Rent NVIDIA A40 in the Cloud

When pricing a NVIDIA A40 cloud VM, start with the on-demand floor of $0.30 per hour (offered by Vast.ai) and layer discounts on top. If the job can checkpoint, spot drops the rate to $0.35 per hour on RunPod — roughly up to 60% off on-demand.

Long-lived services migrate to reserved tiers for an additional ~up to 40% saving. The 5-provider market ensures competitive pricing, but regional availability and GPU count per node affect real-world cost more than headline rates.

Launch a NVIDIA A40 instance on Vast.ai from $0.30/hr — or check RunPod and Massed Compute for alternative regions and availability.

NVIDIA A40 is a Ampere card with 48 GB of GDDR6 memory, 696 GB/s of bandwidth, and 150 TFLOPS of FP16 compute. It launched in 2020 and uses up to 300W of power.

On real workloads the memory subsystem is typically the most important factor: 696 GB/s of bandwidth is enough to sustain most transformer decoding tasks without starving the compute units, while 48 GB of VRAM supports a wide range of model sizes, especially once quantisation is applied.

Rent NVIDIA A40 today from Vast.ai, RunPod, or Massed Compute — compare live pricing and spin up in minutes.

NVIDIA A40 pushes 150 TFLOPS of FP16, 37.4 TFLOPS of FP32, and feeds them from 48 GB of VRAM at 696 GB/s.

Benchmarks: LLM training with mixed precision sees near-peak FLOPS utilisation at batch sizes that fit in VRAM; LLM inference is typically within 5-15% of the theoretical bandwidth-bound ceiling on autoregressive decoding; diffusion models show the biggest jump over older accelerators, where faster attention kernels stack with the raw compute gains.

Rent NVIDIA A40 today from Vast.ai, RunPod, or Massed Compute — compare live pricing and spin up in minutes.

The 5 providers offering NVIDIA A40 today: Cherry Servers, Massed Compute, RunPod, Vast.ai, Vultr. The headline-cheapest on-demand source is Vast.ai at $0.30 per hour; spot pricing bottoms out at $0.35 per hour on RunPod.

Real-world selection factors include: interconnect technology (NVLink vs. PCIe vs. InfiniBand), multi-node availability, per-second vs. per-hour billing, egress fees, and regional coverage. A cheap provider with high egress fees can cost more than a pricier provider with free egress once you factor in data movement.

Top cloud providers offering NVIDIA A40: Vast.ai (from $0.30/hr), RunPod, and Massed Compute. Start comparing live rates.

NVIDIA A40 was designed for Inference, rendering, virtual workstations. In cloud deployments that translates to: AI model training at research-to-production scale, fine-tuning transformers, running batch serving pipelines, and accelerating diffusion/vision workloads.

It sits in the middle of the AI accelerator market — more capable than consumer cards, more affordable than frontier accelerators. For most teams, that's exactly the right tier: enough performance for real work, enough savings to run multiple experiments in parallel. Cloud access from $0.30 per hour via Vast.ai.

Top cloud providers offering NVIDIA A40: Vast.ai (from $0.30/hr), RunPod, and Massed Compute. Start comparing live rates.