The Intel Xeon Phi 7290 is the fastest chip that the company has launched and of course, also the most expensive. The Xeon Phi 7290 integrates no less than 72 processing cores at a frequency of 1.50 GHz each. The company will begin to ship it in September at a price of $ 6,294.
Intel launches 72-core Xeon Phi chip against GPUs in machine learning applications
Built in what seems a graphics card that uses a PCI-Express interface, Intel Xeon Phi 7290 is accompanied by 16 GB of HBM memory, resulting in an impressive bandwidth of 490 GB / s with a TDP of 245W. It also consumes additional 15W due to Intel Omni-Patch Fabric, dual port helps to reduce the cost of the new chip. In addition to the Intel Xeon Phi 7290, the company also announced three more chips.
At the second we have Xeon Phi 72100, a cheaper model with 64 cores @ 1.30 GHz, but loses its memory bandwidth of 7.2 GT / s to stay in the 6.4 GT / s seeing reduced power consumption to 245W + 15W. In both cases, it supported the use of 384 GB RAM: 2400 MHz for the flagship and 2133 MHz of “economic models“.
The new chips are bundled like graphic cards and can work as essential processors or co-processors. The packaging of these chips shows that Intel is endeavoring to tackle NVIDIA in the HPC market. In supercomputers and servers, the chip will serve as a co-processor to Xeon chips as the essential CPU. These chips can likewise be introduced as an essential processor in workstations.
Features of the Intel Xeon Phi family:
Connectivity: Intel Xeon Phi processors now offer the ability to deploy Intel’s next generation fabric architecture, Intel Omni-Path, as a standard, integrated option. The combination provides the foundation for powerful and efficient data traffic control by using an advanced “on-load” design that cost-effectively maximizes performance.
Intel Omni-Path architecture automatically scales to tens of thousands of nodes with extremely low power consumption, making the solution ideal for increasingly demanding HPC and analytics workloads.
Storage: Intel Solutions for Lustre software is a parallel filesystem purpose-built for HPC to provide fast, “massively scalable storage software” for accelerating performance on complex workloads.
Memory: MCDRAM is the integrated memory on Intel Xeon Phi processors, providing up to 5x the bandwidth of DDR4 to deliver breakthrough performance for memory-bound workloads.
Software: Intel HPC Orchestrator simplifies the complexity of installation, configuration and ongoing maintenance of implementing an HPC software stack.