Arc A770 16GB?

BARELY — Tight on Memory

C47Usable
○Estimated from fit model

Codestral 22B needs ~18.4 GB VRAM. Intel Arc A770 16GB has 16.0 GB. With Q4_K_M quantization, expect ~11 tok/s.

Runtime: llama.cppCapacity: OffloadBandwidth: MediumStack: StandardBottleneck: Host offload
Share:

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) — 18.4 GB, 11.3 tok/s, Very compromised (needs ~1.7 GB host RAM)
18.4 GB required16.0 GB available
115% VRAM needed

2.4 GB over capacity — needs offload or smaller quantization

Fit status

Very compromised (needs ~1.7 GB host RAM)

Decode

11.3 tok/s

TTFT

17086 ms

Safe context

4K

Memory

18.4 GB / 16.0 GB

Offload

10%

Memory breakdown

Weights13.4 GB
KV Cache2.4 GB
Runtime0.9 GB
Headroom1.6 GB

See how fast it feels

See how fast it feelsCodestral 22B on Intel Arc A770 16GB
1st promptCold start — includes initialization
>Write a Python function to sort a list
Here's a Python function that sorts a list using quicksort: def quicksort(arr): if len(arr) <= 1: return arr pivot = arr[len(arr) // 2] left = [x for x in arr if x < pivot] middle = [x for x in arr if x == pivot] right = [x for x in arr if x > pivot] return quicksort(left) + middle + quicksort(right) # Example usage numbers = [3, 6, 8, 10, 1, 2, 1] print(quicksort(numbers)) # [1, 1, 2, 3, 6, 8, 10] The algorithm picks a pivot, partitions the list into smaller, equal, and larger elements, then recursively sorts each partition and concatenates the results.
2nd promptWarm — model ready, no init overhead
>Now add type hints and a docstring
def quicksort(arr: list[int]) -> list[int]: """Sort a list using the quicksort algorithm. Args: arr: List of integers to sort. Returns: A new sorted list. """ if len(arr) <= 1: return arr pivot = arr[len(arr) // 2] left = [x for x in arr if x < pivot] middle = [x for x in arr if x == pivot] right = [x for x in arr if x > pivot] return quicksort(left) + middle + quicksort(right) Added type hints for the input and return type, plus a Google-style docstring.
Estimated: 11.3 tok/s decode · 17.1s TTFT (warm) · 28 tok/s prefill

What limits this setup

It fits through host-memory offload, and offload is the main reason performance drops.

CPU or host-memory offload is active

About 10% of the working set spills out of accelerator memory, which usually hurts latency and sustained decode throughput.

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

Remove offload with more accelerator memory

Prioritize a GPU or unified-memory tier that fits the whole model natively. Removing offload usually helps more than small compute gains.

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

WorkloadGradeFitDecodeTTFTContext
ChatCRuns with offload (needs ~0.9 GB host RAM)13.1 tok/s8063 ms4K
CodingCVery compromised (needs ~1.7 GB host RAM)11.3 tok/s17086 ms4K
Agentic CodingFToo heavy8.7 tok/s32321 ms4K
ReasoningCVery compromised (needs ~1.7 GB host RAM)11.3 tok/s20192 ms4K
RAGFToo heavy8.7 tok/s40402 ms4K

Quantization options

How Codestral 22B (22B params) fits at each quantization level on Intel Arc A770 16GB (16.0 GB usable).

QuantBitsVRAMQualityFit
Q2_K
2
8.6 GB
LowB61
Q3_K_SBest for your GPU
3
10.8 GB
LowB61
NVFP4
4
12.3 GB
MediumF0
Q4_K_M
4
13.4 GB
MediumF0
Q5_K_M
5
15.8 GB
HighF0
Q6_K
6
18.0 GB
HighF0
Q8_0
8
23.5 GB
Very HighF0
F16
16
45.1 GB
MaximumF0

Get started

Copy-paste commands to run Codestral 22B on your machine.

Run

ollama run codestral

Upgrade options

Hardware that runs Codestral 22B well

👁 Intel
Intel Arc Pro B60 24GBBudget pick
24 GB VRAM (+8)
B
Removes host-memory offload, which is usually the single biggest latency and throughput win.19.7 tok/s decode

Removes host-memory offload, which is usually the single biggest latency and throughput win.

Raises estimated decode speed by about 74%.

~$599 MSRP

👁 Intel
Intel Data Center GPU Max 1550 128GBBest value
128 GB VRAM (+112)3200 GB/s (+2640)
B
Removes host-memory offload, which is usually the single biggest latency and throughput win.161.5 tok/s decode

Removes host-memory offload, which is usually the single biggest latency and throughput win.

Raises estimated decode speed by about 1329%.

~$15,000 MSRP

👁 Intel
Gaudi 3 128GBIntel upgrade
128 GB VRAM (+112)3700 GB/s (+3140)
B
Removes host-memory offload, which is usually the single biggest latency and throughput win.207.5 tok/s decode

Removes host-memory offload, which is usually the single biggest latency and throughput win.

Raises estimated decode speed by about 1736%.

~$15,000 MSRP

Frequently asked questions

See all results for Intel Arc A770 16GBSee all hardware for Codestral 22B