Training models on Intel Gaudi accelerators is helping improve efficiency in Intel’s factories.

Intel’s smart manufacturing environment relies on AI to improve yield, perform root-cause analysis, and accelerate anomaly detection. As Intel transforms to an internal foundry model, it is crucial that the Manufacturing Automation team helps improve factory efficiency by enabling faster, more cost-efficient deep-learning (DL) and machine-learning (ML) workloads.

We recently benchmarked a manufacturing object detection DL model training workload. We compared the performance of our legacy GPU-based system, a vendor upgrade to the GPUs, and Intel® Gaudi® accelerators. Our AI-based solution using Intel Gaudi accelerators produced the following benefits:

• A 20% improvement in DL model training time

Intel’s AI hardware and software ecosystem is a competitive and efficient solution for high-volume manufacturing (HVM) use cases. The manufacturing efficiency offered by Intel Gaudi accelerators will help us improve production yields and increase Intel factory revenue.

Intel IT adopted a “right workload, right place” multicloud strategy nearly 10 years ago. This strategy has accelerated application development and deployment agility, achieved cost-responsible scalability, and enabled our business partners across Intel to achieve their business goals.

A decade of validation has underscored that emphasizing business value in cloud hosting decisions is advantageous. However, we continue to adapt our multicloud strategy as both business needs and the IT ecosystem evolve.

Some of those adaptations include the following:

But the cloud journey is never finished. New trends in the enterprise—such as the increasing use of AI, as well as innovations in the IT landscape—mean that we must continue to evaluate and evolve our multicloud strategy to consistently support the business in the most cost-efficient manner possible.

To benchmark the end-user PC experience, Intel IT uses tools that clarify PC power consumption and applications’ effect on battery life. This helps us optimize the IT build and answer one of the most common questions we receive from users: “Why doesn’t my corporate laptop battery last as long as my laptop at home?”

We know employees want to be as productive as possible, and battery life is crucial to Intel’s highly mobile workforce. On the other hand, our IT build includes several components that help keep Intel’s data and intellectual property safe. Necessarily, these additional software components consume their fair share of power. To fully understand what users are experiencing with their laptops—as well as how the Intel IT build introduces power consumption overhead—we rigorously benchmark the corporate PC experience.

In collaboration with Intel’s Client Computing group, we recently overhauled our PC experience benchmarking methodology and toolset. We adopted new benchmarking tools that more accurately represent the applications that Intel employees use. We also now use the Intel® Battery Life Diagnostic Tool (Intel® BLDT) to identify specific background processes and applications that are responsible for excessive power consumption.

We use these tools to reduce power consumption rates on users’ PCs and help the Client Computing group drive improvements to client devices and meet customers’ expectations. By sharing our benchmarking approach and results, we hope to encourage other IT departments to take a similar approach.

Intel IT’s Information Security (InfoSec) group has adopted capability-based planning (CBP) because it provides a framework for aligning our group’s strategies with Intel’s overall business strategy. CBP helps us prioritize projects and resources, communicate more effectively with business stakeholders, reduce complexity, and become more agile and responsive to changing business needs.

CBP is a planning methodology that drives business-focused outcomes by providing a systematic, objective, and holistic view of a group’s capabilities. In our case, these capabilities center around InfoSec, such as digital identity and access management and privacy management. Using a four-phase cyclical process of alignment, assessment, planning, and management, we can:

This comprehensive method of capability mapping and scoring delivers a unified view of capabilities and objectives so that teams can collaborate and optimize cross-functional efficiency.

Adopting CBP has provided several benefits to the InfoSec group:

As other Intel IT groups—and even business units—see the positive outcomes of CBP, we hope they also begin to adopt it, which would multiply the benefits across all of Intel. Furthermore, CBP has already fostered collaboration across departments and teams.

