NVIDIA

NVIDIA · Ampere · 2020

NVIDIA A40.

Rent the NVIDIA A40 for AI training and inference. 48GB GDDR6, 74.8 TFLOPS FP16, deploy in 60 seconds.

48 GB
GDDR6
696 GB/s
Bandwidth
74.8
TFLOPS FP16
4× GPU
PCIe cluster
Available now · 4 regions

Cloud rental

$0.75/hr

On-demand · per-second billing

$0.65$0.85/hr range

Or $0.52/hr reserved (30% off)

Deploy

One command. NVIDIA A40 in 60 seconds.

Skip the dashboard if you don't need it. SDK, Python, or cURL — copy the snippet, paste your API key, ship.

import Runcrate from "@runcrate/sdk";

const rc = new Runcrate({ apiKey: "rc_live_••••••••••••••••" });

const instance = await rc.instances.create({
  gpu: "a40",
  region: "auto",
  image: "runcrate/vllm:latest",
});

console.log(`SSH: ssh root@${instance.host}`);

Benchmarks

How NVIDIA A40 stacks up.

Side-by-side specs vs the closest alternatives. Bars are normalized to the highest value in each metric.

FP16

NVIDIA A4074.8 TFLOPS
NVIDIA RTX A600077.4 TFLOPS
AMD MI21047.9 TFLOPS
NVIDIA L40S183 TFLOPS

VRAM

NVIDIA A4048 GB
NVIDIA RTX A600048 GB
AMD MI21064 GB
NVIDIA L40S48 GB

Bandwidth

NVIDIA A40696 GB/s
NVIDIA RTX A6000768 GB/s
AMD MI2101.6 TB/s
NVIDIA L40S864 GB/s

Full specs

NVIDIA A40 specifications.

Memory

VRAM48 GB
Memory typeGDDR6
Bandwidth696 GB/s

Compute

FP3237.4 TFLOPS
FP1674.8 TFLOPS
INT8299 TOPS
Tensor cores10,752
CUDA cores10,752

Power & form factor

TDP300 W
Form factorPCIe
ArchitectureAmpere
Released2020

Cluster

Max GPUs/node4
InterconnectPCIe
ManufacturerNVIDIA
Available regions4

FAQ

Frequently asked.

Ready when you are

Deploy NVIDIA A40
in 60 seconds.

Available across 4 regions · per-second billing · no commitments. Ship today.