Xeon Scalable processors: Built for HPC workloads

Feed your high-performance workloads with IBM Cloud bare metal servers running the second generation of Intel Xeon Gold and Intel Xeon Silver Scalable processors. With increased DDR4 memory speed and capacity, Intel’s processors also provide impressive “per core” performance and advanced reliability.

Intel’s next generation Xeon Scalable processors on IBM Cloud bring innovation to the way you move, store and process data when it comes to HPC workloads. Enhancements include higher per-core performance and CPU frequency, PCIe 3.0 support for 48 lanes, up to three Intel UPI links, increased DDR4 memory and speed capacity, and an optimized cache hierarchy.

Run your most compute-intensive workloads

Earth systems models
Add greater precision and higher speeds to prediction systems, and serve a wide range of meteorological applications.

Manufacturing
Help reduce dev-and-test cycle times for mission-critical design processes, including automotive, nautical, aerospace and more.

Healthcare
Run power-intensive simulations. Deliver fast and precise calculations for complex materials and analysis.

Financial sector
Mirror on-premises environments and speed up statistical sampling techniques.

Explore analytics without boundaries

Get analytics without boundaries with IBM DB2® on Intel Xeon scalable processors. These processors deliver faster, more sophisticated analytics; run more concurrent queries in multi-user environments; and uncover high-value insights as fast as data is created.

Intel Optane: High throughput, low latency and high QoS

Break through data-access bottlenecks and get more work done per server with Intel Optane SSD DC P4800X. This innovative SSD technology is an industry-leading combination of high throughput, low latency, high QoS and ultra-high endurance. It’s the first product to combine the attributes of memory and storage, delivering a solution that can accelerate applications with faster caching and faster storage performance, can increase scale per server and can reduce transaction costs for latency-sensitive workloads.