Intel Unveils Xeon Chips for Entry-Level Workstations

The Xeon E-2100 processors round out a chip lineup that includes Xeon Scalable Processors and Xeon W chips for high-end and midrange workstations.

Intel Xeon workstation

Intel is filling out its lineup of Xeon workstation processors with the latest chips aimed at entry-level systems and mobile workloads.

The 14-nanometer Xeon E-2100 chips, which are Xeon E3 processors based on a derivative of the “Coffee Lake” cores, are designed to offer capabilities that deliver a 36 percent performance improvement over the E-1200 v6 chips that were launched last year and 73 percent more performance over processors from four years ago.

The new processors join the Xeon Scalable Processors and Xeon W chips that were launched in September 2017 for workstations. The Xeon Scalable Processors are aimed at high-end workstations for professionals working on design and modeling applications and such newer workloads like highly immersive augmented and virtual reality and data analytics. The Xeon W-2100 processors target the midrange, with enhance memory, security and reliability features.

The latest Xeon E workstation chips are designed for creative professionals for such workloads as financial services, compute-intensive and 3D modeling and animation applications that touch a broad array of verticals, from energy to education to health care, according to Intel officials.

A range of single-socket system makers are expected to introduce workstations based on the new processors, Jason Kennedy, director of data center product marketing for Intel’s Data Center Group, said during an online press conference. HP Inc., Lenovo and Dell are among the OEMs that will announce new systems, and others will follow, Kennedy said.

The E-2100 processors will come in 10 variations that will range from four to six cores, four to 12 threads, 3.3GHz to 3.8GHz, and 8MB to 12MB of memory cache. With Intel’s Turbo Boost technology, the speeds can be ramped up from 4.3GHz to 4.7GHz. Seven of the chips also will come with Intel’s UHD graphics, and all will have 40 PCIe channels and two channels of DDR4-2666 memory support. They will consume 71 to 95 watts of power, and range in price from $193 to $450 per 1,000 sold.

The six cores are two more than were offered on the chips’ predecessors, and they come with a higher single core turbo frequency. They also feature Error Correcting Code (ECC) memory to ensure data integrity and improve system stability, according to Intel officials. Plus, there is support for USB 3.1 and Thunderbolt technologies and Intel’s Optane memory. Enhanced security comes from Intel vPro technology and Software Guard Extensions support for the vendor’s Ethernet and Wireless-AC networking products.

Working Toward 10nm

The E-2100 processors are part of the Intel’s continued extension of its 14nm manufacturing process as the company works toward 10nm. In response to a question, Kennedy said that the company has “gone through significant enhancements with that same node. We’re continuing to enhance the sophistication and densities and qualities of 14 nanometer as we go along.”

Intel’s struggle to reach 10nm has been highly publicized. Company officials have continued to build out more 14nm releases while moving the expected release dates of 10nm down the road. During a call with financial analysts in April, former CEO Brian Krzanich said the company continues “to make progress on our 10-nanometer process. We are shipping in low volume and yields are improving, though the rate of improvement is slower than we anticipated. As a result, volume production is moving from the second half of 2018 into 2019.”

Intel has taken steps to improve the yield of the 10nm chips, though they will take time to implement, Krzanich said, according to a transcript on Seeking Alpha. In the meantime, Intel has continued to improve the 14nm nodes and has “Whiskey Lake” chips for PCs and “Cascade Lake” processors for data center systems coming later this year, he said.

“While it's taking longer and costing more to deliver and yield advanced process technologies, we are able to optimize our process and products within the node to deliver meaningful performance improvements,” Krzanich said, noting 70 percent performance gains in 14nm chips since the first ones were introduced four years ago.

During the press conference, Kennedy was asked about rival Advanced Micro Devices’ Ryzen chips, which are based on the company’s “Zen” architecture.

“We always take all of our competition—and AMD specifically—very seriously,” he said. “They have some renewed energy and some theoretical excitement around their products. We do always look to make sure that our offerings are as compelling as possible from an overall capabilities perspective. In this case, it’s a bit of a challenge for us because we haven’t really been able to find equivalent workstations in the space to do any head-to-head comparisons even if we wanted to.”