Intel Brings Xeon Scalable Chips to High-End Workstations

The processors are aimed at such emerging workloads as artificial intelligence, virtual reality and data analytics.

Intel Xeon

Intel is targeting emerging workloads such as virtual reality, analytics and artificial intelligence by bringing its new Xeon Scalable chips to workstations.

About six weeks after introducing the latest iteration of its Xeon server processors, Intel this week made two new Xeon Scalable chips available for what company officials call next-generation expert workstations that can be used for such tasks as creating highly immersive 3D virtual reality (VR) media and getting useful insights through faster and deeper data analytics.

“These new Intel Xeon processors for workstations deliver the features and capabilities elite professionals need to bring their digital creations to life,” Lisa Spelman, vice president and general manager of Xeon products and data center marketing at Intel, wrote in a post on the company blog. “Whether it’s professional-grade virtual or augmented reality, content creation in full 8K resolution, advanced computer aided design and simulation, or innovations we’ve yet to discover, these robust processors empower creative and technical professionals with tools to match their imaginations.”

Spelman said she expects a broad array of workstation OEMs—including Dell, HP Inc., Lenovo and Fujitsu—to launch systems with the new dual-socket 14-nanometer Xeon chips.

At the same time, Intel also unveiled new Xeon W processors for mainstream workstations. Officials said workstations with the new Xeon W chips will deliver 1.38 times the performance of their predecessors and 1.87 times the performance of 4-year-old systems. The single-socket chips, which will offer up to 18 cores, 36 threads and 4.5GHz (when leveraging Intel’s Turbo Boost 2.0 technology), include enhanced memory and reliability capabilities and hardware-enhanced security.

Intel introduced the new Xeon Scalable architecture—based on the “Skylake-SP” core—in July, with executives claiming it was the most significant advancement for the data center platform in a decade, with significant improvements in scalability, security, performance and power efficiency. It comes at a time when Intel—the dominant supplier of server chips, with more than 95 percent of the global market—is facing increased competition from a reinvigorated Advanced Micro Devices and ARM-based chip makers like Qualcomm and new data center workloads are emerging, including VR, augmented reality (AR), artificial intelligence (AI), machine learning and data analytics.

The new chips come with a remade microprocessor and Intel’s new Mesh Architecture for faster connections between processor cores and other components on the chip, expanded memory, security and interconnect capabilities and additional cores. The processors offer up to 56 cores and up to 112 threads, with frequencies up to 4.2GHz and up to 3TB of DDR4 memory. They’re being adopted by most major server makers, including Dell EMC, Lenovo, Hewlett Packard Enterprise, Cisco Systems and Cray.

According to Intel, workstations with the new Xeon Scalable chips will offer 1.65 times the performance of current systems and 2.71 times the performance of workstations that are 4 years old.

The new chips will compete with expected workstation processors from AMD, which earlier this year came out with its high-performance and power-efficient Epyc server chips based on the company’s Zen architecture. AMD officials have said the chips will challenge or beat Intel in everything from performance, power consumption and cost, and represents the company’s best chance in more than a decade to gain market share against its larger rival.