For many of us, technology (or gaming in particular) is an all-out battle between AMD and Nvidia to create the best graphics card. And we, as enthusiastic gamers, argue in forums for our preferred brand. There was rarely a clear winner, but that was part of the magic.
Now, 20 years after this war began in stores and internet photos, AMD decides to withdraw from the battlefield… at least when it comes to high-end gaming graphics cards. Let’s explain it.
AMD has spoken out to say that it now prioritizes AI chips over high-end GPUs for gamers. The company has just unveiled a new business strategy, in which it will merge its RDNA efforts for gaming graphics and CNDA for data centers into a single universal UDNA primarily aimed at AI.
In two interviews granted to Tom’s Hardware, AMD’s Chief Computing and Graphics Officer, Jack Huynh, doesn’t beat around the bush. With gaming graphics, he explains, the goal now is to increase scale and market share at lower prices, not flagship GPUs. It is clear that Nvidia has been dominating the market for years with an iron fist, hence the abandonment.
Pivot towards AI and say goodbye to the most demanding gamer
And yet, being the King of the Hill (as they say in the interview) is exactly what AMD wants to build for the AI data center, because that’s where AMD already has a substantial market share with its EPYC CPUs and believes it can grow by selling the best GPUs for data centers as well.
“Don’t worry, I love games. When I present myself to the board of directors, I say that gaming is a strategic pillar in my strategy,” Huynh told Tom’s Hardware later in the interview.
If this sounds vaguely familiar, it’s probably because AMD is following in the footsteps of its rival Nvidia, in more ways than one.
AMD is becoming an AI chip company, just like Nvidia, if we take a close look at their financial reports, which show that half of the company’s sales are now products for data centers.
Both companies have accelerated their businesses to launch new silicon architectures every year, much faster than before. In the Tom’s Hardware interview about AMD’s UDNA, Huynh now reveals that he also wants to pursue the success of Nvidia’s universal CUDA platform.
As gamers, we like the idea of AMD leading the value-oriented GPUs, also known as dollar per performance. This is because GPUs were becoming excessively expensive and Nvidia is too comfortable with the entire graphics market.
