Understanding the 4th Gen Intel Xeon Scalable Processor

January 1, 2023

Gordon Moore, the co-founder of Intel Corporation, predicted the future of computer processing power (CPU) by stating that the number of transistors in an integrated circuit would double every other year. That has certainly held true, and today, the industry is transitioning from the 3rd to the 4th Gen Intel Xeon scalable processor, which companies such as UNICOM Engineering are putting to use.

What are some of the changes with this new processor, and how is UNICOM Engineering leveraging the 4th Gen?

On today’s episode of Flawless Execution, Host Daniel Litwin speaks with Austin Hipes, Chief Technologist and VP Engineering at UNICOM Engineering, to discuss how this update will impact solution providers in the industry.

Litwin and Hipes also discussed…
1. What differentiates the 3rd and 4th generation Inton Xeon processors in terms of the memory architecture and express lanes
2. Where UNICOM is continuing to elevate standards with their clients using the 4th Gen Intel Xeon
3. How the move from the PCI Express form 4 to 5 will impact the longevity and life cycles, and how it connects to Xeon’s scalable architecture

Hipes detailed the best use cases for the 4th Gen Intel Xeon processor: “Anything to do with networks and security. Those are huge focuses and features that will get into some of the acceleration available on the CPUs. Those are things that traditionally have chewed up a lot of CPU cores when you try to use them for general purposes. So, a lot of effort has been put into accelerating those workloads without having to tax those general-purpose cores as much.”

Hipes has worked as a Field Application Engineer at Arrow Electronics and Wyle Electronics. He attended The University of Texas at Dallas and holds an Associate Certificate from Dell EMC in Information Storage and Management Version 3.0. Hipes has various specialties, including ATCA, Carrier Grade Rackmount Servers, Networking, and more.

Recent Episodes

gpu deployment
View episode

During a transformative period, Applied Digital experienced a significant increase in demand for large-scale GPU deployment. This demand required a focus on power density, as the networking within data centers, particularly for InfiniBand, necessitated that servers be located close together. Each server consumed 10.2 kilowatts of power, and the optimal performance was achieved when…

power sources
View episode

Applied Digital optimizes high-performance computing by leveraging unique power sources, and locating facilities at the source of power rather than in traditional cloud regions. This approach is particularly suited for AI workloads, which do not require ultra-low latency like video streaming. By targeting areas with abundant but underutilized power, known as “stranded power,” the…

latency sensitivity
View episode

Latency sensitivity in AI processes varies significantly between training and inference. Training operations, which involve processing large datasets over extended periods, are generally very tolerant of high latency. This tolerance allows training tasks to be performed with minimal concern for immediate responsiveness. Wes Cummins, the CEO of Applied Digital joins David Liggitt, the Founder…