NVIDIA A40

Consumer GPU · Ampere

Summary

The NVIDIA A40 is an Ampere-generation workstation GPU with 48GB of GDDR6. It shares its GPU with the RTX A6000 but has lower memory bandwidth and no NVLink. For cloud inference and rendering it has largely been superseded by the Ada-generation NVIDIA L40 and L40S.

Launched
2020
VRAM
48 GB
Mem. bandwidth
696 GB/s
On-demand from
$0.59 /GPU/h
Compare prices from 4 providers →

Tech specs

VRAM 48 GB GDDR6
Memory bandwidth 696 GB/s
Interface PCIe
CUDA cores 10,752
Tensor cores 336 (Gen 3)
TDP 300 W
Supported data types
FP64FP32TF32FP16BF16INT8INT4

Cloud rental prices

4 providers · available in 8 countries · lowest price per GPU, per hour

Prices updated

On-demand
$0.59 – $2.05 /GPU/h
Provider On-demand Reserved Spot
Denvr ? $0.65 — —
Runpod ✓ $0.59 — —
Sesterce ✓ $2.05 — —
Vultr ? $1.71 — —
Compare all NVIDIA A40 cloud providers & configurations →

Models that fit in VRAM

Total VRAM From 16-bit inference 8-bit inference 4-bit inference
1× A40 48 GB $0.59/h gpt-oss-20bQwen3-4B-Instruct-2507Rio-3.0-Open-Mini GLM-4.7-FlashQwen3-Coder-30B-A3B-InstructNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 Qwen3-Coder-NextGLM-4.7-FlashNVIDIA-Nemotron-3-Nano-30B-A3B-BF16
2× A40 96 GB $1.18/h GLM-4.7-FlashQwen3-Coder-30B-A3B-InstructNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 Qwen3-Coder-NextGLM-4.7-FlashNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16
3× A40 144 GB $1.77/h GLM-4.7-FlashQwen3-Coder-30B-A3B-InstructNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 gpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16GLM-4.5-Air MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b
4× A40 192 GB $2.36/h Qwen3-Coder-NextGLM-4.7-FlashNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16 MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b
5× A40 240 GB $2.95/h GLM-4.5-AirQwen3-Coder-NextNVIDIA-Nemotron-3-Nano-30B-A3B-BF16 DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16 MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b
6× A40 288 GB $3.54/h gpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16GLM-4.5-Air MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b
7× A40 336 GB $4.13/h gpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16GLM-4.5-Air MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b
8× A40 384 GB $4.72/h DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16 MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b
9× A40 432 GB $5.31/h DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16 MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b DeepSeek-V3.2MiniMax-M2.7gpt-oss-120b
10× A40 480 GB $5.90/h DeepSeek-V4-Flashgpt-oss-120bNVIDIA-Nemotron-3-Super-120B-A12B-BF16 MiniMax-M2.7DeepSeek-V4-Flashgpt-oss-120b DeepSeek-V4-ProMiniMax-M2.7gpt-oss-120b

Open-weights models that fit in VRAM. Estimated using 🤗 accelerate, plus approximation for up to 8K context.

Resources

Detailed documentation from the manufacturer.

Datasheet ↗ Official spec sheet Product page ↗ Manufacturer overview