NVIDIA / Ampere
A40
| Provider / plan | Price / GPU-hour | Rental type | Terms & configuration | Last checked | Provider pricing link |
|---|---|---|---|---|---|
| RunpodSecure Cloud Pods | $0.49Lowest flexible rate | On-demandNo reservation | 1 GPU minimumSee providerView detailsSecure Cloud Pods published GPU-hour rate. CPU/RAM allocation varies; storage/networking extra. Not live capacity.
|
2026-10-03Auto-collected | View pricing ↗ |
| Vast.aiOn-demand offer 52628046 | $0.6022 | On-demandNo reservation | 1 GPU minimumUnited Kingdom, GBView detailsObserved marketplace offer, not a provider-wide list rate. Total hourly quote for one GPU with the API default disk allocation; bandwidth and additional storage can cost extra. Up to 5 cheapest matching offers per catalog GPU; availability can change.
|
2026-10-03Auto-collected | View pricing ↗ |
| Vast.aiOn-demand offer 51337965 | $0.7889 | On-demandNo reservation | 1 GPU minimumBelgium, BEView detailsObserved marketplace offer, not a provider-wide list rate. Total hourly quote for one GPU with the API default disk allocation; bandwidth and additional storage can cost extra. Up to 5 cheapest matching offers per catalog GPU; availability can change.
|
2026-10-03Auto-collected | View pricing ↗ |
| Vast.aiOn-demand offer 51338030 | $1.0556 | On-demandNo reservation | 1 GPU minimumBelgium, BEView detailsObserved marketplace offer, not a provider-wide list rate. Total hourly quote for one GPU with the API default disk allocation; bandwidth and additional storage can cost extra. Up to 5 cheapest matching offers per catalog GPU; availability can change.
|
2026-10-03Auto-collected | View pricing ↗ |
Prices are dated listings, not live availability. Compare minimum GPU counts and commitments. Storage, networking and taxes may add charges; serverless rates cover GPU time only.
A40 is a 48 GB Ampere server GPU for visualization, rendering and AI. Its passive design targets data centers, while the memory capacity supports workloads that exceed 24 GB cards. Optional NVLink can connect two supported cards.
CHOOSING THIS GPU
Is it right for your workload?
A40 is a graphics-capable Ampere GPU, not an A100. Its GDDR6 memory and compute design differ from A100 HBM accelerators.
Best suited to
- Remote visualization and GPU rendering
- Inference needing more than 24 GB of memory
- Adapter fine-tuning with a carefully sized memory budget
Strengths
- 48 GB ECC memory
- Ray tracing hardware
- Optional two-card NVLink connectivity
Things to consider
- No MIG
- No native FP8 Tensor acceleration
- Passive cooling and bridge availability depend on the server
HARDWARE DETAILS
Inside the GPU
- Manufacturer
- NVIDIA
- Architecture
- Ampere
- VRAM (GB)
- 48
- Memory type
- GDDR6
- Form factor
- PCIe
- Memory bandwidth (GB/s)
- 696
- Maximum board power (W)
- 300
- CUDA cores
- 10752
- Tensor cores
- 336 / 3rd generation
- Ray tracing cores
- 84 / 2nd generation
- CUDA capability
- 8.6
- Memory ECC
- Supported
- Host interface
- PCIe Gen4
- GPU interconnect
- Two GPUs; 112.5 GB/s aggregate bidirectional
- Hardware partitioning (MIG)
- Not supported
- Cooling
- Passive; server airflow required
Hardware capabilities are not a guarantee of access in a cloud instance. Check the exact edition, GPU allocation and server topology in the rental offer.
MODEL MEMORY
Plan your model size
Compare 8B/14B FP16 and larger quantized models using the weight-only estimates below. A 70B 4-bit model has about 35 GB of raw weights before quantization metadata and runtime memory; nominal capacity alone does not establish a fit.
| Example model class | Weight format | Raw weights ≈ | Minimum GPUs for weights only |
|---|---|---|---|
| Llama 3.1 8B ↗ | 4-bit | 4 GB | 1Runtime needs more memory |
| Llama 3.1 8B ↗ | FP16 | 16 GB | 1Runtime needs more memory |
| Qwen2.5 14B ↗ | 4-bit | 7 GB | 1Runtime needs more memory |
| Qwen2.5 32B ↗ | 4-bit | 16 GB | 1Runtime needs more memory |
| Llama 3.1 70B ↗ | 4-bit | 35 GB | 1Runtime needs more memory |
| Llama 3.1 70B ↗ | 8-bit | 70 GB | 2Runtime needs more memory |
| Llama 3.1 70B ↗ | FP16 | 140 GB | 3Runtime needs more memory |
Arithmetic lower bound: rounded parameter count × bits per weight ÷ 8, in decimal GB. Excludes quantization metadata, KV cache, activations, CUDA workspaces and training states. Actual model sizes differ from their rounded names. No context length, batch size or concurrency is guaranteed. GPU counts assume supported model sharding and can be higher in practice. Quantized checkpoints and compatible kernels are required for 4-bit/8-bit execution.
SOFTWARE & PERFORMANCE
Before you deploy
Software compatibility
Use an NVIDIA CUDA-enabled framework/container compatible with this GPU and the host driver. Check PyTorch build and inference-engine kernel requirements before deployment. vLLM documents NVIDIA compute capability 7.5+ as a baseline; support for each model and quantization method still needs checking.
Measured benchmarks
A reproducible, comparable application benchmark has not yet been recorded for this edition. Peak TFLOPS above are not measured tokens per second or image-generation speed.
Compare tests using the same model, precision, GPU count, engine version, input/output length and batch or concurrency. A provider’s server configuration can change the result.
COMPARE YOUR OPTIONS
Alternatives to consider
BUDGET YOUR RUN
GPU cost estimator
GPU compute only: hourly rate × billable hours × GPU count. CPU, RAM, storage, egress, taxes, billing increments and minimum terms may add charges. Serverless hours mean active GPU task time; spot capacity can be interrupted. Reserved commitments and quote-only offers are excluded from this simple hourly estimator.