Our comprehensive method of capability mapping and a formulaic scorecard captures the strategic importance, business goal importance, capability maturity score, and quantified risk. It highlights what is required to execute initiatives successfully. Our solution addresses our longstanding InfoSec planning challenges but can also be applied to other IT groups and even other industries. Adopting CBP is an excellent way to align every decision with its strategic vision and capabilities and drive sustained enterprise success.

“Learn how Intel IT is gaining operational efficiencies resulting from a 5G private network in Intel’s manufacturing environment.

Minimizing factory downtime and improving operational efficiency have always been the driving forces behind Intel IT’s push to digitize equipment in both factory and subfab environments. Now, we’re expanding these efforts by working with Intel’s Corporate Services (CS) team to connect critical factory-support infrastructure, including surveillance systems in guardhouses, electrical supply equipment, water systems, and more.

Connecting equipment in remote outdoor areas to the corporate network has always been a challenge. Wi-Fi coverage is limited; hard wiring is costly and time-intensive; and relying on public networks introduces concerns with cost, performance, and security. After reviewing the options, we chose to roll out 5G private networks at Intel factories to extend wireless connectivity for key use cases in Internet of Things (IoT), robotics, operations, and security. This deployment, combined with 5G-enabled IoT Gateways and an array of sensors, lets us reach further and deliver more robust connectivity where it’s needed.

With our 5G private network in place, CS technicians can tap into sensor data to gain deeper visibility into equipment health, enabling predictive and preventative maintenance. This enhanced connectivity also allows technicians to connect field devices, like tablets, directly to the corporate network to increase efficiency. Currently, our 5G private program supports only 13 use cases—a fraction of the potential applications. Our 5G deployment spans five factories and is projected to generate a net present value of USD 35M over the next five years.

We hope that by sharing our 5G private network journey, we can inspire other IT departments to explore using this technology to continue the digital transformation of their previously unconnected environments.”

Thousands of Intel’s employees connect to the corporate Wi-Fi network every day. When connectivity issues occur, both productivity and user experience (UX) suffer. Intel IT makes every effort to optimize the network’s performance. Still, until recently, we had little visibility into client behavior through traditional network monitoring solutions because they report issues only from the network infrastructure perspective. At the same time, endpoint IT tools are limited to operating system data.

Based on its existing market presence with PC Wi-Fi products, Intel’s Client Computing product group developed Intel Connectivity Analytics technology. The product group collaborated with Cisco—a well-known enterprise Wi-Fi network infrastructure supplier—and Intel IT to develop compelling use cases and solutions that utilize Intel Connectivity Analytics. Intel IT provided product feedback and practical applications of the technology in our enterprise environment and helped optimize the use cases over Intel’s network infrastructure. Our input helped to better define the data and analytics needed to achieve the desired business outcomes, such as faster troubleshooting, mean time to repair (MTTR), reduced network total cost of ownership, and a better UX.

Intel IT now uses Intel Connectivity Analytics—delivered through the Cisco Catalyst Center (previously Cisco DNA Center) and cloud dashboard—to improve the management of Intel’s Wi-Fi network. We are also working with the overall ecosystem to expand the use cases for Intel Connectivity Analytics and further enhance AI for IT Operations (AIOps). We have reduced some client-side troubleshooting from upwards of 15 minutes to 10–15 seconds. Similarly, finding the root cause for ubiquitous network issues—which could take days—now often takes seconds as well.

We encourage other IT departments to consider deploying IT tools powered by Intel Connectivity Analytics for Wi-Fi, which is built into Intel Wi-Fi adapters and requires no software installation or maintenance. It can be utilized by Cisco and other Intel Connectivity Analytics Program members.

In addition, we are now working with the Intel Client Computing product group to develop an additional Intel Connectivity Analytics offering that shares data through a PC agent to the cloud and expands the available connectivity analytics beyond Wi-Fi to include Thunderbolt technology, Bluetooth, Ethernet, and more.

