Ampere Architecture
NVIDIA A40
NVIDIA A40 Datacenter GPU - Rent GPU servers or lease cloud GPUs for AI, machine learning, and high-performance computing.
NVIDIA A40
Ampere Architecture • Released 2020
Memory
48 GB
GDDR6
FP32 Performance
37.4 TFLOPS
Single Precision
Memory Bandwidth
696 GB/s
Data Transfer
Avg. Price
$0.75/hr
Pay as you go
Technical Specifications
| Manufacturer | NVIDIA |
| Architecture | Ampere |
| Release Year | 2020 |
| VRAM | 48 GB GDDR6 |
| Memory Bandwidth | 696 GB/s |
| FP32 Performance | 37.4 TFLOPS |
| FP16 Performance | 74.8 TFLOPS |
| INT8 Performance | 299 TOPS |
| Tensor Cores | 10752 |
| CUDA/Stream Processors | 10752 |
| TDP | 300W |
| Form Factor | PCIe |
| NVLink Support | No |
| Max GPUs per Node | 4 |
Performance Overview
Compute Performance
37.4TFLOPS
FP32 Single Precision
AI Training
74.8TFLOPS
FP16 Half Precision
AI Inference
299TOPS
INT8 Integer Ops
Ideal Use Cases
AI Inference
Graphics Rendering
Virtual Workstations
Data Science
Transparent Pricing
Hourly
$0.75
per hour
Most Popular
Daily
$18.00
per day
Monthly
$547.50
per month
Reserved Clusters
Enterprise-grade infrastructure
24h Quote
Need More? We've Got You Covered
Get a custom quote in 24 hours for reserved GPU clusters with high-speed interconnect, any region, any GPU model, and any number of GPUs you need.
Any GPU
Choose your hardware
Any Quantity
Scale as needed
Any Region
Global availability
Interconnect
High-speed networking
Why Rent NVIDIA A40 on Runcrate?
Deploy in 60 Seconds
Get instant access to NVIDIA A40. No setup required, start training immediately.
Pay Per Second
Only pay for what you use. Stop instances anytime. No hidden fees.
Enterprise Ready
99.9% uptime guarantee. 24/7 support. Production-grade infrastructure.