As Intel prepares for the launch of its ARC GPU range on the market and information on some products has started to come out, we all noticed that the company’s executives have made it their mission to occupy the field on the subject. For example, we noticed that Raja Koduri had taken a liking to social networks, regularly leaving us little messages about the gestation of his GPUs.
Raja and Roger give more details about the ARC GPUs.
But recently, the team in charge of the GPUs was able to give more details with some instructive elements during an interview. The topics of partners for the manufacture of future cards, the possibility of “reference” products or the availability at launch were mentioned. Here is the transcript of the important elements of this exchange. How does Intel intend to work with the graphics card ecosystem? Will it work with partner brands? Roger Chandler: Intel has a long history of working with the Taiwanese ODM community as well as OEMs around the world. So at the platform level, we’re extending those relationships and changing them because the situation is different with consumer graphics cards. But we’ve actually found that many of the relationships and processes are extremely applicable to this product line. That’s also one of the reasons why we created DG1, our previous graphics product launched last year, so that we could really make sure that all of these pipelines and connections and relationships are really strong. There are ODMs and OEMs, but I also think it’s important to tap into the ISV ecosystem, and Intel has decades of experience working with ISVs and game developers. Will Intel make reference boards or will board partners only have custom models to offer? Roger Chandler: We are providing the designs necessary to eliminate friction so that our partners can truly accelerate and differentiate themselves from our products. Now, whether or not it’s comparable to what’s on the market today, we’re not going to get into those details. (So that’s largely a non-issue.) Is Intel confident that it will have enough Arc GPUs at launch? Raja Koduri: The response to our Architecture Day announcements from our OEM customers and end users has been phenomenal. Everyone is welcoming us, so this is great news. We will be stepping in as a third player. I will always be very careful, when demand is so high and when the market is so tough. I can always use more supply. So I’m not going to say I have enough supply in this high-demand market. I think every one of my competitors will say the same thing right now.( second big wild card ) Will Intel consider making its GPUs in-house rather than at third-party foundries? Raja Koduri: Depending on the design points of the product, [we look at] what is the best technology that we have access to both internally and externally. We also look at capacity. We need to optimize, especially for GPUs, which is a new line for Intel, [so] we always have capacity, whether it’s internal or external. If you look at our first-generation graphics card, the Xe Max , it was produced on Intel SuperFin technology. We’ll look at the roadmap that was released at Intel Accelerated, all the way back to the Angstrom era. I’m looking at that roadmap with great interest, to see how I can leverage it.
What is the status of XeSS support in games? How many games will be supported on launch day? ( XeSS is the in-house oversampling technology that is meant to compete with DLSS) Roger Chandler: There has been a lot of excitement about XeSS, and when we talked about it at Architecture Day, we actually demonstrated its capabilities. We are actively working with dozens and dozens of studios right now. That’s only going to grow as well, because what really motivates developers is the fact that they understand how many platforms support this feature. There’s really more excitement about our open approach than almost any other feature in the ecosystem right now. You should expect a very healthy collection of games to support this, and that will grow aggressively over time. The “enthusiast” GPUs are now pushing into the 350-400W range. Will Alchemist consume as much power as possible to maximize performance, or is power efficiency more important? Raja Koduri: Power efficiency always matters. Always. We’re focusing on that, optimizing that, and as far as power envelopes go, we’re going to work with partners and we’re going to enable them with various tools. You know our basic architecture will be focused on different frequencies, different voltages, different choices. We’re leaving a lot of choices with that. But, my short answer is that efficiency still matters a lot. These excerpts represent a summary of the interview delivered by the Intel team. Some answers have been edited for length and translation. Other topics discussed have been given a brief like for clamping on cryptos.
