A100 40GB PCIe
The A100 40GB PCIe is a cloud GPU with 40 GB of memory and 250 W TDP. It is currently available from 3 providers starting at $0.41/hr, with a market median of $1.72/hr across 13 offerings.
Best deal
Hardware specifications
- Architecture
- Ampere
- VRAM
- 40 GB
- Memory type
- HBM2e
- Memory bandwidth
- 1,555 GB/s
- FP16
- 312 TFLOPS
- FP32
- 20 TFLOPS
- TDP
- 250 W
Same across all providers.
Market comparison
All-time price history
Offerings
| Provider | Cloud | Count | vCPU | RAM | Region | Per GPU hr | SPOT | Total/hr | |
|---|---|---|---|---|---|---|---|---|---|
AI Galaxy 0.0% 30dReferral link | — | ×1 | — | — | CN | $0.41/hr | — | $0.41/hr | Launch ↗ |
HardwareHQ 0.0% 30dReferral link | — | ×1 | — | — | — | $0.89/hr | — | $0.89/hr | Launch ↗ |
HardwareHQ 0.0% 30dReferral link | — | ×1 | — | — | — | $1.00/hr | — | $1.00/hr | Launch ↗ |
RunPodPromoted 0.0% 30dReferral link | Community | ×1 | — | — | — | $1.00/hr | — | $1.00/hr | Launch ↗ |
HardwareHQ 0.0% 30dReferral link | — | ×1 | — | — | — | $1.10/hr | — | $1.10/hr | Launch ↗ |
HardwareHQ 0.0% 30dReferral link | — | ×1 | — | — | — | $1.10/hr | — | $1.10/hr | Launch ↗ |
HardwareHQ 0.0% 30dReferral link | — | ×1 | — | — | — | $1.89/hr | — | $1.89/hr | Launch ↗ |
HardwareHQ 0.0% 30dReferral link | — | ×1 | — | — | — | $3.60/hr | — | $3.60/hr | Launch ↗ |
HardwareHQ 0.0% 30dReferral link | — | ×1 | — | — | — | $3.67/hr | — | $3.67/hr | Launch ↗ |
LLMs that fit
| Model | Params | 16-bit | 8-bit | 4-bit |
|---|---|---|---|---|
DeepSeek V4 Flash DeepSeek | 284B | Not released | Not released | Barely fits on 6× |
Solar Open2 250B Upstage | 250B 15B active | Needs >8 GPUs | Needs >8 GPUs | Barely fits on 6× |
Qwen3 235B-A22B Alibaba | 235B 22B active | Needs >8 GPUs | Needs >8 GPUs | Barely fits on 6× |
Laguna-S 2.1 Poolside | 118B 8B active | Needs >8 GPUs | Barely fits on 5× | Barely fits on 3× |
gpt-oss-120b OpenAI | 117B 5.1B active | Not released | Not released | 3× · 26.4 GB free |
Llama 3.1 70B Meta | 70B | 5× · 32 GB free | 3× · 28.2 GB free | 2× · 26 GB free |
Qwen3.6 35B-A3B Alibaba | 35B 3B active | 3× · 14% spare | 2× · 23.6 GB free | 1× · 6.4 GB free |
Qwen3 32B Alibaba | 32.8B | 3× · 25.2 GB free | 2× · 28.4 GB free | 1× · 10 GB free |
Qwen2.5 14B Alibaba | 14B | Barely fits on 1× | 1× · 20.7 GB free | 1× · 28.6 GB free |
Llama 3.1 8B Meta | 8B | 1× · 19.1 GB free | 1× · 28.6 GB free | 1× · 33.3 GB free |
Mistral 7B Mistral AI | 7B | 1× · 21.8 GB free | 1× · 30 GB free | 1× · 34.1 GB free |
Qwen3 4B Alibaba | 4B | 1× · 28.4 GB free | 1× · 33.6 GB free | 1× · 36.3 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.
Similar GPUs
See more GPUs like this in
Frequently asked questions
How much does the A100 40GB PCIe cost per hour?
The cheapest A100 40GB PCIe offering starts at $0.41/hr, and the market median is $1.72/hr.
Which cloud providers offer the A100 40GB PCIe?
We track A100 40GB PCIe offerings from AI Galaxy, HardwareHQ, RunPod.
How much VRAM does the A100 40GB PCIe have?
The A100 40GB PCIe has 40 GB of VRAM.
What LLMs can run on the A100 40GB PCIe?
With 40 GB of VRAM, the A100 40GB PCIe can run 12 of our tracked models in 4-bit quantization, including DeepSeek V4 Flash, Solar Open2 250B, Qwen3 235B-A22B. 7 models fit in 16-bit precision.
What's the biggest LLM I can run on multiple A100 40GB PCIes?
With 8× A100 40GB PCIe (320 GB total), the largest tracked model that fits in 16-bit is Llama 3.1 70B (70B params), which needs 5× GPUs.
Where is the A100 40GB PCIe cheapest?
The cheapest A100 40GB PCIe offering in our catalog is AI Galaxy at $0.41/hr per GPU-hour (×1).