Intel recently announced the Arc GPU line, which is set to compete with AMD and NVIDIA’s top graphics cards. XeSS, a competitor to DLSS as the best AI upscaling solution for gamers, was also announced by the company. But, in the eyes of the average tech enthusiast, will these creations be enough to put Intel alongside NVIDIA and AMD in graphics debates?
Windows Central enlisted the help of professionals to find out. These were their opinions on the three-way race between Intel, NVIDIA, and AMD, as well as where team blue fits into the equation.
Performing on a crowded stage
Although Intel is behind NVIDIA and AMD in terms of the market Arc and XeSS are attempting to corner, timing isn’t everything. The degree to which Intel’s latest products can be integrated into their respective consumer ecosystems is a major factor in corporate competition.
“To win share in the gaming market Intel will need to work with all the major game providers to have them support its GPU features — something it already does for its integrated GPUs,” Gartner VP Analyst Alan Priestley said. “Intel will also need to deliver as good, if not better, performance on the leading games when compared to the current NVIDIA/AMD GPUs. Based on publicly available information it’s hard to assess how successful Intel will be, but given Raja Koduri’s background it’s reasonable to assume that Intel will deliver a performance-competitive GPU.”
Raja Koduri has a long history with AMD, having spearheaded the company’s Radeon efforts. As a result, he has a track record of producing competitive GPUs and shaping big-picture graphics card plans, which is why Priestley thinks Koduri’s presence will help Intel’s overall chances in this market.
Benefits and drawbacks
However, Priestley pointed out that developers are currently tuning their efforts for NVIDIA and AMD, implying that Intel’s differentiation tactics will take time to earn business. “So initially Intel is likely to provide an alternative to the incumbents (product availability may be an advantage if NVIDIA/AMD remain supply-constrained), and it will take time for it to build momentum behind its feature set versus NVIDIA,” Priestley concluded.
Given that NVIDIA and AMD have both commented on supply shortages affecting their output, his mention of supply is intriguing. If Intel can bridge the gap for consumers, it will have a leg up on the competition in its foundational efforts.
The XeSS factor
“NVIDIA is clearly the dominant platform for GPUs,” Forrester VP, Research Director Glenn O’Donnell said. “What’s particularly challenging is not so much the graphics applications, but the AI applications for tasks like machine learning. THIS is the growth area for GPUs. NVIDIA has the upper hand now, but because so much more potential exists for these kinds of applications, just about anyone has a shot, including Intel.”
He also mentioned another area where Intel should concentrate its efforts if it wants to remain competitive. “Intel needs to demonstrate how its complete family of silicon can fit more seamlessly — easy development of applications will be successful. Also, as we see the so-called xPU (hybrid processors) gain momentum, Intel actually has a good story here. This is precisely why NVIDIA wants Arm.”
To put it another way, Intel has a lot of options besides the well-traveled paths of the competition. Take, for instance, XeSS. Given how early in its lifecycle that technology category is, it could give Intel an advantage. However, we do not yet know how Intel’s solution will fare.
For the time being, all we know is that XeSS will compete against NVIDIA’s AI-driven deep learning super sampling (DLSS) technology, which is currently pitted against AMD’s FidelityFX Super Resolution technology (FSR). Check out our guide on AMD FSR vs. NVIDIA DLSS for a detailed comparison.
Predicaments with a patent
Both experts had ideas when asked how Intel would navigate the sea of patents and established technologies held by AMD and NVIDIA.
“Major chip companies typically have extensive patent cross-licensing agreements so as to avoid any conflicts,” Priestley went on to say that, given Intel’s history with integrated GPUs, the company is likely to have a slew of patents to work with.
O’Donnell took a different approach, emphasizing the uncharted territory that remains to be conquered. “Again, Intel could concede much of the graphics market,” he noted. “That’s not where the action lies. Machine learning and custom applications of silicon are the future market. Intel should focus there and do its own patent sweep there.”
To be continued
It’s too early to speculate on what might happen. Perhaps Intel will make a big splash and become involved in not only the debate over the best CPUs for your PC, but also the debate over the best GPUs. Alternatively, the company’s efforts may fail across the board, and everyone will return to discussing NVIDIA and AMD’s current two-party system.
Will Intel concentrate on machine learning and artificial intelligence? Or will it rely on traditional hardware efforts, perhaps with the goal of filling current market gaps through supply capabilities? Whatever happens, it will undoubtedly be fascinating.