Coming soon
NVIDIA Blackwell Cloud GPUs
A New Class of AI Chip...
- Train up to 10 Trillion parameters
- Up to 800Gb/s Networking Speed
- 1.8Tb/s throughput per GPU
Solutions
Accelerated AI
Whether you need lightning fast inference for Generative AI, or you are model training neural networks for AI, our VMs outperform the competition. Our GPU cloud is scalable for Machine Learning and enables more advanced execution.
Rendering
We have a range of SKUs purpose built for render-based workloads using NVIDIA's RTX family. Ideal for 3D visuals, game development, photorealism, and film production. Real-time rendering is now a reality with Hyperstack.
Machine Learning
Train your models on lightning-fast GPUs, scale seamlessly without resource constraints, and experience the power of AI. Hyperstack will handle the infrastructure, you focus on breakthroughs.
Data Analytics
Hyperstack helps handle huge amounts of data and can be used for specific data analysis and anomaly detection in real-time. Sort through big data in no time, or improve deep video analytics performance with our VM clusters.
Deep learning
Push the boundaries of your neural networks with cutting-edge technology and expert support. Tackle complex problems in computer vision, NLP, and more, with Hyperstack.
Simulations & Visualisations
Visualise and simulate for scientific/medical research, architecture, and design. HyperStack provides the tools and infrastructure to unlock insights, optimise designs, and accelerate discovery.
High-Performance Computing
Our HPC solutions deliver the raw power you need for demanding simulations, scientific computing, and engineering workloads. Access cutting-edge hardware and software, all managed and optimised for peak performance.
Virtual Workstations
Ditch the desk, and embrace flexibility. Hyperstack allows you to create high-performance virtual workspaces in seconds, accessible from anywhere. Access demanding applications remotely, with seamless collaboration built-in.
Why Hyperstack
Managed Kubernetes
Deploy your Virtual Machine cluster in seconds and automate software deployment, scaling, and management.
Green Cloud
All of our infrastructure is 100% powered by renewable energy within our Data Centres in Europe and North America, and runs at peak efficiency.
Networking & Storage
Designed to handle high-demand workloads in a faster, 75% more cost-effective manner compared to legacy public cloud providers.
Speak To a Solutions Engineer
Our Solutions Engineers will personally assess your workload needs and provide cost optimisation recommendations based on your specific use cases.
Enterprise-grade features
- Easy-to-Use: 1 click deployment available
- Wide range of pre-configured SKUs available
- Ultra-fast networking & storage
- Custom-build flavours
- Role-based access control
- Resource callbacks
- Logically organise different VMs into environments
- Ever-growing fleet of GPUs available
- Human support team
- First class API
GPU Pricing
We have Data Centres in Europe and North America. Our billing cycles are accurate to the minute, so you only pay for what you use.
GPU Model | VRAM (GB) | Max pCPUs per GPU | Max RAM (GB) per GPU | Pricing Per Hour | Reservation Pricing |
---|---|---|---|---|---|
NVIDIA H100 SXM 80GB | 80 | 24 | 240 | $ 3.00 per Hour | Starts from $2.25/hour |
NVIDIA H100 PCIe NVLink 80GB | 80 | 31 | 180 | $ 1.95 per Hour | Starts from $1.37/hour |
NVIDIA H100 PCIe 80GB | 80 | 28 | 180 | $ 1.90 per Hour | Starts from $1.33/hour |
NVIDIA A100 80GB PCIe NVLink | 80 | 31 | 240 | $ 1.40 per Hour | Starts from $0.98/hour |
NVIDIA A100 PCIe 80GB | 80 | 28 | 120 | $ 1.35 per Hour | Starts from $0.95/hour |
NVIDIA L40 | 48 | 28 | 58 | $ 1.00 per Hour | Starts from $0.70/hour |
NVIDIA RTX A6000/A40 | 48 | 28 | 58 | $ 0.50 per Hour | Starts from $0.35/hour |
NVIDIA RTX A5000 | 24 | 8 | 24 | $ 0.25 per Hour | Starts from $0.18/hour |
NVIDIA RTX A4000 | 16 | 6 | 24 | $ 0.15 Per Hour | Starts from $0.11/hour |
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
GPU Model
VRAM (GB)
Max pCPUs per GPU
Max RAM (GB) per GPU
Pricing Per Hour
Reservation Pricing
See What Our Customers Think...
I have tried a number of different GPU-accelerated cloud providers and Hyperstack has by far the best support and pricing on the market.
Brian, CEO
We've been running our workloads on HyperStack for a long time now. No other cloud provider offers a similar experience: cost-effective virtual machines with NVLinks, and super-fast storage and low-latency networking. Our company can run many more GPU-based machines compared to any other provider! And you pay for what you use: need one - deploy a VM! Finished your task? Hibernate it without losing your environment! You can have a powerful VM in minutes - something that you have to wait for even hours or days in AWS/Azure to provision and join the cluster!
But the real benefits are:
- Ultra-fast VMs backed by a low-latency network and very fast volumes either ephemeral or external ones
- Multiple data centre locations
- Cheap VMs honouring NVIDIA Inception Program discounts (even cheaper under a contract!)
- API to manage VMs
- Very fast and helpful support team
Grzegorz, DevOps Engineer
“This is the fastest GPU service I have ever used.”
Ahmad, CEO
You guys rock!! You have NO IDEA how badly I need a solid GPU cloud provider. AWS/Azure are literally only for enterprise clients at this point, it's impossible to build a highly technical startup and get hit with their ridiculous egrees fees. You guys have excellent latency all the way down here to Atlanta from CA.
By far the most important aspect of a cloud provider, only second to cost/quality ratio, is their API. The UI/UX of the console is extremely well designed and I appreciate the quality. So, I’ll be diving into your API deeply. Other GPU providers don’t offer a programmatic way of creating OS images, so the fact that you do is key for me.
Anonymous user
Subscribe to Hyperstack!
Enter your email to get updates to your inbox every week