NVIDIA Reportedly Ending the RTX 5090, Reserves GB202 Silicon for Workstation GPUs

erek

8-D
2FA
Joined
Dec 19, 2005
Messages
18,506
See:
  1. 5090 Production Halted
    Nvidia reportedly halts GeForce RTX 5090 production in favor of AI data center and professional...

“RTX 5090 buyers were already having a rough time. The card launched in January 2025 at $1,999, and last month European listings went past €5,000 while Walmart sold out at $4,299. Cards built from GB202 chips that partners already have could keep trickling out for a while, though it's reasonable to expect prices to climb further once that stock dries up. Of course, the RTX 5090 was hardly a gaming card for most people by this point, with AI firms reportedly picking up stock in bulk before it ever reached gaming retail. With the RTX 60-series reportedly slipping to 2028, the RTX 5080 24 GB may stay the fastest GeForce card in production for a long time.
Sources: MEGAsizeGPU on X, hongxing2020 on X, via Videocardz”

Source: https://www.techpowerup.com/353578/...0-reserves-gb202-silicon-for-workstation-gpus
 
Without any competition I can see why. There is nothing that AMD has that can beat the 5080 anyway so even with the 5090 discontinued, Nvidia would still be #1. Besides, all 5090's right now are just being funneled towards AI anyways and not gamers so that's another reason for them to not bother.
 
A RTX 5080 24GB release would be interesting, but it's gonna be pricey.
 
This doesn't bode well for the next gen cards. I can see them dropping the 90 series altogether now to allocate those GPUs to the Pro/AI data center class GPUs since that's where most 5090s are going now anyways and Nvidia can get significantly more money for them in that market.

So I can totally see Jenson pulling a "6070 is 2x faster than 5090" type situation again next gen to justify dropping the 6090 while still charging 2k for the next 80-class GPU. 😑
 
This doesn't bode well for the next gen cards. I can see them dropping the 90 series altogether now to allocate those GPUs to the Pro/AI data center class GPUs since that's where most 5090s are going now anyways and Nvidia can get significantly more money for them in that market.

So I can totally see Jenson pulling a "6070 is 2x faster than 5090" type situation again next gen to justify dropping the 6090 while still charging 2k for the next 80-class GPU. 😑
"The more you buy, the more you save. Do you like my jacket? AI, AI, AI..."
 
This doesn't bode well for the next gen cards. I can see them dropping the 90 series altogether now to allocate those GPUs to the Pro/AI data center class GPUs since that's where most 5090s are going now anyways and Nvidia can get significantly more money for them in that market.

So I can totally see Jenson pulling a "6070 is 2x faster than 5090" type situation again next gen to justify dropping the 6090 while still charging 2k for the next 80-class GPU. 😑

They might as well drop the 6090 from the lineup since AI firms would gobble them up regardless and the street price would be over $10k anyways.
 
It's really a shame the 5080 is such a downgrade even at 24GB. Being a lifelong PC gaming enthusiast I despise what this AI mania has done to our hobby. At the same time I can't wish for a crash... the overall economic implications would be disastrous.

"May you live in interesting times."
 
Nvidia is going to pull a AMD with the 6000 serise cards. No halo nor even a high end cards. 6070 with 12gb will be the highest end card.
 
Without any competition I can see why. There is nothing that AMD has that can beat the 5080 anyway so even with the 5090 discontinued, Nvidia would still be #1. Besides, all 5090's right now are just being funneled towards AI anyways and not gamers so that's another reason for them to not bother.
It is worse, AMD has nothing that can beat a RTX 4090 either, the RTX 4090 is getting close to GTX 1080 Ti status.
 
Nvidia and AMD with lucrative SI income could just decide to drop gaming cards, gamers are just whiners type thought. Not sure how viable for Nvidia and AMD to build up a community, hardware for gamers and then crap on them. What is Sony going to do when memory is 6x plus, wait inline for APUs, limited on what node to use? Microsoft? Probably just go SI.

Without new hardware being pumped out, developers will be next to go for games, wait, that is already happening.
 
Nvidia and AMD with lucrative SI income could just decide to drop gaming cards, gamers are just whiners type thought. Not sure how viable for Nvidia and AMD to build up a community, hardware for gamers and then crap on them. What is Sony going to do when memory is 6x plus, wait inline for APUs, limited on what node to use? Microsoft? Probably just go SI.

Without new hardware being pumped out, developers will be next to go for games, wait, that is already happening.
They can and Nvidia will I believe.
 
This timeline sucks.
I hope whenever we do get RTX 60 series that they focus on efficiency and getting power usage (TDP) down.
I'm still on an RTX 3090 and I say that the RTX 5080 really should have been 24gb at launch. But at this point Blackwell is coming up on 2 years old. I'm not paying these crazy prices this late into the generation. I'm holding for next gen, whenever that is at this point.
 
I think that the same leaker that said there will be floads of RTX 5090 and that scalper would be left with theirs because Blackwell GB200 and RTX 6000 were a failure that could not be sold and Nvidia had to do stuff with their pre-bought capacity...

https://x.com/Zed__Wang/status/1890608126329586017

So... could be taken with some grain of salt
 
This timeline sucks.
I hope whenever we do get RTX 60 series that they focus on efficiency and getting power usage (TDP) down.
I'm still on an RTX 3090 and I say that the RTX 5080 really should have been 24gb at launch. But at this point Blackwell is coming up on 2 years old. I'm not paying these crazy prices this late into the generation. I'm holding for next gen, whenever that is at this point.
I want more performance, not "greener" GPU's...
 
TSMC-Nvidia has been obsessed with efficacy and power usage (from the mobile and datacenter world that are a lot about that and way bigger client-focus than desktop), a 5080 is about twice the 3080 in efficacy and match/beat in many titles even a 9070 non XT despite the lesser node used.

What would be a big surprise from Nvida 6000 series, too big the first generation since ampere that not focused on price cut-efficacy and power usage (5090 aside of course, which did go for performance). Maybe the 7000 series if TSMC node family had time to cut more clearly between high power (Nvidia compute gpus) and mobile (Apple) that it would be an option.
 
Back
Top