HARDWARE PLATFORMS

Christina Cardoza: Hello and welcome to the IoT Chat, where we talk about the latest developments in the Internet of Things. I’m your host, Christina Cardoza, Editorial Director of insight.tech. And today we’re going to take an inside look at the latest 4th Gen Intel^® Xeon^® Scalable processors, code name “Sapphire Rapids,” with Christine Boles from Intel. Christine, welcome to the show.

Christine Boles: Hello, Christina. It’s great to be here.

Christina Cardoza: Before we jump into the conversation, I would love to learn a little bit more about you. And you have quite the impressive career over there at Intel. So, what can you tell me about it?

Christine Boles: Yes, Christina. Yes. It has been a long and prosperous career here. So, today I am the leader of the Federal Industrial Solutions Division within the Network and Edge group, where I’m a Vice President within that Network and Edge organization. And I like to say I have one of the best jobs at Intel, and frankly in the industry, where I have the opportunity to utilize incredible assets that Intel has from products and technologies that are really helping many in the industry utilize the capabilities they need for this transforming space that is really rapidly transforming in both the federal and industrial areas.

Christina Cardoza: Great. And of course one of those technologies we’re talking about today—the 4th Gen Xeon Scalable processors, which Intel announced at the beginning of the year. I’d love to start off with that release. What can you tell me about these latest processors? And, as it relates to the federal and industrial space, what makes it so exciting for them?

Christine Boles: Yes, it is exciting to see these processors that really have been designed to deliver incredible capabilities for very demanding workloads. And many of those workloads are in the industrial and federal space. And if you really look at what’s been happening in the—and I’ll focus on the industrial space—in the industry space, the Industry 4.0 transformation that’s been underway, this sector and these areas have really been looking for technologies that help deliver capabilities that are really going to extend and increase the business value and address some real challenges that manufacturers or utilities in other areas are looking for.

And so this 4th Generation Intel Xeon Scalable processor, with all of the new capabilities that it’s brought in, particularly in the areas of acceleration for AI and machine learning and data analytics types of capabilities, in addition to networking and storage, really are exciting areas. And Intel really stepped back and re-architected the microarchitecture in order to really address some of these dynamic workloads—whether it is in the networking space or in the area that I focus on: what is happening in the industrial edge. And really extending the processing capability, but within a great performance-to-power performance area, while also extending the memory capacity in the IO that you need in these industrial kinds of capabilities.

So it really is an exciting kind of solution that is made possible with this 4th Generation processor. And perhaps, Christina, I can extend this a bit more and talk about some of the specific areas where we have added some capabilities in. One of the areas is really around the deep learning/machine learning areas, by putting some additional acceleration into the CPUs versus what we’ve had in the past from our Xeon Scalable processors. And two of those areas are the Advanced Matrix Extensions—the AMX extensions—as well as the Intel^® Data Streaming Accelerator.

These AMX extensions are really what is that addition for accelerating the AI capabilities in the workloads that you will see in the industrial space, such as machine vision, defect detection, or even quality assessment of what is happening in the equipment as well as of the products moving down the line. And as you can imagine, when you’re in the industrial space there’s a lot of data that manufacturers or oil and gas providers really need to deal with. And having the Data Streaming Accelerator to really prioritize and manage the data through the virtualized environments, as well as presenting the information, is really a benefit that the industrial-solution providers can tap into to meet the demanding needs that the industrial space has.

And I guess one other area that I’d call attention to—and I’m really excited to see how this will be utilized by the industrial-solution providers—is the Intel Speed Select Technology. And this is SST technology that really will help with the concepts of where the industry has been going to consolidate workloads onto form factors that really have multiple workloads running. And the ability to select where you’re going to be processing and optimizing that performance where you need it in some of the virtual machines, and versus, say, other workloads that might not need as much, and really balancing that overall load.

So, there’s some really exciting technologies that we’ve built into this 4th Gen Scalable Xeon processor, that really are going to address some of the needs that the industrial and federal space is looking for.

Christina Cardoza: Absolutely. It sounds like it’s jam-packed with new and great features, especially all the AI stuff. Manufacturers, they want to become more intelligent. They want to leverage AI capabilities, and this sounds like it’s not going to only help them perform better, but perform faster and improve their operations. And you mentioned that the microarchitecture in this release was sort of reinvented also.

So, I think all of these capabilities sound great. And a lot of the things that people want to hear when Intel has new releases like this is what makes it different from the last release? Why should I move now from the release I’m on onto this one? And so you mentioned a couple of the new capabilities there, but what would you say are the main takeaways of how this latest Xeon Scalable processor performs or is compared to the previous generations?

Christine Boles: Yes, Christina. As I mentioned, there’s a myriad of improvements that we brought in in new capabilities. And there’s probably more than we can cover in a short conversation. But I guess four areas that I really see that are really going help the IoT edge, and what customers are looking for, and what solution providers will develop around.

