With the launch of Nvidia’s RTX 4080 GPUs, plural, on the horizon, the corporate has taken an unprecedented step: It’s “unlaunching” the RTX 4080 12GB. As you would possibly recall, this GPU was broadly mocked for being a RTX 4080 whereas providing specs just like earlier x70 GPUs. That features utilizing a smaller AD104 die, lots fewer CUDA cores, and a slender 192-bit reminiscence bus. No one was pleased with this case, as Nvidia has by no means supplied two GPUs with the identical title and totally different specs earlier than. Why it selected to take action this time might be a thriller for the ages. Regardless, it has now responded by eradicating the GPU from its secure. It’s unclear if it will likely be rebadged as an RTX 4070, or simply disappear completely. As of press time, the GPU has been deleted from Nvidia’s web site.
Nvidia introduced the abrupt u-turn in a weblog publish. We double-checked to verify it wasn’t a parody web site, because it’s shocking Nvidia listened to its prospects. The corporate holds the lion’s share of the discrete GPU market, so it may well just about do no matter it desires. This time, nonetheless, it appears to have gone too far.
In its publish, the corporate states the plain: “The RTX 4080 12GB is a unbelievable graphics card, however it’s not named proper. Having two GPUs with the 4080 designation is complicated.” Um, you’re simply now figuring this out, Nvidia? We knew it was complicated on launch day and mentioned as a lot. The remainder of the weblog simply reiterates the 16GB model continues to be coming. Nvidia has additionally given it a launch date: Nov. 16. The corporate states that if the lengthy strains at Microcenters for the 4090 launch are any indication, the 16GB model will undoubtedly promote out, too. It’s in all probability proper about that. RTX 4090s are already littering eBay with $1,000 added to the value. Plainly although mining is useless, scalping continues to be alive and properly.
As you possibly can see above, the 2 GPUs are radically totally different. In all earlier generations, the 12GB card could be labeled as an RTX 4070. Nvidia hasn’t begun speaking about that exact GPU but, so it’s nonetheless potential it might flip into one. Nevertheless, it’s additionally probably it will likely be scuttled completely. We hate to beat a useless horse, however it’s stuff like this that in all probability helped EVGA determine it was time to give up making Nvidia GPUs. It looks as if companions will need to have already made a few of these GPUs, because the launch was simply weeks away. Additionally, the 12GB model was a partner-only card, with Nvidia abstaining from making FE variations. This in all probability made the choice simpler because it had no pores and skin within the recreation.
The timing of the announcement makes us suppose Nvidia is studying to the 4080 benchmarks it posted on its web site. Each web site that covers these items leads with a headline about how a lot slower the 12GB model is than the 16GB mannequin. This simply strengthened the concept in everybody’s minds that the 12GB mannequin was a money seize. Would any gamer actually pay $899 for a GPU with a 192-bit reminiscence bus? You would possibly suppose that’s spec minutia that no person pays consideration to, however an individual spending that a lot cash on a GPU would know that. Nvidia’s announcement comes on the heels of a web-based uproar over these benchmarks, making us suppose the 2 occasions are linked.
There’s additionally another excuse for Nvidia to backpedal. As a substitute of simply doing the best factor to scale back client confusion, the corporate might be simply making an attempt to keep away from launching a dud. AMD is about to unveil its RDNA3 GPUs on Nov. 3. In response to Igor’s Lab, the corporate won’t be able to the touch the RTX 4090. Nevertheless, its flagship GPU would possibly find yourself proper in the midst of each Nvidia RTX 4080s. Recall the 16GB model is $1,199 and the 12GB mannequin was $899. Now think about AMD placing a card in between these two that performs nearer to the high-end, at a worth of $700 or so. That may make the 12GB card DOA instantly.
No matter its motivations, Nvidia has achieved the best factor right here. Launching a $900 card with x70 specs had led to accusations of Nvidia being grasping. That was an accusation hurled on the firm through the pandemic too, because it did issues like relaunching the RTX 3080 with a $500 worth premium, together with all the assorted Ti card launches. We all know that corporations have to make a revenue, however providing “new” GPUs with a 5 to 10 % efficiency enhance over an present card appears pointless. Nvidia even bragged to buyers in that interval about how its prospects had been paying $300 greater than traditional to improve. Again then, it offered each GPU it made, so it tried to make as many playing cards because it might. This helped result in the present oversupply of GPUs.
For now, we must wait and see what occurs with the RTX 4070. Nvidia will in all probability simply sit again and see what AMD declares on Nov. 3 after which alter accordingly. It’s already left massive gaps between its 40-series GPUs. There’s a large delta between the RTX 4090 and 4080, so a 4080 Ti might be within the playing cards. We already know a 4090 Ti will come in some unspecified time in the future, too. The place AMD’s chiplet-based GPUs will match into this matrix stays to be seen.