Nvidia A100 H100 cloud servers

Get access to unmatched performance and accelerated AI capabilities with bare metal cloud servers powered by Nvidia A100 / H100 GPUs.

Get started

Nvidia A100 H100 configurations, perfect for AI, deep learning, visualization and high-performance computing.

check-mark

NVMe disks

check-mark

24/7 support

check-mark

Unbeatable prices

check-mark

5 minutes deployment

NVIDIA A100 GPU

A100 GPU
NVIDIA A100 specs:
Video memory capacity 40GB / 80GB HBM2
CUDA cores 6912 pcs.
Max Bandwidth 1.6 TB/s
The Nvidia A100 GPUs offers the performance, scalability, and efficiency necessary for AI and deep learning applications, making it an excellent option for businesses and researchers looking for cutting-edge computational power.
Ampere infrastructure

Ampere infrastructure

With 54 billion transistors, the NVIDIA Ampere architecture is one of largest 7 nanometer chips ever built.

High-bandwidth memory

High-bandwidth memory (HBM2)

HBM2 is designed to offer quick and effective data access, with memory bandwidth of up to 1.6 TB/s.

AI and Deep Learning

AI and Deep Learning

The Nvidia A100 is specifically designed for artificial intelligence and deep learning applications, delivering up to 20 times the performance of previous-generation GPUs.

Ray tracing

Ray tracing

Experience real-time ray tracing, ideal for demanding visualization tasks. The A100 GPU delivers the rendering power needed for realistic and immersive visuals.

NVLink support

NVLink support

Utilize the capabilities of 3rd-generation NVLink for lightning-fast data transfers up to 10x to 20x faster than PCI-Express.

Multi-interface GPU

Multi-interface GPU

Empower each A100 GPU to simultaneously run seven distinct and isolated applications or user sessions.

See configurations

NVIDIA H100 GPU

NVIDIA H100 specs:
Video memory capacity 80GB HBM3
CUDA cores 8448 pcs.
Max Bandwidth 3 TB/s
The latest released NVIDIA H100 GPU offers unprecedented performance, scalability, and security for various workloads. It is at least two times faster than its predecessor, the A100 GPU.
H100 gpu
Hopper infrastructure

Hopper infrastructure

H100 GPUs deliver outstanding performance in HPC applications, 7x higher performance compared to A100 GPU.

Connectivity

Connectivity

Up to 256 H100 GPUs can be connected using the NVIDIA NVLink Switch System, allowing exascale tasks to be accelerated.

AI training

AI training

The H100 GPU stands as the fourth generation of the most advanced AI system, delivering exceptional performance capabilities.

Multi-interface GPU

Multi-interface GPU

The H100 features second-generation MIG technology, which allows each GPU to be securely partitioned into up to seven separate instances.

Instant Nvidia A100/H100 configurations

Loading...

Custom Nvidia A100/H100 configurations

Loading...

Why Primcast?

Fast delivery

Fast delivery

Instant A100 configurations are delivered within 5 minutes, with your verified payment.

SLA

99.9% sla

Your GPU dedicated server is backed by an industry-leading, with 99.9% uptime SLA.

Support

24/7 support

Live GPU cloud experts are available via live chat, phone and email.

Data centers

Data centers

Deploy your NVIDIA A100 dedicated server from New York, Miami, San Francisco, Amsterdam or Bucharest.

Deploy your A100 H100 bare-metal cloud server in minutes!

Get started