Skip to main content

Intel promises new experiences for its upcoming discrete GPUs in 2020

http://s3.amazonaws.com/digitaltrends-uploads-prod/2018/10/samsung-chg90-ultrawide-monitor-review-5481.jpg
Intel

Competition in the graphics segment is about to heat up for Nvidia and AMD, as Intel’s Vice President of the Core and Visual Computing group and General Manager of the Visual Technologies Team Ari Rauch confirmed that the company is on schedule to launch its own discrete graphics solution in 2020. In an interview with Hot Hardware, Rauch claimed that Intel’s latest efforts with its new discrete GPU architecture will be completely different from an earlier Larrabee prototype, which Intel canceled in 2010.

Recommended Videos

“In our discussion with Ari, he specifically stated that the GPU coming in 2020 is a massive enhancement to Intel’s graphics IP, optimized for future graphics, compute, and machine learning workloads,” Hot Hardware reported. “When highly constrained by performance per watt, like Intel’s on-processor graphics are, architectural decisions are made that wouldn’t necessarily apply to discrete solutions.”

In his interview, Rauch claimed that Intel’s GPU solution will deliver unique features to drive new consumer experiences, though he did not give any examples or speculated on what those experiences may look like. “We’re listening to customers and want to deliver the best experiences we can based on what they tell us,” he said. “It’s not just about pricing and framerates — it’s about things like great technology, driver quality, and features. We want to look at every dimension that is important to a user when it comes to graphics.”

Rauch seemed confident that the company’s solution will be scalable, which would allow Intel to create GPU solutions to target different computing demographics. “Last November, we said we’ll bring discrete GPUs to both client and data center segments aiming at delivering the best quality and experiences across the board including gaming, content creation, and enterprise,” Rauch said. “These products will see first availability over a period of time, beginning in 2020. ”

Rauch also said that Intel will also target Linux gamers with its new GPU, though initially at launch, Intel may not be chasing the ultra-premium gaming segment. Intel’s first release may be targeted at the “sub-$300 sweet spot,” Hot Hardware speculated.

By shifting its focus on compute cores to graphics chips, Intel hopes that innovation in the GPU space will help it keep Moore’s law intact. “We believe we’re ready to scale the experience even further with discrete graphics, which will see first availability by 2020,” Rauch said.

Earlier this year, Intel turned 50 years old and the milestone was marked by what many analysts speculated would be the end of Moore’s law, a principle stating that every two years the number of transistors on a chip would double. Since it’s founding, Intel has largely stayed true to Moore’s law, but in recent years, delays with its 10nm design pushed back the cadence of Intel’s release cadence. The company even partnered with rival AMD to deliver a hybrid chip that combines Intel’s compute cores with Radeon graphics cores in the hopes of offsetting some of the demands of Moore’s law.

Chuong Nguyen
Silicon Valley-based technology reporter and Giants baseball fan who splits his time between Northern California and Southern…
Intel quietly opens preorders on new Arrow Lake CPUs
Fingers holding an Intel 285K.

With CES 2025 right around the corner, most of us expect the big announcements to arrive in a week -- but some companies are already teasing new products. In Intel's case, the manufacturer plans to add more CPUs that might compete against some of the best processors. To that end, Intel has now announced preorders for new Arrow Lake CPUs, but most of us can't get our hands on them yet.

As spotted by VideoCardz, Intel China announced that preorders for the Core Ultra 200 non-K CPUs are opening today, with availability planned for January 13. These CPUs will presumably just be non-overclockable versions of existing Arrow Lake chips, such as the Core Ultra 9 285K. In its announcement, Intel teases "new architecture" and "better power consumption."

Read more
Details leak on the upcoming RTX 5070 Ti and RTX 5070 GPUs
The back of the Nvidia RTX 4080 Super graphics card.

As we draw closer to January, leaks and speculation around Nvidia’s next-generation RTX 50-series GPUs are echoing all over the internet. The latest scoop comes from renowned leaker Kopite7kimi providing insights into Nvidia's midrange Blackwell graphics cards, the RTX 5070 Ti and GeForce RTX 5070.

As per the leaks, the RTX 5070 Ti is set to feature the GB203-300-A1 GPU, equipped with 8,960 CUDA cores, offering a substantial increase in processing power compared to its predecessor. The card is rumored to include 16GB of GDDR7 VRAM on a 256-bit memory interface, providing enhanced memory bandwidth for demanding applications such as 4K gaming and video editing.

Read more
I tested Intel’s new XeSS 2 to see if it really holds up against DLSS 3
The Intel logo on the Arc B580 graphics card.

Although it technically arrived alongside the Arc B580, Intel quickly disabled its new XeSS 2 feature shortly after it was introduced. Now, it's back via a new driver update, and with a few fixes to major crashes issues. I took XeSS 2 out for a spin with the Arc B580, which has quickly climbed up the rankings among the best graphics cards, but does XeSS 2 hold up its side of the bargain?

XeSS 2 is Intel's bid to fight back against Nvidia's wildly popular DLSS 3. The upscaling component at the core of XeSS is the same, but XeSS 2 includes both a Reflex-like latency reduction feature and, critically, frame generation. The latency reduction, called XeLL, is enabled by default with frame generation.

Read more