NVIDIA A40

48 GB VRAM150 FP16 TFLOPS300 W TDP5 providers10 offerings4 regions

The NVIDIA A40 is an enterprise-class GPU built on the Ampere architecture, featuring 48 GB of memory with 696 GB/s of memory bandwidth and a 300 W TDP. It delivers 37.4 TFLOPS of FP32 performance and 150 TFLOPS of FP16 tensor performance. Currently available from 5 providers starting at $0.44/GPU/hour, with a market median of $2.00/GPU/hour across 10 configurations.

Cheapest

$0.440/hr

Runpod

Median

$2.003/hr

across 10 offerings

Most expensive

$2.333/hr

Vultr

90-day trend

+36.0%

median rate

Hardware specifications

Same across all providers

Memory

48 GB

What LLMs fit? ↓

FP16 performance

150 TFLOPS

peak

TDP

300 W

Memory bandwidth

696 GB/s

Architecture

Ampere

FP32 performance

37.4 TFLOPS

Brand

NVIDIA

Series

A40

Market price history

Median price across all providers

Prices updated 7 hours ago

5 offerings from 5 providers

Sorted by price ascending. Compare configurations side-by-side.

Provider	Count	vCPU	RAM	Region	Per GPU hour	Total/hr	Action
RunpodReferral linkTrendingCheapest Up to$500credit	×1	9	50 GB		$0.440	$0.44	Launch
Vultr	×1/20	1	5 GB		From$1.528	From$0.08	Launch
Omega Gradient	×1	24	120 GB	--	$1.960 0.0% 30d	$1.96	Launch
Sesterce CloudReferral link	×1	24	120 GB		From$2.046 0.0% 30d	From$2.05	Launch
Runcrate	×1	24	120 GB		$2.050 0.0% 30d	$2.05	Launch

LLMs that fit on the NVIDIA A40

Estimated VRAM for popular open-weight LLMs at 16-bit (FP16/BF16), 8-bit (FP8/INT8) and 4-bit (INT4/MXFP4) precision, against 48 GB per GPU. Showing the 12 largest that fit.

Model	Parameters	16-bit	8-bit	4-bit
Qwen3.6 35B-A3B Alibaba	35B 3B active	86 GB Too large	47 GB Barely fits on 1×	28 GB 1× · 20 GB free
Qwen3 32B Alibaba	32.8B	79 GB Too large	43 GB 1× · 4.8 GB free	25 GB 1× · 23 GB free
Gemma 4 31B Google	31.3B	75 GB Too large	41 GB 1× · 6.7 GB free	24 GB 1× · 24 GB free
GLM-4.7 Flash Z.ai	31.2B	75 GB Too large	41 GB 1× · 6.8 GB free	24 GB 1× · 24 GB free
Qwen3.6 27B Alibaba	27.8B	67 GB Too large	37 GB 1× · 11 GB free	21 GB 1× · 27 GB free
Mistral Small 3.2 24B Mistral AI	24B	58 GB Too large	32 GB 1× · 16 GB free	18 GB 1× · 30 GB free
gpt-oss-20b OpenAI	21B 3.6B active	— Not released	— Not released	17 GB 1× · 31 GB free
Gemma 3 12B Google	12.2B	29 GB 1× · 19 GB free	16 GB 1× · 32 GB free	9.4 GB 1× · 39 GB free
Granite 4.1 8B IBM	8.8B	21 GB 1× · 27 GB free	12 GB 1× · 36 GB free	6.8 GB 1× · 41 GB free
Qwen3 8B Alibaba	8.2B	20 GB 1× · 28 GB free	11 GB 1× · 37 GB free	6.3 GB 1× · 42 GB free
Llama 3.1 8B Instruct Meta	8B	19 GB 1× · 29 GB free	11 GB 1× · 37 GB free	6.2 GB 1× · 42 GB free
Mistral 7B Instruct v0.3 Mistral AI	7.2B	17 GB 1× · 31 GB free	9.6 GB 1× · 38 GB free	5.6 GB 1× · 42 GB free

Estimates: model weights plus 20% for KV cache and runtime overhead, at short context lengths. Long contexts and large batches need more. Rows marked “tight” hold the weights but not that full margin. A ×N figure is the total VRAM across N of these GPUs and makes no claim about interconnect throughput. How we estimate this · Start from a model instead

See more GPUs like this in

Best bang for buck

Want more from this page?

See incorrect data?

Frequently asked questions

How much does the A40 cost per hour?

The A40 starts at $0.44 per GPU-hour, with a median of $2.00 across 10 available configurations. Actual pricing depends on the provider, region, and whether you rent on-demand, spot or reserved.

Which cloud providers offer the A40?

5 providers currently list the A40: Sesterce Cloud, Runcrate, Omega Gradient, Runpod, Vultr. Compare their hourly pricing, regions and machine specs in the table above.

How much VRAM does the A40 have?

The A40 comes with 48 GB of VRAM, which determines the largest models and batch sizes it can hold in memory for training and inference.

Which LLMs can I run on the A40?

With 48 GB of VRAM, a single A40 can serve models such as Qwen3.6 35B-A3B, Qwen3 32B, Gemma 4 31B, GLM-4.7 Flash. The table above lists the estimated VRAM for each model at 16-bit, 8-bit and 4-bit precision, including the multi-GPU configurations needed for larger models.

Where is the A40 cheapest?

The lowest on-demand price we currently track for the A40 is $0.44 per GPU-hour at Runpod. Because providers update pricing and availability frequently, check the live table above before you launch.