Saturday, March 1st 2025
Early Leak Claims AMD Radeon RX 9060 XT Might Reach NVIDIA GeForce RTX 4070 Territory
Judging by the current state of gaming GPUs, it might appear to some that true budget-class cards are a thing of the past. That said, it is almost certain that both NVIDIA and AMD are cooking entry-level GPUs to cater to folks who can't shell out the astoundingly high prices that modern mid-range and high-end GPUs command, with AMD having already confirmed the launch for RX 9060 class cards sometime in Q2 of this year. Previous leaks have indicated that the RX 9060 will likely hit the scene with 12 GB of GDDR6 VRAM, whereas its XT sibling will boast an additional 4 GB. NVIDIA is also expected to drop the RTX 5060 and 5060 Ti cards sometime towards the end of this month, likely in 8 GB and 16 GB flavors of the shinier GDDR7 spec.
Now, a fresh leak by Moore's Law Is Dead (MLID) has claimed that the Radeon RX 9060 XT will outperform the RTX 4060 Ti in performance, slotting in between the RTX 4060 Ti and the Radeon RX 7700 XT. Moreover, he added that AMD may even push clocks to bring the card closer to the RTX 4070 territory - a sweet position to hold indeed. Regarding launch date, MLID expects the card to hit the arena sometime in April. Of course, as with all leaks and rumors, accept this information with a grain of salt, especially considering that MLID's assertions are sourced from a single party. The RTX 5060/Ti is expected to be priced in the $400-$500 range, which means the RX 9060 XT will likely have to be priced in the lower-end of that in order to make for a compelling value proposition.
Sources:
Moore's Law is Dead, Spotted by Notebookcheck
Now, a fresh leak by Moore's Law Is Dead (MLID) has claimed that the Radeon RX 9060 XT will outperform the RTX 4060 Ti in performance, slotting in between the RTX 4060 Ti and the Radeon RX 7700 XT. Moreover, he added that AMD may even push clocks to bring the card closer to the RTX 4070 territory - a sweet position to hold indeed. Regarding launch date, MLID expects the card to hit the arena sometime in April. Of course, as with all leaks and rumors, accept this information with a grain of salt, especially considering that MLID's assertions are sourced from a single party. The RTX 5060/Ti is expected to be priced in the $400-$500 range, which means the RX 9060 XT will likely have to be priced in the lower-end of that in order to make for a compelling value proposition.


78 Comments on Early Leak Claims AMD Radeon RX 9060 XT Might Reach NVIDIA GeForce RTX 4070 Territory
All the tech companies right now are constrained by the fact that most of the products, even the ones that are direct competitors, are being made by a single maker - TSMC, who can't just simply expand his production volumes. And common consumer products from those tech companies are also not their biggest revenue bringers - Nvidia makes most of their money by selling server AI equipment, AMD is selling server CPUs.
So I don't see any possibility of AMD suddenly focusing on consumer GPUs, regaining market share - but actually lowering their income by not making as many server CPUs. I think not having presentation of new GPU lineup on CES was telling enough on that they aren't a priority, even more so than not aiming at high end GPUs at all.
Top line seems about right.
9070 may equal the 7900XT give or take but I suspect the 5070 does not even match the 4070S let alone the 4070Ti.
The 9060XT has too little bandwidth to match the 7800XT so I see it closer to the 7700XT / 4070
With how small N44 probably is I don't see the point in a cut part at launch. Just release the XT with 16GB of VRAM for $330 and call it done.
If they do get some faulty dies save them and then release a 9050XT with a 96bit bus and 12GB of VRAM for ~$250 when there is enough of a stockpile for it to last. I really think 96bit with 12GB will be a far better 1080p part than an 8GB 128bit part would be.
That is the stack I would have if I were AMD.
I've picked an Nvidia card countless times for builds that are thermally-constrained because the Nvidia equivalent GPU is pulling 30-40% less power for the same end result. It's why I begrudgingly put up with the 4060Ti in one of my machines, and I can't even remember the last time I put a Radeon in a mini ITX build...
www.techpowerup.com/forums/threads/nvidia-geforce-rtx-5070-founders-edition.333502/post-5466185
It's pretty terrifying if those numbers are even vaguely accurate for the "330W" 9070XT
With the exception of that last bit, your statement edges on the blatantly ignorant and ridiculous. Every time a new advancement comes along that pushes the boundary's, someone chimes in and said it'll never happen and then it invariably does. They said that with 1440p, they said it with 4k and now here we are with someone saying it about 8k.
Are we sensing a pattern yet?
Now before you respond with some smarmy egocentric diatribe, sit back, think back to the past and just accept that your current thought on this matter does not meet with historical trends or merit. Everyone who said that 1080p, 1440p and 2160p would never be a thing were wrong, because here we are.
The future is coming, it needs only time and the work of technological engineers.
It's all about budget constraints, about resources, about time and manpower. Nobody wants to waste any of that on something that offers no value.
2) You're talking about the consumer point of view again. YouTube wanted to lock 4K behind Premium, you think they'll want to pay for 8K in every video? People are watching podcasts in 4K, which is a complete waste of bandwidth. They're watching videos with graphs, where 4K offers no benefit whatsoever. It costs them crazy amounts of money, and allowing 8K would raise those costs even more, with no financial benefit to them.
And what about streaming services like Netflix? They all keep raising prices, introducing ads, and 4K quality is already constrained. How much would they have to charge for 8K, where you wouldn't even see a difference because of the reasons I stated in point 1.
People are getting sick of streaming services. It might all just completely crash if they keep doing what they're doing.
3) With a bump in resolution has to come a bump in screen size. That's actually the definition of resolution, it's a combination of pixel density, screen size and viewing distance. Calling 4K a "resolution" is just a marketing term that's been normalized in everyday use. Your eyes need to actually be able to resolve the detail. You say it's an opinion, but there are plenty of studies that show optimal viewing distance for a given screen size and "resolution".
And there's a limit to what screen size people want. Most people don't want a screen bigger than 32" on their desk. Many say that even 32" is too big for a monitor.
I really wish the obsession with resolution stopped. On one hand you get those crying about upscaling and saying "native or bust", even when they're presented with direct comparisons showing little to no difference. On the other hand you have you have console gamers playing titles that upscale from 720p and they think they're getting 4K because that's the signal being output from the console.
I guess you're right, though. People will just gobble up whatever is thrown at them. They'll buy shiny new 8K monitors that offer no benefit, but they'll be happy because bigger number better.
The only thing is that for that to happen, there needs to be a semiconductor revolution. The current advancement of silicon manufacturing will not be enough to make 8K viable. Not from the consumer side, but from the content provider side.
*With FSR/DLSS ultra performance + FG
So now with AMD having a slight edge with RX 9070, and a good chance of well positioned RX 9060 / 9060 XT, they should above all flood the market with as many as possible, which will also prevent stores from overcharging.Having more VRAM doesn't give you more performance, it only makes you capable of displaying higher details provided you have the bandwidth and computational performance to go along with it. A theoretical "9050 XT" would be way too slow to make any real use for 12 GB during gaming, and would probably approach ~10 FPS in loads that actually requires it. Meanwhile, having faster and slightly more expensive VRAM would probably yield an extra ~3-5% performance across the board, which is way more valuable to the customer than extra VRAM which would only come into play when running unrealistic uses cases no buyer would ever do.