Question 1

What AI models can run on A100 40GB?

Accepted Answer

The A100 40GB (40 GB VRAM) can run models including Llama-2-7b-chat, Llama-3.2-1B-Instruct-QLORA_INT4_EO8, MediPhi-Instruct, Llama-3.2-3B-Instruct-SpinQuant_INT4_EO8, Llama-3.2-1B-Instruct-SpinQuant_INT4_EO8 and more — 10 compatible models indexed so far. Compatible models are determined by whether the model's memory footprint fits within the GPU's 40 GB VRAM, potentially using quantization (Q8, Q6, or Q4) to reduce memory usage.

Question 2

What is the cheapest cloud option for A100 40GB?

Accepted Answer

The cheapest on-demand option for the A100 40GB is Vast.ai at $0.93/hr (1x-A100_SXM4-32-65536). Spot/preemptible pricing is available at $0.00/hr. We track pricing across 46 providers to help you find the best deal.

Question 3

Is A100 40GB good for fine-tuning?

Accepted Answer

Yes, the A100 40GB is well-suited for fine-tuning. With 40 GB of VRAM, it can handle LoRA/QLoRA fine-tuning of models up to ~30B parameters and full fine-tuning of smaller models (up to ~7B parameters). For larger models, consider multi-GPU setups.

Question 4

How much VRAM does A100 40GB have?

Accepted Answer

The A100 40GB has 40 GB of VRAM (HBM2). It also offers 1555 GB/s memory bandwidth, 312 TFLOPS of FP16 compute.

Question 5

What quantization formats work on A100 40GB?

Accepted Answer

The A100 40GB supports the following quantization formats: FP16 (full precision), Q8 (8-bit), Q6 (6-bit), Q4 (4-bit). FP16 works for models up to ~7B parameters. Q4 quantization lets models up to ~136B parameters (by VRAM estimate) fit on the A100 40GB.

Provider	On-demand	Spot	Actions
Vast.ai	$0.930/hr	$0.000/hr	Rent
Google Cloud	$1.614/hr	$1.228/hr	—
Lambda	$1.990/hr	—	—

A100 40GB

Specifications

Cloud pricing

Compatible models

Compare this GPU

Frequently asked questions

VRAM	40 GB
Memory type	HBM2
Bandwidth	1555 GB/s
FP16 TFLOPS	312
FP8 support	No