Will It Run AI

Can Qwen 3.5 35B A3B run on Tesla P40 24GB?

BARELY — Tight on Memory

A80Great
Estimated from fit model

Qwen 3.5 35B A3B needs ~26.1 GB VRAM. Tesla P40 24GB has 24.0 GB. With Q4_K_M quantization, expect ~17 tok/s.

Runtime: llama.cppCapacity: OffloadBandwidth: LowStack: 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) — 26.1 GB, 17.1 tok/s, Very compromised (needs ~1.7 GB host RAM)
26.1 GB required24.0 GB available
109% VRAM needed

2.1 GB over capacity — needs offload or smaller quantization

Fit status

Very compromised (needs ~1.7 GB host RAM)

Decode

17.1 tok/s

TTFT

11325 ms

Safe context

4K

Memory

26.1 GB / 24.0 GB

Offload

10%

Memory breakdown

Weights21.3 GB
KV Cache1.5 GB
Runtime0.9 GB
Headroom2.4 GB

See how fast it feels

See how fast it feelsQwen 3.5 35B A3B on Tesla P40 24GB
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: 17.1 tok/s decode · 11.3s TTFT (warm) · 43 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.

Older PCIe generation

PCIe 3.0 is workable, but it compounds the penalty when you offload heavily or try to scale across multiple cards.

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.

Buy headroom, not only minimum fit

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

Increase host RAM if you keep offloading

This setup may need roughly 1.7 GB of extra host RAM just for the offloaded portion, before OS and other tools.

Performance by workload

WorkloadGradeFitDecodeTTFTContext
ChatARuns with offload (needs ~1.2 GB host RAM)18.2 tok/s5803 ms4K
CodingAVery compromised (needs ~1.7 GB host RAM)17.1 tok/s11325 ms4K
Agentic CodingAVery compromised (needs ~2.8 GB host RAM)15.2 tok/s18570 ms4K
ReasoningAVery compromised (needs ~1.7 GB host RAM)17.1 tok/s13384 ms4K
RAGAVery compromised (needs ~2.8 GB host RAM)15.2 tok/s23213 ms4K

Quantization options

How Qwen 3.5 35B A3B (35B params) fits at each quantization level on Tesla P40 24GB (24.0 GB usable).

QuantBitsVRAMQualityFit
Q2_K
2
13.7 GB
LowS92
Q3_K_SBest for your GPU
3
17.2 GB
LowS91
NVFP4
4
19.6 GB
MediumF0
Q4_K_M
4
21.3 GB
MediumF0
Q5_K_M
5
25.2 GB
HighF0
Q6_K
6
28.7 GB
HighF0
Q8_0
8
37.5 GB
Very HighF0
F16
16
71.8 GB
MaximumF0

Get started

Copy-paste commands to run Qwen 3.5 35B A3B on your machine.

Run

ollama run qwen3.5:35b-a3b

Frequently asked questions

Can Tesla P40 24GB run Qwen 3.5 35B A3B?

Yes, Tesla P40 24GB can run Qwen 3.5 35B A3B with a A grade (Very compromised (needs ~1.7 GB host RAM)). Expected decode speed: 17.1 tok/s.

How much VRAM does Qwen 3.5 35B A3B need?

Qwen 3.5 35B A3B (35B parameters) requires approximately 26.1 GB of memory with Q4_K_M quantization.

What is the best quantization for Qwen 3.5 35B A3B?

The recommended quantization for Qwen 3.5 35B A3B is Q4_K_M, which balances quality and memory efficiency.

What speed will Qwen 3.5 35B A3B run at on Tesla P40 24GB?

On Tesla P40 24GB, Qwen 3.5 35B A3B achieves approximately 17.1 tokens per second decode speed with a time-to-first-token of 11325ms using Q4_K_M quantization.

Can Tesla P40 24GB run Qwen 3.5 35B A3B for coding?

For coding workloads, Qwen 3.5 35B A3B on Tesla P40 24GB receives a A grade with 17.1 tok/s and 4K context.

What context window can Qwen 3.5 35B A3B use on Tesla P40 24GB?

On Tesla P40 24GB, Qwen 3.5 35B A3B can safely use up to 4K tokens of context. The model's official context limit is 131K, but available memory constrains the safe maximum.

What should I upgrade first if Qwen 3.5 35B A3B feels slow on Tesla P40 24GB?

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.

See all results for Tesla P40 24GBSee all hardware for Qwen 3.5 35B A3B
Embed this result▼

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

<iframe src="https://willitrunai.com/embed/qwen-3.5-35b-a3b-on-tesla-p40-24gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>

Preview: