AMD Radeon RX 9070 XT outsells entire combined NVIDIA RTX 50 series GPUs at German retailer
-
Well good thing the 9070 series is the series that is trying to adopt a more Nvidia-like naming scheme.
Yeah it has two names. But slightly less confusing
-
Consistency? Nvidia does the exact same thing, so yeah I guess that’s consistent. That reliable numeric with the Nvidia GeForce Titan, Nvidia Titan, and Nvidia Titan RTX (yes those are all very different cards).
And don’t even get me started on the workstation Quadro card naming, with the Quadro RTX 4000, RTX A4000, RTX 4000 (Ada generation), and RTX PRO 4000 Blackwell.
Then you have the server cards. L4, L40, A100, H100 then rtx 6000 Blackwell (workstation and server editions are same name but different cards)
It goes on
-
I thought it was confusing because this is the 3rd changeup to their naming in recent years.
RX 580 to RX Vega 56 and 64 to RX 5700 to 9070. Yes it’s still the intuitive “bigger number better” and “first number is generation”, but I can see how people might be frustrated with it whereas nvidia has consistency.
Don’t forget the Radeon VII (announced 2019) that didn’t fit in any scheme.
-
How is it confusing?
Well minus the part where theres 8GB versions with no name differences…
It’s confusing because they constantly change their numbering scheme. Nvidia has stayed consistent in the last 17 years
-
This post did not contain any content.
Right now on Amazon you can get an RTX 5070 for cheaper than an RX 9070 XT. The obvious choice is still the RTX 5070.
RTX 5070 ~$550 RX 9070 XT ~$630
-
Yeah it has two names. But slightly less confusing
The 9070 had two names? You’re talking about the regular, the XT, and GRE?
-
I am not retaking the GRE to use that card.
-
I am not retaking the GRE to use that card.
I’m wooshing. Is that some kind of test where you live?
-
This post did not contain any content.
NVIDIA made like 4 billion on gaming GPUs and like 58 billion on AI GPUs last quarter. I’m sure 80% of those gaming GPUs were used for AI as well. To them it’s likely irrelevant how they’re doing in the gaming market.
-
Yeah it’s the exam you take for grad school in the US for some colleges. I think it is the “graduate readiness exam” but it’s been a few years and I cant spare 10 seconds to look lol.
-
This post did not contain any content.
I saw a Black Friday sale for a 5080 for around €1k, a decent OC model too. I have no interest in supporting NVIDIA or their fire hazards with the way they’ve been acting.
-
It’s confusing because they constantly change their numbering scheme. Nvidia has stayed consistent in the last 17 years
Well yea… But nvidia didn’t change their naming scheme because they use it to mislead people.
A 5080 today is equivalent to what a 1060 used to be when it comes to raw sillicon if I remember correctly.
-
I saw a Black Friday sale for a 5080 for around €1k, a decent OC model too. I have no interest in supporting NVIDIA or their fire hazards with the way they’ve been acting.
It’s an obscene price to pay for a GPU no matter how fast it is.
-
I thought it was confusing because this is the 3rd changeup to their naming in recent years.
RX 580 to RX Vega 56 and 64 to RX 5700 to 9070. Yes it’s still the intuitive “bigger number better” and “first number is generation”, but I can see how people might be frustrated with it whereas nvidia has consistency.
I think they had to make that change because there was a Radeon 9700 back in the ATI days or something. I just wish they hadn’t done the RX thing. Because those charts where people compare GPU performance will have legends like:
RX-7600 RTX-5060 RX-9060XT TXT-6060RXT TVX-5040T RTX-4060
and you say "There’s no consistency with generation or manufacturer and I’m pretty sure one of those is the part number of a cylinder head for a Toyota Tacoma.
-
Is this just this week’s exception or is this representative of the retail market?
A couple generations back, Nvidia was the obvious choice and AMD just couldn’t compete. Nvidia had real-time ray tracing, AMD didn’t, Nvidia’s hardware video encoding was great, AMD’s sucked, Nvidia had CUDA, AMD pretty much didn’t, Nvidia had their DLSS frame gen technology, AMD either didn’t or it wasn’t very good.
Well, in most of those places, AMD has caught up, and they offer more VRAM in their lower tier products, at better prices.
Oh, and I’ve been hearing through the grapevine that Nvidia is dropping the ball with drivers. That used to be AMD’s bag, but AMD’s drivers are more solid these days.
Oh, and Nvidia’s weird new power socket keeps catching on fire. I’ve got a 7900GRE that attaches with two good old 8-pin PCIe connectors that offer a distinct lack of combustion.
-
The 9070XT is a good card for the price, and that doesn’t tell us a lot, there’s going to be markets where AMD does better and some where Nvidia does better, as an AMD user I feel like they still fall behind with their drivers, I’ve had many Driver timeouts with my RX6700XT just because I was hitting alt tab on CS2, I’ve never had it on Nvidia.
Also someone who does a lot of upscaling with Topaz I would have loved to see AMD making some serious grounds there there’s no reason why they should be that behind with content creators, I don’t think we’ll ever see the GPU department catching up as the CPU did with Ryzen but I feel like everytime Nvidia fucks something up, AMD does the same so It’s a game of who’s going to fuck up more.
I bought the 6700xt just because I was contemplating on the 3070 but I saw it was an 8GB VRAM card and I thought yeah miss me with that shit.
Those driver issues are so disheartening, especially when you’re just trying to enjoy a game! It’s good to see AMD gaining ground though, even if it’s a slow climb.
-
This post did not contain any content.
The 9070xt is a great card and much better value than the equivalent Nvidia offering, last time I checked the benchmarks. The Linux compatibility was the other major draw. My system is now 100% AMD and I’m preparing to trial Steam os and Bazzite.
-
Well yea… But nvidia didn’t change their naming scheme because they use it to mislead people.
A 5080 today is equivalent to what a 1060 used to be when it comes to raw sillicon if I remember correctly.
Wym with raw silicone and how is it important?
-
Wym with raw silicone and how is it important?
I mean that the 5080 size, I believe, is about 55% that of the 5090.
Whereas, when the 10 series launched, the 1060 die size was also at around 55%.
The percentaches are definitely not correct but you uget the idea. The non-90 dies of today are MUCH more cut down that the old ones.
