Powering the next generation of AI & high-performance computing.
Engineered for large-scale AI training, deep learning, and high-performance workloads, delivering unprecedented compute power and efficiency.
Why rent the H100 SXM instead of buying?
The highest-performance GPU for serious AI training
The H100 SXM's Hopper architecture delivers up to 4× faster LLM training than the A100, with Transformer Engine support, FP8 precision, and the highest NVLink bandwidth available at 900 GB/s. The SXM form factor's direct NVLink interconnect makes it the right choice for distributed training across multiple GPUs — near-linear scaling without PCIe bottlenecks.
Pay only for what you use
An H100 SXM costs $25,000–$40,000 per card on the open market. Runpod's on-demand pricing gives you access to the same hardware instantly, with no capital commitment, no depreciation, no idle hardware.
Deploy in seconds, scale without limits
Provision an H100 SXM instance in seconds. Scale to a multi-GPU cluster for a pre-training run, then scale back down when the job is done. Runpod handles the infrastructure so you can move fast without overprovisioning.
Use Cases
Popular use cases.
Designed for demanding workloads —learn if this GPU fits your needs.
Technical Specs
Ready for your most demanding workloads.
Essential technical specifications to help you choose the right GPU for your workload.
"The Runpod team has clearly prioritized the developer experience to create an elegant solution that enables individuals to rapidly develop custom AI apps or integrations while also paving the way for organizations to truly deliver on the promise of AI."
Amjad Masad
"Runpod is the only place I can deploy high-end GPU models instantly—no sales calls, no rate limits, no nonsense."
Daniel Chang
“The main value proposition for us was the flexibility Runpod offered. We were able to scale up effortlessly to meet the demand at launch.”
Josh Payne
“Runpod helped us scale the part of our platform that drives creation. That’s what fuels the rest—image generation, sharing, remixing. It starts with training.”
Matty Shimura
Comparison
Powerful GPUs. Globally available. Reliability you can trust.
30+ GPUs, 31 regions, instant scale. Fine-tune or go full Skynet—we’ve got you.
FAQs
Questions? Answers.
What are the current hourly rates for renting an H100 SXM on Runpod?
Pricing varies based on instance. For the most current pricing, see the Runpod pricing page.
How many H100 SXM GPUs do I need for my workload?
For large language model training (70B+ parameters), 8 or more GPUs are often recommended. For smaller models or fine-tuning, 1-4 GPUs may be sufficient. Single-GPU inference workloads can leverage the H100's MIG technology to serve multiple models concurrently. Benchmark your specific workload to determine the optimal configuration.
What multi-GPU configuration options are available?
Runpod supports configurations from single H100 SXM instances to multi-node clusters. The SXM form factor enables high-speed GPU-to-GPU communication via NVLink at 900 GB/s, which is essential for distributed training. See the Runpod pricing page for current multi-GPU options.
How do I choose between on-demand and reserved instances?
On-demand instances offer maximum flexibility for variable or short-term workloads. Reserved instances provide significant discounts for longer-term, predictable usage. See the Runpod pricing page to compare current rates.
Is the H100 SXM suitable for sensitive or regulated workloads?
Yes. Runpod Secure Cloud instances provide enterprise-grade isolation and reliability. The H100 supports hardware-level confidential computing features, and Runpod implements encryption in transit, strong authentication, and role-based access controls. For GDPR, HIPAA, or SOC 2 requirements, review Runpod's compliance documentation.
10,100,100,100
Requests since launch & 400k developers worldwide