Increasing Electronic Design Automation (EDA) performance and throughput is critical to Intel’s silicon Design engineers.

Silicon chip Design engineers at Intel face ongoing challenges: integrating more features into ever-shrinking silicon chips, bringing products to market faster, and keeping Design engineering and manufacturing costs low. Design engineers run more than 273 million compute-intensive batch jobs every week. Each job takes from a few seconds to several days to complete.

As design complexity increases, so do the requirements for compute capacity, so refreshing servers and workstations with higher-performing systems is cost-effective and offers a competitive advantage by enabling faster chip design. Refreshing older servers also enables us to realize data center cost savings. By taking advantage of the performance and power-efficiency improvements in new server generations, we can increase computing capacity within the same data center footprint, helping to avoid expensive data center construction and reduce operational costs due to reduced power consumption.

To meet Design engineers’ computing capacity requirements, Intel IT conducts ongoing throughput performance tests using real-world Intel silicon Design workloads. These tests measure EDA workload throughput and help us analyze the performance improvements—and in turn, business benefits offered by newer generations of Intel® processors.

We recently tested two-socket servers based on the Intel® Xeon® Platinum 8400 and 8500 processor Series, along with the Gold 6400 processor Series. The test included operating single- and multi-threaded EDA applications running Intel silicon Design workloads for more than four days. Select results include the following:

• Higher frequency for per-core performance. For critical-path EDA workloads, selecting a high-frequency CPU like the Intel Xeon Gold 6444Y processor (32 cores per server) can deliver up to 1.14x higher per-core performance compared to lower-frequency, higher-core-count CPUs in the same generation of processors.

• Higher core counts for throughput. For volume validation runs, selecting a higher-core-count CPU at optimal frequency like the Intel Xeon Platinum 8462Y+ processor (64 cores per server) can deliver up to 1.75x higher Register Transfer Level (RTL) Simulation throughput per server when compared to a lower-core-count CPU (32 cores per server) in the same generation of processors. The Intel Xeon Platinum 8462Y+ processor (64 cores per server) completed workloads up to 2.17x faster than a previous-generation Intel Xeon Gold 6346 processor-based server, which has only 32 cores. Compared to a 2nd Gen Intel Xeon Gold 6246R processor (32 cores per server), the server with the newer processor outperformed the older processor by up to 2.40x in throughput.

• Additional benefits from 5th generation processors. At the same Thermal Design Power (TDP) of 350W, the Intel Xeon Platinum 8580 processor completed workloads up to 1.27x faster than a previous-generation Intel Xeon Platinum 8468 processor-based server.

Based on our performance assessment and our refresh cycle, we are deploying servers based on the 4th and 5th Gen Intel Xeon Scalable processor family in our data centers. By doing so, we have significantly increased EDA throughput performance to improve the overall EDA design cycles and optimize time to market of Intel chips.

Learn how Intel’s manufacturing engineers are using natural language processing (NLP) to streamline failure mode and effects analysis (FMEA).

Intel Manufacturing Automation has developed an innovative methodology for performing FMEA in manufacturing by using artificial intelligence (AI) techniques to analyze users’ emotional tone—positive, negative, or neutral—about the manufacturing equipment.

Failure mode and effects analysis (FMEA) is a critical process in manufacturing that is used to identify potential failures, assess their impact, and prioritize preventive actions. The FMEA process involves breaking down the system into its individual components; analyzing each component for potential failure modes; determining the effects of these failures; and assigning a risk priority number based on severity, occurrence, and detectability.

Traditionally, FMEA involves time-consuming manual extraction and analysis of large text data from tool logs and other sources, leading to weeks of engineering effort per analysis. This approach must be repeated across large fleets of tools by the engineers who support these tools.

Our new approach uses natural language processing (NLP) and sentiment analysis (SA) to reduce the labor required for FMEA from weeks to seconds. When we compared the traditional FMEA approach to our SA-based system, the software discovered everything the engineers found with their manual system approach—plus even more that they missed. Our system performed FMEA on six months’ worth of data in under one minute, saving weeks of engineering time.

