We need an architecture that securely delivers high performance with low latency, and integrates a full stack of capabilities from networking to compute at data center scale, the new unit of computing. The H100 is that new architecture.

NVIDIA DGX™ H100 is an AI powerhouse that lets enterprises expand the frontiers of business innovation and optimization.

Logo ©copyright 2015. All rights reserved.

NVIDIA H100 Tensor Core GPU

The New Architecture For AI, Modeling and HPC

Click on the icon to learn more about each product

Padova can meet all your H100 needs. We will provide any type of support you need to get going on your Processing/AI/Modeling journey.

The higher processing output of each of the H100s creates an environment conducive to using less equipment to do the same amount of work, from AI to Modeling, from the Data Center to the Edge. Less equipment means lower costs for the end user.

Much like an increase in the power draw the H100 has a larger footprint. Padova understands the need for greater physical space on the server and will configure an environment tailored to fit the customers needs and desires.

With each new generation of GPUs, the power demands increase. Like its predecessors, the H100 is no different. The wattage has increased from the 250 watt A100 to 350 watts. Along with this wattage increase comes more robust usage. The A100 had 6912 shading units while the H100 has a boosted 7296 shading units allowing the H100 to speed up the process for the workload. Padova understands the greater power draw and builds a whole system to work with the greater power needs.

The H100 boasts an unprecedented 32 petaFLOPS with 640gb total GPU memory. Built with 80 billion transistors using a cutting edge TSMC 4N process custom tailored for NVIDIA’s accelerated compute needs, H100 is the world’s most advanced chip ever built. It features major advances to accelerate AI, HPC, memory bandwidth, interconnect and communication at data center scale. The H100 can crunch through large data sets faster than ever.

The World’s First Confidential Computing Chip

Today’s confidential computing solutions are CPU-based, which is too limited for compute-intensive workloads like AI and HPC. NVIDIA Confidential Computing is a built-in security feature of the NVIDIA Hopper™ architecture that makes NVIDIA H100 the world’s first accelerator with confidential computing capabilities. Users can protect the confidentiality and integrity of their data and applications in use while accessing the unsurpassed acceleration of H100 GPUs. It creates a hardware-based trusted execution environment (TEE) that secures and isolates the entire workload running on a single H100 GPU, multiple H100 GPUs within a node, or individual MIG instances. GPU-accelerated applications can run unchanged within the TEE and don’t have to be partitioned. Users can combine the power of NVIDIA software for AI and HPC with the security of a hardware root of trust offered by NVIDIA Confidential Computing.