Artificial intelligence has its uses. While it’s tempting to think of AI as chatbots, image generators, and perhaps something that could spell the end of us all if we’re not careful, big tech companies are busy applying AI and machine learning optimization to many aspects of their business. We’re on a quest to integrate a once-fringe concept into the processes that create products you already own.
I recently had a chance to chat with Gokul Subraniam, President, Intel India and VP and Head, Customer Computing Group Sustainability, in a wide-ranging discussion about Intel’s sustainability goals, product lifecycles and more.
I took the opportunity to ask how Intel uses AI and machine learning models for efficiency in its products and whether AI is used to optimize the process.
“We have AI as a big area of focus in engineering, starting from the front end of our silicon design at the RTL level, all the way to the back end, which is basically post-powering the silicon until it’s ready for production. And then. In our software development and debugging processes, as well as our production use.” We also use artificial intelligence and test how we use artificial intelligence.
“We also use AI when it comes to a lot of the telemetry data that we collect to understand what decisions we can make from a usage standpoint and things like that. So that’s a big area of focus across the lifecycle, from front-end silicon design, to back-end, software development and manufacturing.”
It seems that according to Intel, AI is already involved in numerous aspects of chip design and manufacturing. But given that our discussion revolved around Intel’s goals for sustainability, I also jumped at the chance to ask him about his views on the sustainability of AI’s scalability and the increasing power demands that come with it.
“One of the things Intel believes in is the AI continuum from the cloud data center to the network edge to the PC. What that means is it’s an AI continuum, it’s not a one-size-fits-all, you can’t just have one structure. monolithic view There are a variety of needs across the AI continuum, from large and very large models to small and agile models that can potentially reside in devices at the edge.
“So there’s everything from high-performance computing like Argonne labs with 10s to 1000s of servers to on-premises, smaller companies that only need a few Xeons or maybe Xeons and a few GPUs. And so, “We have this heterogeneous calculation. It allows people to be able to do this in a way that fits their power envelope and calculate the gains.”
While these are noble goals, it has been hard to ignore recent headlines about concerns about the sustainability of an AI future that often involves large amounts of processing power and a potentially surprising amount of real power.
“AI is new, I don’t think anyone has cracked the code,” says Subramaniam. “What does this mean for sustainability? This will be a journey we’ll take and learn as long as we focus on power efficiency and allow our customers to build with the technologies we offer beyond silicon. That’s where things will work.”
While data centers like Argonne National Laboratory and Intel’s Aurora Exascale Supercomputer built on the Purpose-oriented computing types appear as an interesting distinction.
With the rise of consumer chips featuring NPUs (neural processing units) specifically designed for AI processing, first seen in Intel’s Meteor Lake mobile CPUs and expected to appear in yet-to-be-announced Arrow Lake desktop CPUs, this is where the focus is It seems that it is not that much. not only with large amounts of data center AI computing capability, but AI processing as a whole has spread to future chips.
Your next upgrade
Best CPU for gaming: The best chips from Intel and AMD.
best gaming motherboard: Correct boards.
best graphics card: Your perfect pixel pusher is waiting for you.
Best SSD for gaming: Enter the game before others.
Given the rise of the “AI Computing”, it seems that the chip industry is leaning heavily towards an AI future, although there is an argument that any computer with a modern GPU is already equipped for AI processing. However, the current definition of an AI PC seems to range from any PC offering more than 45 TOPs from a dedicated NPU to having the appropriate label on your keyboard.
Regardless, it looks like it’s all turtles for Intel, or in this case AI from the top down. It turns out that your next CPU, or even the one you’re currently using, may have already felt the touch of AI, at least if it’s an Intel unit.