Superchip
Also known as: Grace Hopper, Grace Blackwell
A CPU and GPU on one module with a fast link between them, such as the GH200 and GB200. The GPU can use the CPU's memory as an extension of its own.
GPUs with Superchip, best value first
Ranked by the lowest on-demand price per GPU-hour.
Compare every GPU →Related terms
- SXM vs. PCIeSXM GPUs sit on a baseboard with NVLink between them and get more power, so they run faster. PCIe cards plug into a normal slot, cost less and are usually slower. An H100 SXM and an H100 PCIe are different products.
- NVLink and NVSwitchNVIDIA's high-speed links between GPUs in one server, far faster than PCIe. They matter when a model is split across GPUs. AMD's equivalent is Infinity Fabric.
- VRAMThe GPU's own memory. The model weights, activations and KV cache must fit in it (or be split across GPUs), so VRAM often decides which GPU you can use at all.
Hardware and interconnect
- InfiniBandThe low-latency network used to connect GPU servers in training clusters. Multi-node training needs it (or fast RoCE Ethernet); single-node jobs don't.
- NodeOne server. Datacenter GPUs usually come as 8-GPU nodes; some providers only rent whole nodes. PetaPrice always shows the price per GPU.
- TDPThermal design power: roughly the most power the GPU draws, in watts.
- MIG and fractional GPUsMulti-Instance GPU splits one A100, H100 or newer into isolated slices. Slices and other fractional GPUs (vGPUs) are cheaper but aren't full GPUs, so PetaPrice leaves them out.