Shortly after launching the 13th generation Intel Raptor Lake processing chip series, they seem to have wasted no time in introducing a new series of processing chips for their data centers, through the Intel Xeon Max and Intel Data Center GPU Max processing chips.
The Intel Xeon series of processing chips are typically known for use in data centers, and with this latest model, it's no different, only this time the processing chips are developed with a specific task, leveraging the latest Sapphire Rapids chip technology.
Intel Xeon Max is equipped with up to 56 physical cores and up to 64GB of built-in HBM memory and support for PCIe 5.0 and CXL 1.1 data transmission technologies specifically for data center use. It also has a total power consumption of 350W which Intel says is 68 percent more efficient than the AMD Milan X.
Typically, a processing chip for a data center is introduced for various purposes, either to process a lot of data, and also various other uses such as storing large-sized data and so on.
For this Xeon Max processing chip, Intel says that it was developed specifically for tasks such as weather forecasting and also for medical research purposes such as finding cures for cancer and so on.
As for the Intel Data Center GPU Max graphics card, this graphics card comes with 128 Xe-HPC cores with a total L2 cache of 408MB with 48GB of built-in HBM2e memory. The Data Center Max graphics card series consists of three models, namely Max Series 1100, Max Series 1350 and Max Series 1550.
The graphics card comes with support for Ray-tracing, which Intel says will be leveraged for faster science visualization purposes. In terms of performance, this graphics card is seen with the ability to process credit risk financial models and digital reactor simulations at a faster rate than the NVIDIA A100 graphics card.
Intel says that the Intel Xeon Max and Data Center Max processing chips and graphics cards will be introduced in the market around the beginning of next year for companies that need them. Intel Xeon Max processing chips are expected to power supercomputers at Los Alamos National Laboratory, Kyoto University, while Intel Data Center Max is expected to be used in Aurora supercomputers at Argonne National Laboratory.