The first is really in the area of performance and memory and IO. So, in the overall architecture we have a higher per-core performance than previous generations, and with up to 52 cores or different sockets for the range of IoT-edge use cases that customers will be looking for. We also have extended the capabilities in the area of memory, with eight channels of DDR5. Now, DDR5 allows for an overall 1.5x improvement in bandwidth over the DDR4 generation, and really will ultimately improve the performance and the capacity for memory utilization.

Now, when we talk about industrial use cases, one of the areas that really pushes limits is IO capabilities. And this generation has up to 80 lanes of PCI Express Gen 5, and test and measurement really pushes the limits of the IO capabilities. And having this range of IO is definitely a great area. And I guess the final IO area that I’ll mention is we do also provide the ability—we have great acceleration of our AI capabilities in the Xeon processor—but if you need additional CPUs or accelerators external, we do have the CXL 1.1 connectivity for interconnect to external devices. So one big category is that memory and IO and overall performance.

Now I did mention the AI acceleration that is the biggest—one of the biggest—additions to this product in the 4th Gen with those AMX—those Advanced Matrix Extensions—that we’ve added in. And of course we take it one step further and make sure you have the right toolkits to take advantage of that capability for workload inferencing and optimization utilizing Intel’s distribution of the OpenVINO^™ toolkit. And having both that improved AI acceleration with the toolkits will really help customers have the right support for deep learning and overall training of workloads utilizing this Xeon Scalable processor product line.

And the next area is what I mentioned a bit ago about that bringing together workloads utilizing the Intel Speed Select Technology. Now, this is really to allow for better control over the CPU performance and how you’re utilizing that performance and the compute power across the Xeon processor. And so we also do make available tools that allow for that monitoring control with the Intel Resource Director Technology toolkit that really enables the control and the sharing of the resources and managing the overall environment.

And then the fourth big area is of course resiliency and security—particularly in when you’re putting solutions into operations such as the manufacturing or federal types of environments. Having that efficiency and resiliency that Intel is known for with our processors, allowing for that trusted compute for the life of the product the deployment—so that’s a big part of what we continue to provide. And then of course we have the security extensions with the Software Guard Extensions to really allow for secure enclaves of execution of different applications, and really monitoring and assessing for security support in the platforms that we provide. So, a lot of great areas of improvements over the previous generation.

Christina Cardoza: That definitely sounds like it. And you guys are hitting all of the areas for Industry 4.0 success—security, reliability, flexibility, scalability—everything that manufacturers want to see. And so you mentioned a little bit about how this is being used in federal environments and manufacturing environments. I would love to dig a little bit deeper into this release, and talk about some use cases that you really think are going to benefit from the Xeon Scalable processors.

Christina Boles: Yes, there’s quite a few areas where the Xeon processors have really been set up for a range of edge IoT use cases in the federal and industrial, but actually additional areas, and I’ll touch on a few. The first big area within the federal and industrial spaces—really those use cases that have a high demand for compute, whether that’s machine vision kinds of applications or detecting defects that might be happening and then taking action upon them.

Of course there’s an emerging area of digital twin capabilities, of having both visibility and ability to have a representation of what is happening within the factory area. And then it’s expanding into evolving areas of automation—whether that’s within the utility space, with the modernization of the grid and bringing greater levels of software-defined capability and management of the utility and the electric-grid infrastructure or process automation.

And the capabilities that have been built into the next generation—one of the big areas will be in this machine vision area. How do manufacturers really improve the detection of defects, as well as quality inspection from a range of cameras, gathering that information in, accurately analyzing, and then acting upon the images that it’s bringing in. And those capabilities that we’ve built into the Xeon processor with the AMX extensions will really allow for these workloads to be processed and managed with the amount of information flowing through. And that’s inclusive of also the Data Streaming Accelerator area.

Now one area I’ve mentioned quite a bit is the manufacturing space; you’ll also see those same benefits in the use cases in the fed aerospace, where there they really need to utilize those capabilities of those new instructions and the vector extensions that we’ve integrated in for doing signal processing on algorithms—that signal processing algorithm on workloads that really need that additional capability for an analysis.

Now, those same kind of areas that we have in the federal and industrial—you can think about some of those improvements being utilized in more of those consumer-focused industries of the retail environment, the hospitality types of spaces. And really over the last few years you’ve really seen a change in what is made available as you’re going into your stores or into your hotel locations—with self-checkout kiosks, or even robotics for assessing what is happening with the logistics, say, in the warehouse and the back room and managing the overall inventory.

So, one area that the Xeon processors will really help is in that—that front end of the store, and as well as of course the back end as well. But we’ll focus on the front end, where these storefronts have really been changing, and they’ve been utilizing new capabilities to have interaction with consumers as they’re coming in, as well as assessing what is happening and preferences that the customers may have. So, having that 4th Gen Intel Xeon Scalable processor–based solution with the additional AI and analytics capabilities will really allow utilities to offer new products, offer new service positions for customers as they’re coming in, offer new service revenue, as well as overall assessment of what is happening in the environment in those commercial-storefront areas. So, great opportunities to utilize the capabilities of these Xeon-based solutions.

