Tuesday, January 7th 2025
GIGABYTE Shows Off Custom Radeon RX 9070 and 9070 XT GPUs at CES 2025
At CES 2025, GIGABYTE unveiled their latest Radeon RX 9070 series GPUs, showcasing the AORUS Radeon RX 9070 XT ELITE and the Radeon RX 9070 GAMING OC models. Powered by AMD's new RDNA 4 architecture, GIGABYTE has prepared a massive triple-fan cooling setup for both cards. The AORUS RX 9070 XT ELITE stands out with its WINDFORCE cooling system, featuring a newly designed Hawk fan and a vapor chamber with composite copper heat pipes for enhanced thermal management. The card also glows with an RGB Halo lighting system, a reinforced structure with an RGB-illuminated metal backplate, and dual BIOS modes for performance and silent operation. Meanwhile, the RX 9070 GAMING OC offers similar cutting-edge features in a sleek package, tailored for gamers who demand a balance of performance and style.
These new GIGABYTE models, with features such as dual BIOS for overclocking, help AMD's RX 9070 series deliver performance comparable to the RX 7900 XT or even higher in rasterization while introducing innovations like 2nd gen AI accelerators, 3rd gen ray tracing accelerators, and the FidelityFX Super Resolution 4 (FSR 4) with frame generation and anti-lag technology. With these GPUs set to launch in Q1 2025, we will learn more about the pricing structure in the coming days.
These new GIGABYTE models, with features such as dual BIOS for overclocking, help AMD's RX 9070 series deliver performance comparable to the RX 7900 XT or even higher in rasterization while introducing innovations like 2nd gen AI accelerators, 3rd gen ray tracing accelerators, and the FidelityFX Super Resolution 4 (FSR 4) with frame generation and anti-lag technology. With these GPUs set to launch in Q1 2025, we will learn more about the pricing structure in the coming days.
18 Comments on GIGABYTE Shows Off Custom Radeon RX 9070 and 9070 XT GPUs at CES 2025
Genuinely hope none of these sell.
The way I see it, Nvidia copied ATI naming conventions who have now changed it to suit. That's ignoring both of these companies odd-ball named products such as the 'Titan X, Xp, V, CEO/w@nker edition' or 'Fury / Fury X', etc.
All Nvidia had back in the day was Geforce(256), Geforce 2 > 3. So imaginative... from the company who decided the follow up to the TNT would be 'TNT2'... real friday night branding work there (I kid... I couldn't care less if they just carried on from 1>∞).
ATI changed the branding when they renamed the Radeons to the 7000/7500, then introduced the 8000 range within this same time period.
Nvidia then sort of copied for the Geforce 4 onwards which, post 'tesla' cores, they recategorised. I don't remember people acting like chimps and throwing crap around back then because of it.
Unfortunately this Radeon '9000' series isn't going to be like 20+ years ago where the 9500/9700 destroyed the competition*.
* technically the Geforce FX was a more 'capable' product but ATI played to the DX9 requirements perfectly in terms of maximum performance for the money.
RDNA1: 5000
RDNA2: 6000
RDNA3: 7000
The reason people hate it is because everyone knows there nowhere to go after 9000 series.
They won't call UDNA1 10000 series and since UDNA1 brings a name change anyway it makes RDNA4 name change all the more meaningless.
Not the best way to transition to new series.
Also now Gigabyte too with 3x8pin. Totally meaningless on a 330W (reported AIB, 260W reference) card.
And before anyone says OC headroom - no. Modern cards dont have meaningful any OC headroom that would warrant the extra connectors.
It's almost as if AIB's are trying to prove a point with 16pin by unnecessarily equipping AMD's midrange'ish card with 3x8pin.
Or they went from 9000 GT(X) to GT200, seemingly ignoring starting at GT100?
Or AMD bypassed the HD 8000 (and 9000) series going from HD 7000 to the R3/5/7 200
And like i said - UDNA1 will necessitate name change regardless. Thus AMD will have two name changes in two years which looks even worse.
And dont get me started on their mobile CPU naming scheme that required a dedicated decoder ring for tech press to make head and tails of the lineup - one they now also abandoned for AI everything. IMHO just sends a message that AMD cards are power hungry and using outdated, bulky power connectors.
Makes them look bad.
Apart from OEM only options (in the 100/300), where were the GT 800 options? Why even create a hole in the first place...
And when they didn't create a hole, you had issues like the GT 730, which literally had a chip from every family in the DX11 range Nvidia made...
I don't really care about the branding - they can do what they want with the one main exception (which I think everyone would agree with).... do not use it to screw over customers.
By that logic AMD using APUs filled the HD8000/9000 hole using old GPU logic from their pre-GCN silicon, because it existed purely as a named / branded product.
Do you not see the irony of criticising branding that skips a sequence number whilst giving a pass to branding that fills in a skipped sequence with products from the existing (or even worse, older than the last) generation which then, in some cases, people have to warn others from buying or explain a scenario why some products in the same series do not support the same functions as others...
Yeah, going from the RX 7000 to 9000 is kinda dumb / pointless. Maybe it's a plan to release product branding on CPU and GPU in step going forwards - if the release schedule of Ryzen stays the same that's not an impossible assumption seeing as GPU releases have slowed down compared to a decade or so ago. Have your Ryzen 9000 and RX 9000 GPUs, etc.