NVIDIA Blackwell
Rent GB200 servers, reserved from data centers
GB200 joins a Grace CPU to two Blackwell GPUs and is deployed as the GB200 NVL72 rack, 72 GPUs connected by NVLink across the rack. It is the first Blackwell system built around rack-scale NVLink rather than the eight-GPU server.
- Memory per GPU
- 186 GB HBM3e
- Architecture
- Blackwell
- Listings
- On request
- From
- Ask
GB200 capacity listed now
No GB200 capacity is listed right now. Send a request and we will source it from our owners.
Tell us what you needWhat GB200 is good for
- Large training runs that benefit from one 72-GPU NVLink domain
- High-throughput serving of the largest open models
What to ask before you reserve GB200 capacity
- Whether you get whole racks or a share, and the rack's power and cooling.
- How the servers are interconnected (InfiniBand or RoCE, and at what speed per GPU) if you will train across nodes.
- Whether the price covers storage, egress and support, or only the GPU-hours.
- The uptime they commit to, and what you get back when they miss it.
Questions
- How much does it cost to rent a GB200?
- Owners set their own price per GPU-hour, which depends on the number of servers, the term and the start date. Send a request with what you need and we come back with offers.
- What is the minimum term for reserved GB200 capacity?
- Usually one month or more, set by each owner. Longer terms usually get a lower price per GPU-hour.
- How much memory does a GB200 have?
- 186 GB HBM3e per GPU (NVIDIA Blackwell).
- Can I rent a single GB200 server?
- Yes, where the owner allows it: each listing shows how many servers it has and you can ask for one or for all of them.
Tell us what you need
Sign in to send a request. It opens a conversation with the owners and the Paralon team, and you get an email for every reply.
Sign in to requestFree, with Google, GitHub or an email code. No commitment.