Our SA analysis extracts keywords from comments, notes, charts, and quality systems data to find words with negative connotations about a tool, such as “abort” or “fail.” We’ve customized the analysis for domain-specific language, such as words and phrases that are specific to the semiconductor fab environment (e.g., “defect” and “excursion”). We also implemented custom replacement keywords (“ABORT” vs. “ABT” vs. “ABRT”) to accommodate various abbreviations and spelling errors commonly found in tool logs and other data sources.

The system analyzes sentence structure and filters out brackets for HTML and other special characters used in programming languages, allowing for a wide variety of inputs. While Intel’s manufacturing logs are predominantly in English, the libraries can be extended to handle multilingual input.

Integration with our in-house factory data analytics platform, Data on the Spot (DOTS), further democratizes FMEA results by enabling users to easily access this data and use it to identify root causes from the original sources. The system is now deployed across all Intel semiconductor factories, with a focus on front-end manufacturing.

The Intel® architecture-based components of our new FMEA system enable us to deploy new features rapidly. The FMEA system is powered by Intel® Xeon® processors, which provide the computational power and performance needed for real-time analysis of large datasets. The system uses high-speed SSDs and advanced memory configurations to ensure rapid data access and processing. Fast Intel® Optane™ persistent memory is employed to enhance caching capabilities and accelerate data retrieval.

This transformative approach promises to dramatically streamline the FMEA process, freeing up engineers’ time to focus on developing innovative solutions to further enhance Intel’s manufacturing tools and processes.

“Why is Intel IT modernizing Windows client management? Because a large part of Intel IT’s job is managing client devices—175,000 end-user and shared PCs for 124,000 employees. We need to consider many factors. Does a PC have the right drivers and security patches? How do we get new hires provisioned quickly? How do we ensure end users have access to the corporate resources they need, while still protecting resources from unauthorized access?

Creating efficient client management processes is crucial to maintaining IT efficiency as well as bolstering a positive user experience (UX). That is why we continue our journey toward modern, cloud-native client management with cloud-first capabilities.

Over the last few years, we have migrated both existing devices and new devices to a modern client stack composed of cloud-based components such as Microsoft Intune, Entra ID, Autopilot, and Windows Update for Business with Delivery Optimization. Together, these components support seamless client provisioning that improves the UX for end users as well as substantially reduces IT effort associated with client management.

Benefits accruing from the move to the cloud are plentiful. A few examples include:

We still have additional plans for continuing our modern client journey. However, we—and our users—are already reaping huge rewards, such as the following:

This white paper shares our client management strategy, our roadmap, and our key learnings. We hope this story encourages other IT departments to pursue the benefits of cloud-based client management and makes their journey a little easier.”

Intel IT’s data center strategy is to apply breakthrough technologies, solutions, and processes to accelerate Intel’s business. We also run Intel data center services like a factory, e?ecting change in a disciplined manner. Our approach enables us to optimally meet Intel’s business requirements while providing our internal customers with e?ective data center infrastructure capabilities and innovative business services.

Building on previous investments and techniques, our data center strategy has generated savings exceeding USD 9.6 billion from 2010 to 2023. We are continually improvising our data center strategy to further our data center transformation. We will accomplish this by using disruptive server, storage, network, infrastructure software, and data center facility technologies. These can lead to unprecedented quality-of-service levels and a reduction in total cost of ownership (TCO) for business applications. And they will enable us to continue to improve IT operational efficiency and be environmentally responsible.

Transformational elements of our data center strategy include the following:

Some of our most notable achievements include:

Our data center transformation strategy is critical for Intel IT to stay competitive. We are committed to providing a foundation for continuous innovation that will improve the quality, velocity, and efficiency of Intel IT’s business services.