Rent A10G
Compare live on-demand and spot rental prices across 97+ cloud providers. Data center variant of RTX 3090 die. Very low TDP (150W). Ideal for inference at scale.
Choosing the right billing model for A10G
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
- Inference serving
- Graphics workloads
- Low-power deployments
A10G — Specs & Benchmarks
Performance bars, compute tiers (FP32/FP16/BF16/FP8/INT8), memory specs, LLM model size guidance, and related GPU comparisons.
A10G Rental Guide
The A10G is the most cost-effective data center GPU for inference serving of models up to 13B parameters on AWS. Available in g5 instances at $0.50–$1.50/hr, it delivers strong cost-per-token for 7B–13B models with ECC memory and data center reliability — advantages over the consumer RTX 4090 at similar price points. For Stable Diffusion and image generation, its 31.2 TFLOPS FP32 makes it one of the best options available.
Spot instances for A10G are available through AWS g5 spot instances at 40–70% below on-demand rates. For batch inference jobs and image generation pipelines, A10G spot instances often deliver the best cost-per-output of any ECC-grade GPU. On-demand is appropriate for production serving where uptime matters.
The main limitation to plan around is the 24GB GDDR6 ceiling. For models above 13B parameters at FP16, you will need INT8 quantization (fits up to 24B) or INT4 (fits up to 48B). Without NVLink, multi-card VRAM pooling is not supported. If your workload requires more than 24GB, consider the L40S (48GB) or A100 80GB instead. The A10G's data center reliability (ECC memory, validated for 24/7 operation) makes it preferable to the RTX 4090 for production deployments.
Frequently Asked Questions
How much does it cost to rent a A10G?
A10G 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 A10G?
The cheapest A10G 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 A10G?
With 24GB of GDDR6, the A10G can run LLM models up to approximately 12B parameters at FP16, 24B at INT8, or 48B at INT4/GGUF quantization. Common workloads include: Inference serving, Graphics workloads, Low-power deployments. Data center variant of RTX 3090 die. Very low TDP (150W). Ideal for inference at scale.
Should I use on-demand or spot pricing for A10G?
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 A10G 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 A10G's 125 TFLOPS FP16 and 600 GB/s bandwidth. The H100 is significantly more expensive — typically $2.50–$5.00/hr vs lower rates for the A10G. For workloads that fit within 24GB and don't require FP8 precision, the A10G often delivers better cost-per-token than the H100.
What is the memory bandwidth of the A10G and why does it matter?
The A10G has 600 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 A10G for Stable Diffusion or image generation?
Yes — the A10G is capable for Stable Diffusion and image generation workloads. Image generation is primarily FP32 and FP16 compute-bound, and the A10G's 31.2 TFLOPS FP32 throughput determines images-per-second. The 24GB 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.