The Intel AI PC Sustainability Story
By Roberta Zouain, Sustainability Product Strategy & Marketing Manager, Client Segments
It is no secret that these days, sustainability is top of mind for most organizations. Whether due to a concern for the planet, regulatory requirements, corporate goals, or the fact that over 50% of young professionals consider a company’s sustainability practices as a key factor in choosing an employer[1]– decision makers across industries are seeking solutions that are better for the planet, and thus the bottom line.
It is also no secret that an organization’s tech stack can present one of the biggest challenges and opportunities when it comes to achieving sustainability goals. When discussing technology and sustainability – especially as AI scales across industries – our minds often go first to the expansive data centers and humming servers that power our digital lives. But there’s another piece of technology that can have a massive impact on your organization’s sustainability goals: your business PC.
At Intel, we’ve been dedicated to sustainability for decades, across every facet of our operations. So, it should be no surprise that when we designed Intel® Core™ Ultra, the technology that powers the AI PC, sustainability was top of mind at every stage of the journey. In this blog, I’ll share how we started with engineering and scaled through our ecosystem partners to make AI PCs on Intel Core Ultra sustainable by design.
Efficiency at the Processor Level
In developing new Intel® Core™ Ultra processors, our engineering teams have sought to reduce energy use at the processor level with our next-generation hybrid architecture design, which is comprised of a GPU (graphics processing unit), CPU (central processing unit), and an NPU (neural processing unit).
- What is next-gen hybrid architecture? This architecture uses the concept of Efficient-cores (E-cores) and Performance-cores (P-cores). Tasks that don’t require significant power are run on E-cores, and tasks that need more power are run on P-cores. For example, the task of watching a video locally could easily run on E-cores, whereas the task of editing a video is most efficiently run on P-cores. This optimization strategy saves energy.
We have also created tools like Intel® Thread Director that analyzes different types or workloads and places the right application thread on the right core, optimizing performance per watt. Using this approach, enterprise fleets of Intel Core Ultra-based notebooks can consume up to 64 percent less energy than is dictated by ENERGY STAR 8.0 requirements, efficiency benchmarks set by the U.S. Environmental Protection Agency.[2]
- A new integrated neural processing unit (NPU). The Intel Core Ultra processor includes an integrated NPU, a specialized component for running AI tasks. NPUs are specifically engineered for the types of calculations most common in AI and ML (machine learning), such as tensor operations and matrix multiplications. Running AI workloads on NPUs is more energy efficient than running them on a CPU or GPU.
- A new Xe-LPG GPU architecture. The Intel Core Ultra processor also includes an integrated graphics processing unit (GPU) built on a new Xe-LPG GPU architecture. The new architecture significantly improves performance per watt, contributing to power savings
Efficiency at the system level
We are also committed to helping our OEM partners improve energy efficiency at the system level. We offer power savings through tuning, display, and thermal technologies. Some examples include:
- Tuning. Intel® Adaptix™ is a software toolkit that helps OEMs tune systems for improved efficiency and performance. It includes Intel® Energy Performance Optimizer, which helps to optimize energy efficiency automatically based on user behavior.
- Display. displays can use a high percentage of system power. To help lower their energy use, we have developed Intel® Intelligent Display Technology. This technology senses the presence of a user, and if the user leaves their desk, it dims the display light while the user is away, saving power.
- Thermal. Intel® Automated Dust Cleaning Technology reduces dust accumulation on fans in high-performance laptops. This not only improves energy efficiency but it also helps extend the durability of the system components by reducing operating temperatures. Over time, this can lead to lower IT maintenance costs.
Efficiency at the application level
Applications make the AI PC real. To bring these next-generation PCs to life, we are engaging with more than 100 AI independent software vendors (ISVs) for PC platforms and have plans to have more than 300 AI-accelerated applications in the market by the end of 2024. One focus of our work with ISVs such as Zoom and Microsoft is improving the energy efficiency of applications. For example, a 1x1 Zoom call uses 38 percent less power on an Intel Core Ultra processor than a 13th Gen Intel Core processor.[4] Similarly, a Microsoft Teams 3x3 call with AI-enhanced capabilities on an Intel Core Ultra processor uses 13 percent less power than the previous generation processor.[3]
Lower energy consumption and tech-related environmental impact
Intel’s commitment to sustainability has resulted in multiple innovations for AI PCs. These efforts are important as 81% of senior IT leaders say that “green IT” is high on their organization’s corporate agenda.[5] Together with our OEM and ISV partners, we are helping enterprises that adopt AI PCs to lower energy costs and reduce tech-related environmental impact.
Learn more about Intel AI PCs by visiting intel.com/aipc.
------------------------------------------------------------------------------------------------------------
1. Wall Street Journal, “Gen Zs and Millennials Look to Employers to Address Climate Concerns,” July 26, 2023
2. Based on OEM design implementation. 64% more efficient claim is based on ASUS Zenbook 14 Laptop with Intel® Core™ Ultra7 155H, which uses 64% less energy (Typical Energy Consumption – TEC) than total ENERGY STAR computer specification v8.0 allowance. (TEC of model = 13.6 kWh, vs. Total allowance of 38.3 kWh). AI power measurements for government energy regulations are platform holistic with AC power measured at the wall. Platform configurations and implementation have an obvious impact on overall platform power, as they apply to power measurements for these government regulations. This data is relative to one ASUS Zenbook laptop tested and may not be characteristic of all platforms on the market regardless of vendor and/or configurations.
3. Use up to 38 percent less power for AI-enhanced collaboration with an Intel® Core™ Ultra 7 165H processor, compared to a 13th Gen Intel® Core™ i7-1370P processor. As measured by system on chip (SOC) package power during a 1x1 Zoom call with background blur on the NPU. See intel.com/performanceindex for details. Results may vary.
4. Consume up to 13 percent less power while collaborating on a Microsoft Teams 3x3 call with Windows Studio Effects enabled on an Intel® Core™ Ultra 7 165H processor, compared to a 13th Gen Intel® Core™ i7-1370P processor. As measured by system on chip (SoC) package power consumption. See intel.com/performanceindex for details. Results may vary.
5. Intel, The Sustainable CTO Infographic. www.intel.com/content/dam/www/central-libraries/us/en/documents/2023-06/sustainable-cto-exec-summary-infographic.pdf