Rent A100 40GB
Compare live on-demand and spot rental prices across 97+ cloud providers. Same compute as 80GB variant but half the VRAM. Lower cost, suitable for models up to ~30B params.
Choosing the right billing model for A100 40GB
Provision and terminate at any time. Ideal for development, short experiments, and workloads with unpredictable duration.
Instances can be reclaimed when demand spikes. Best for fault-tolerant batch jobs, training with checkpointing, and preprocessing.
Lock in a rate for 1–3 months. Right for sustained production inference or long training runs where cost predictability matters.
Best use cases
- ML training
- Mid-size model inference
- Research
A100 40GB — Specs & Benchmarks
Performance bars, compute tiers (FP32/FP16/BF16/FP8/INT8), memory specs, LLM model size guidance, and related GPU comparisons.
A100 40GB Rental Guide
Choose the A100 40GB when you need dependable data-center hardware and HBM bandwidth but do not need the 80GB model’s capacity. It is generally cheaper than the A100 80GB and can be an efficient choice for 7B–20B-scale training and inference with mature CUDA tooling.
Spot A100 capacity is broadly available and works especially well for checkpointed training, offline inference, and research queues. On-demand instances are preferable for always-on endpoints or when a fixed schedule matters more than the savings from interruption-tolerant capacity.
Check whether the provider offers SXM or PCIe hardware and whether NVLink is actually exposed for multi-GPU work. Compare with H100 40GB if FP8 throughput is useful, and step up to A100 80GB or H200 when memory capacity—not compute—is the limiting resource.
Frequently Asked Questions
How much does it cost to rent a A100 40GB?
A100 40GB on-demand rental prices vary by provider and region. On-demand rates typically range based on availability and provider margins — use the comparison table above to see current live rates across all providers. Spot instances are generally 40–70% cheaper than on-demand but can be interrupted. Monthly cost estimates (hourly rate × 730 hours) are shown in the table for sustained workloads.
Which cloud provider has the cheapest A100 40GB?
The cheapest A100 40GB provider changes as providers update their pricing. The comparison table above shows live rates sorted by price, so the cheapest option is always at the top. Factors beyond headline price include region (latency to your users), availability (high/medium/low), and billing granularity (per-second vs per-hour minimums).
What can I run on a A100 40GB?
With 40GB of HBM2e, the A100 40GB can run LLM models up to approximately 20B parameters at FP16, 40B at INT8, or 80B at INT4/GGUF quantization. Common workloads include: ML training, Mid-size model inference, Research. Same compute as 80GB variant but half the VRAM. Lower cost, suitable for models up to ~30B params.
Should I use on-demand or spot pricing for A100 40GB?
Spot instances save 40–70% vs on-demand but can be interrupted when the provider needs capacity back. Use spot for: batch inference jobs, training runs with checkpointing, preprocessing pipelines, and any workload that can tolerate interruption and restart. Use on-demand for: production inference serving, interactive workloads, and jobs that cannot be interrupted. Most providers bill per second, so short on-demand jobs are not penalized by hourly minimums.
How does the A100 40GB compare to the H100 for cloud rental?
The H100 80GB delivers 1,979 TFLOPS FP16 with 3,350 GB/s HBM3 bandwidth, compared to the A100 40GB's 312 TFLOPS FP16 and 1555 GB/s bandwidth. The H100 is significantly more expensive — typically $2.50–$5.00/hr vs lower rates for the A100 40GB. For workloads that fit within 40GB and don't require FP8 precision, the A100 40GB often delivers better cost-per-token than the H100.
What is the memory bandwidth of the A100 40GB and why does it matter?
The A100 40GB has 1555 GB/s of memory bandwidth. For LLM inference, memory bandwidth is often more important than raw TFLOPS — each autoregressive token generation reads the full model weight matrix from VRAM, so bandwidth directly determines tokens-per-second throughput. Higher bandwidth means faster inference for the same model at the same batch size. For batch inference (processing many requests simultaneously), compute throughput becomes more important.
Can I use the A100 40GB for Stable Diffusion or image generation?
Yes — the A100 40GB is capable for Stable Diffusion and image generation workloads. Image generation is primarily FP32 and FP16 compute-bound, and the A100 40GB's 19.5 TFLOPS FP32 throughput determines images-per-second. The 40GB VRAM fits SDXL (requires ~6GB) and most ControlNet pipelines. For high-throughput image generation at scale, compare cost-per-image across providers using the GPU cost calculator.