And I guess the last area—if you extend that use of visualization and gathering of information—there isn’t an area that has more of an opportunity than most than the health and healthcare and life sciences area. And how they might utilize the AI extensions and support that has been built into this 4th Generation Xeon processor. And really utilizing the information to assess images; to do advanced analysis on genomics, sequencings and other areas; and really utilizing the improvements that we’ve put into the products. So it’s going to be exciting to see how medical-equipment manufacturers utilize some of the new capabilities we’ve put into these processors.

Christina Cardoza: Yeah, absolutely. That is such a wide range of use cases. So, it’s great to see how every industry can start taking advantage of this. And one of the things you were talking about with all of these improvements and opportunities out there for these industries is they want those real-time insights, that real-time visibility, so that they can start making these informed-based decisions to really improve everything and move faster.

And one thing that I just want to talk about quickly is, as they start adding these new and advanced solutions into their operations, it starts to put some constraints on the network, and then they want to start moving closer to the edge so that they can utilize all this data much faster and perform much better. So I’m curious what role do you see the 4th Gen Intel Xeon processors playing as all these opportunities start adding more network workloads and start moving closer to the edge?

Christine Boles: Yes. This is one of the exciting parts of the next-generation platform that we’ve been working on with these Xeon processors, is a range of workloads that it enables. And one of those areas is really this shift that is happening from what have traditionally been more fixed-function network architectures, to what is evolving to be more of a software-based, virtualized network environment, particularly in the radio and access-network area. And ultimately these communication-service providers can utilize these capabilities for the optimizations that we put in, not just for the examples I’ve given, but also for the network workloads in how we need to optimize and manage the range of data, as well as the overall load balancing that you need within the network.

And so the Intel Xeon processor capability that we’ve built in really allows for the solutions providers to have more of a software-defined network environment, while at the same time having the AI and machine learning capability for the information that’s flowing through the network and optimizing it. So it’s a very interesting time, as more of the network becomes software and managing. And then you have this range of use cases that you can process alongside them in a converged, network-edge kind of capability. And the Xeon processor really helps make that possible, while at the same time really optimizing the overall performance-power kind of measurement that the overall network solutions are looking for. And at the same time having the right latency, low-latency capability, that you need in these network capabilities. So it’s an exciting place for us to see how the Xeon processors are going to help the industry move along.

Christina Cardoza: Yeah, absolutely. I can’t wait to see how industries just continue to transform and leverage these 4th Generation Xeon Scalable processors, and see what else they come out with, because, like you’ve mentioned in this call, there’s lots to take advantage of in this latest release. Unfortunately we are running out of time. But before we go, I just want to throw it back to you one last time. We talked about a lot of different features and capabilities here; so are there any final takeaways or thoughts you want to leave our listeners with today?

Christine Boles: Christina, I’ve talked a lot about the new features, but there’s—I’ll call it at its roots—this Intel Xeon Scalable processor that has brought capabilities to the edge really helps with not only the things I mentioned on the performance, the security, etc., but we’ve also kept in mind what you need going into more ruggedized kinds of environments, where you need to ensure that you have the reliability and the support for that.

And so we offer a broad range of SKUs of these processors. And the SKUs have really been specifically built catering for the long-life needs, as well as the reliability needs, of 10-year reliability and availability that industrial-commercial offerings need. And it’s also available in a more industrial temperature range of 0°C–84°C, such that they really can be used in the industrial environments.

In addition we offer the range of SKUs—so, where I mentioned the range of cores and performance and how it can be up to the numerous number of cores that the Xeon processor brings—it also can be scalable to really bring, “What does the edge workload really need?” and really balancing the power-performance needs, as well as bringing them into the specific environment. So, bottom line, we really have redefined and ensured that the solutions providers that are utilizing this 4th Gen Xeon processor can really have the right capabilities that they need in performance acceleration for AI workloads or networking workloads and analytics, but also have the range of power and performance that they need for the environments they’re going into. So, really excited to see the applications coming to market based on this new generation.

Christina Cardoza: Yeah, and some great final points there, especially when you think about the manufacturing industry. They operate in harsh environments, like you mentioned, sometimes, and their equipment has to last a long time. They’re not one to really “rip and replace” every couple of years; they sometimes have their equipment there for decades. So these 4th Generation Intel Xeon processors are really helping them future-proof their investments and meet the needs that they have of the factory today, while also looking forward to tomorrow. So, this has been a great conversation. I just want to thank you again so much for joining us and for the thoughtful conversation.

Christine Boles: Thank you, Christina.

Christina Cardoza: And thanks to our listeners for tuning in. Until next time, this has been the IoT Chat.

The preceding transcript is provided to ensure accessibility and is intended to accurately capture an informal conversation. The transcript may contain improper uses of trademarked terms and as such should not be used for any other purposes. For more information, please see the Intel^® trademark information.

This transcript was edited by Erin Noble, copy editor.