Monday, June 9th 2025
NVIDIA GeForce RTX 5050 Uses Slower GDDR6 Memory, Based on GB207 Silicon
NVIDIA is preparing to launch a new lower mid-range graphics card SKU in July, the GeForce RTX 5050. Positioned below the RTX 5060, the RTX 5050 possibly targets a price-point under the $250 mark, looking for a slice of the pie commanded by the Intel Arc B580. We are now learning that NVIDIA is making design choices that enable it to sell this card with an aggressive price, specifically, the choice of older generation GDDR6 memory. The card will likely feature 8 GB of GDDR6 memory across a 128-bit memory interface. At this point, we don't know the memory speeds, but if we were to hazard a guess, it could be 18 Gbps, for 288 GB/s of memory bandwidth.
The RTX 5050 is also expected to debut and max out the new "GB207" silicon, the smallest chip based on the GeForce Blackwell graphics architecture. This chip is expected to come with 20 SM, for 2,560 CUDA cores, 80 Tensor cores, 20 RT cores, 80 TMUs, and an unknown number of ROPs. The RTX 5050 is expected to be given a total graphics power (TGP) value of 130 W. It will be possible to build cards with 6-pin PCIe power connectors (75 W from connector, 75 W from the PCIe slot), although we expect single 8-pin PCIe to be the standard. The 130 W TGP will make it possible to build low-profile or compact, ITX-friendly cards.
Source:
VideoCardz
The RTX 5050 is also expected to debut and max out the new "GB207" silicon, the smallest chip based on the GeForce Blackwell graphics architecture. This chip is expected to come with 20 SM, for 2,560 CUDA cores, 80 Tensor cores, 20 RT cores, 80 TMUs, and an unknown number of ROPs. The RTX 5050 is expected to be given a total graphics power (TGP) value of 130 W. It will be possible to build cards with 6-pin PCIe power connectors (75 W from connector, 75 W from the PCIe slot), although we expect single 8-pin PCIe to be the standard. The 130 W TGP will make it possible to build low-profile or compact, ITX-friendly cards.

76 Comments on NVIDIA GeForce RTX 5050 Uses Slower GDDR6 Memory, Based on GB207 Silicon
Technically, the XX50 was entry level in the past, but spec-wise this should have been the GT 5030.
It also should be a $125-149 product, but with performance being down-tiered and prices up-tiered with 50-gen once more, this will be one expensive “add-in dedicated graphics processor” for $199 +
with a power connector to boot…
Bravo!
Would that be something under the cost of a low entry cpu + say ram and board costs?
Something with at least 3x the performance of the best igpu?
or just the lowest tier dgpu card available regardless of performance or price??
This is the only way 8GB makes sense. Also the lack of VRAM bandwidth and cut down core configuration means that 1080p is all this will be good for, making 8GB even less of an issue.
My beef with the 9060/5060/5060Ti 8GB is that they are fast enough for 1440p and more demanding AAA games, without the VRAM to back it up. At least this time the VRAM quantity is proportionate to the price and performance.
Even On DAY 1 Release date, every Reviewer suggest Avoid the 3GB version its trash and always been trash, Misleading series carry 1060 name but totally different card it absolete only 1 yeaes later
RX570 8GB at the time selling only for 110$ and Only require 110 watt when i gaming, and still can play game 8 years laters
Even Better RX580 selling only 130$, same performance as 6GB version of 1060 and still perform 8 years later when 1060 even 6GB version now is useless run out Vram to play game and need mod the game to run on 6GB
Let alone 3GB version become Ewaste
It was obsolete a year later, so go ahead and tell me what games it couldn't play in 2017. Im all ears.
www.techpowerup.com/review/msi-gtx-1060-gaming-x-3-gb/31.html
I still personally feel that between the 1050Ti and 1060 6GB it was a good value for what you got. AMD had more interesting prospects in terms of value but again, what NVIDIA offered was not anywhere near as bad as what the person you quoted tried implying.
I dont personally consider anything thats hovering around 180 Watts or below to be concerning enough to wanna be worried about power in terms of price.. though that obviously depends on a ton of other factors, such as your PSU / its pricing, economy, your living situation, country, etc. 200 and up, sure though.
Polaris was decent value. Artic Islands? Eh. I can't really say so.
The RX580 was always faster than a 1060, even if you look back to reviews from 2018. At that point nobody cared though - the 1060 was already 2 years old and the RX580 was just an overclocked 480, so pretty unexciting and people were probably disappointed, hoping for more from AMD than a factory overclocked refresh that offered 5-10% more performance than a two-year-old Geforce and sucked down 180-200W instead of the 120W of the old Geforce.The 1060 3GB was doubly trashed by reviewers at launch.
- for having only 3GB which was detrimental at a time when 2GB cards were already struggling for VRAM in AAA games and 4GB cards were on borrowed time.
- for calling it a 1060 when it had fewer cores and TMUs than the 6GB variant of the 1060.
Nvidia didn't even try to clock it higher to compensate - it was simply a lesser card, deceptively sold with the "1060" name to trick people into thinking it had 1060 performance. Even if it had been given 6GB of VRAM it wouldn't have matched a 1060, but the lack of VRAM meant that it was much slower in some games where 3GB wasn't enough.Same with the 3050 and now the 5050. The specs are fully published and people know what they are buying. Niche or not, there is a demand for such cards and NVidia, AMD & Intel all know it.Yeah, and that was just as dumb then as it is now. Though to be fair, there was FAR less drama then now, comparatively.Your opinion is NOT supported by either merit or general consensus. The 1060 3GB was in the top 5 of the Steam survey for a very long time. Not as long as the 6GB model, but still a very long time. It was a very popular seller because, and try to follow along here, it was an EXCEPTIONAL value for money. Why? Because it performed within a close margin of it's 6GB sibling.
Just because you can hypothetically find the specs doesn't mean that someone actually will go out of their way to do so. Sure you can shift blame onto the consumer, especially if its a digital purchase, but physical purchases? I would disagree. I dont have a problem with the 1060 3GB because not such a horrendous cut down from its 'full' version that its bad enough to call out. The 3050 6GB absolutely is though.
Again, doesnt mean that the 3050 6GB is a awful card, or that it was made to be misleading (though the way it was announced and quietly released there after was suspicious as hell.), it just means the card is misleading for the average person. Which, I should add, matters more to me than tech enthusiasts being mislead. I still like the 3050 6GB. I won't pretend that its quiet annoucement and release alongside having the same name as the 3050 8GB while basically being a completely different card is not suspicious, and I think it should always be pointed out, atleast the fact that it and the 8GB model are very much different cards.
It's a matter of logic in my mind.. if the card is different enough to warrant having a different GPU die name (GA106 vs GA107) then its absolutely different enough to warrant a different name, nuff said. But hey, the card is old enough now that most people know what they're buying when they get a 3050 6GB so I really dont care as much anymore.I cannot personally find that many reviews that which mentioned this fact vs the positive reviews. Maybe times were different back then? Not sure, im not the best at looking back at history on the internet.Like I said before to lex, the performance difference is not high enough that its worth me getting bothered by. Still sucks, dont get me wrong, but personally speaking I dont mind the 1060 3GB. I did mind the 3050 6GB though, which I explained in the response to Lex.