NVIDIA H100 computing card landed in the Japanese market
At GTC 2022, NVIDIA released a new generation of H100 based on the Hopper architecture for the next generation of accelerated computing platforms. The NVIDIA H100 has 80 billion transistors, is a CoWoS 2.5D wafer-level package, a single-chip design, is manufactured using TSMC’s 4nm process, and is a tailored version for NVIDIA.
Nvidia said it expects to start supply in the third quarter of this year, but did not give a price for the H100 computing card. Recently, a retailer in Japan has listed the Nvidia H100, showing a price of 4,745,950 yen ($36,567.5). The price includes shipping and taxes. If only the card itself is calculated, it is 4,313,000 yen ($33,231.7). The H100 is available in SXM and PCIe form factors to support different server design requirements. The Japanese retailer released a PCIe-based version this time.
The full GH100 chip is equipped with 8 sets of GPC, 72 sets of TPC, 144 groups of SM, and a total of 18432 FP32 CUDA cores. It uses the fourth-generation Tensor Core, a total of 576, and is equipped with a 60MB L2 cache. However, not all of them are opened in the actual product. Among them, 132 groups of SMs are enabled in the SXM5 version, with a total of 16896 FP32 CUDA cores, 528 Tensor Cores, and 50MB of L2 cache, while the PCIe 5.0 version has 114 groups of SMs enabled, and the number of FP32 CUDA cores is only 14592. In addition, the TDP of the SXM5 version reaches 700W, and the PCIe 5.0 version is 350W.
In addition, the H100 supports NVIDIA’s fourth-generation NVLink interface, which can provide up to 900 GB/s of bandwidth. At the same time, H100 is the first GPU to support PCIe 5.0 standard, and also the first GPU to use HBM3. It supports up to six HBM3s with a bandwidth of 3TB/s, which is 1.5 times that of A100 using HBM2E. The default memory capacity is 80GB.
