Nvidia is soon to dominate the high-end GPU market, but not because of the standard ‘it’s going to release a GPU’ reason you might think. Nvidia is not planning to release some high-end AMD killer GPUs, instead, AMD is planning to stop targeting the high-end GPU market. AMD Vice President, Jack Huynh, alluded in an interview with Tom’s Hardware, that AMD is seemingly no longer interested in trying to compete with Nvidia at the high end of the scale.
In the interview, Huynh stated “I don’t want AMD to be the company that only people who can afford Porsches and Ferraris can buy. We want to build gaming systems for millions of users.” And when asked about going for the flagship market, the VP says “One day, we may. But my priority right now is to build scale for AMD.”
The rumor that AMD may not release high-end GPUs in the RX 8000 series has been circulating for a while. Now that it seems more likely than not, we’re worried about what Nvidia will do with this eventuality. If there’s no competition at the high-end (we know that the 7900 XTX doesn’t really compete, but it is still an option for gamers who want to save a buck), then Nvidia can pretty much run off and do what it wants at the top end.
If this remains to be true, AMD seems to be going for the middle of the pack, and focusing more on developing the mainstream GPU market and making a better product there. Since AMD can’t really compete at the high end, it makes sense to focus resources on bringing value to the middle of the pack, maybe furthering efficiency or value for money in that bracket. Huynh reflects on this by saying “I want to build the best products at the right system price point. So, think about price point-wise; we’ll have leadership.”
What makes this worse, is we know that Nvidia GPUs already have a bit of a reputation for being very expensive, particularly at the enthusiast level. To put it into perspective, the MSRP of the RTX 4090 was $1,599 compared to the MSRP of the 7900 XTX which was just $999. That’s 37% more than the 7900 XTX and when compared with our 4090’s in-house benchmark performance, the result is not 37% less performance.
Here’s how the current-gen flagships compare
CS_GO 7900 XT vs 4090 testing @ PC Guide
Doom Eternal 7900 XT vs 4090 testing @ PC Guide
3DMark 7900 XT vs 4090 testing @ PC Guide
Shadow of the Tomb Raider 7900 XT vs 4090 testing @ PC Guide
As you can see, the 7900XT certainly falls behind, but does it fall behind enough to warrant an extra $600? If you want the best of the best then there’s absolutely no better than the RTX 4090, but if you’re looking at an RTX 4080, or lower, there’s still the option to jump to a comparable AMD GPU and potentially get a better price. Though it isn’t confirmed which models will get the axe, Nvidia could benefit from this greatly, and clamp down on the top three SKUs for example. Raising the price for entering the “high-performance” arena of GPUs.
It’s this that scares me, the fact that Nvidia could be left unchecked, could have some potentially costly implications for the GPU market. We believe that Intel isn’t really in a position to compete at the high end (that we know of), even though the Arc A770 is fantastic for the price point but it’s not up to the task of taking on the 4090 (or at least the next generation probably won’t be). So we’re not going to be able to rely on Intel to come and offer up some fantastic value alternatives at this level.
I think that Nvidia will keep things as they are for the first uncontested generation but might sneak into a habit of charging more where they can get away with it. Business is business at the end of the day, and if you control that market, the supply dictates the prices. All in all, nobody knows what’s going to happen, this is just our opinion, however, it is something pretty scary to think about. And it’s not even October.
Even Tom’s Hardware said that “Hyunh didn’t say directly that the company wouldn’t have any high-end graphics cards” but it’s looking like there’s little chance for a “5090” killer flagship in the RX 8000 series AMD GPUs.
