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
How about the GTX 260 VS the GTX 260, eh?
www.techpowerup.com/gpu-specs/geforce-gtx-260.c217
www.techpowerup.com/gpu-specs/gigabyte-gtx-260-core-216-oc.b7817
Or perhaps the GTX 560 VS GTX 560, hmm?
www.techpowerup.com/gpu-specs/geforce-gtx-560.c287
www.techpowerup.com/gpu-specs/geforce-gtx-560-oem.c891
Or maybe the RTX 3060 Vs RTX 3060?
www.techpowerup.com/gpu-specs/geforce-rtx-3060-8-gb.c3937
www.techpowerup.com/gpu-specs/geforce-rtx-3060-12-gb.c3682
Are any of THOSE cards scams too? Hmm? And the correct answer is...No, not a scam. It's called a difference in model specs and nothing more. By your daft logic, the 3060 8GB must be a scam because it's got a lower VRAM bus width and only 8GB of VRAM compared to it's 12GB sibling. :rolleyes:Neither is your version of supposed "facts". Do you think anyone is taking you seriously? Your special brand of drivel is not convincing anyone with any modicum sense.
Your only argument here is "you post nonsense". Please, explain to me and all those who don't take me seriously, but take you seriously, how a 3GB card costing as much as 8GB models that where performing more or less the same, was a better buy.
Now, we're discussing the imminent release of the RTX 5050, then you were attempting, and failing, to claim some sort of scam was going on. That's your failure, not mine, not the other readers or forum members, not NVidia's, not the AIB's, yours. Unless you can prove your assertion...
The most fun with Nvidia is the low end parts. GT 730 in particular. Under the same model number you could find Kepler, Fermi, 64bit, 128bit, GDDR3, DDR5 specs. A huge mess, but it was a low end card and with an Nvidia logo on it, so no one bothered.Examples? Again, it's easier to find example from Nvidia. I don't remember any from AMD. AMD used to just re badge, because of their financial constrains, especially before Ryzen CPUs, that was limiting their ability to produce enough new models. Like R9 280X for example that was the HD 7970 GHZ edition under a different new name. That's the example you are trying to find.
Nvidia was thinking clever with the 3GB model. The average user would go and buy simply a "GTX 1060". Because if someone wanted a good enough Nvidia GPU, the 6GB GTX 1060 was a great option. So, they used the popularity of the 1060 to sell even cut down dies. I am pretty sure that many people bought the 3GB model only to hear from friends who had more knowledge of GPUs that they bought "the wrong 1060".Whatever makes you happy.Don't sugar coat it. You become rude in an arrogant way. The type of "There are two opinions, my opinion and the wrong opinion". Or maybe more like "My truth and BS that needs to be crushed". Anyway what you call crap or bs or nonsense, it's just a different opinion. I guess you don't like different opinions. Who would have thought, people having different opinions in a forum.No, you where talking about the GTX 1060 3GB. Maybe you forgot what I quoted. So, please, at least before throwing lies, try to verify them.
One of the funny things is that when people say 1080p they don't realize people aren't running 1080p on those monitors. They are running lower than that. PC games is more of a 30fps at 720p mess than the Switch could ever hope to be. Something like this is actually a massive upgrade. When you see what PC gaming actually entails for most people you realize it's the worst looking most hilarious clusterfuck out there.
Even nvidia doesn't know the answer to that
Especially a single slot sub 75w version. That is a niche of a niche of a niche, and nobody is servicing that. Just not a big enough market. If they DID make one, it would be priced like a new A series quaddro card anyway.
I remember back in the day stores being flooded with the 3GB 1060, and some vendors trying to hide the 3GB part on the box very well, Gigabyte and MSI being chief among them, so perhaps he's thinking of that.
This is a 3GB SKU but unlike let's say EVGA there's no 3GB on the front, I am sure some people have been scammed this way. It honestly should've had another name.
Anyhow this 5050 card is getting a bit too much hate, it has some market appeal with its pricing.
You can put a graphics card into almost any system that has a free PCI-E slot.
You can't put an APU in anything except a board designed for said APU and you'll have to like the CPU that the APU contains. (i.e. If say for instance Person A found a good deal on a strong CPU without an iGPU (Like a 14900KF, 5950X...etc.) and intends to use the system exlusively for CPU-heavy tasks. What good would an APU do them? They can't stick an APU in there if they want to use the CPU.)
Just in case you needed a reminder that the 5060 Ti 8GB was pointless.
This type of card will really shine if it doesn't require external power. But priced too closely to the 5060/9060 unless it's genuinely not that much slower.
That is not a scam in my book. With the same argument, 4070 Ti (remember: it was going to be called 4080 12GB before the totally justified backlash) is a scam. And while I don’t particularly like that card in price-to-performance, it has its place. And the 4070Ti super is one of the best value cards in that generation and price range.
For me, this 5050 is very uninteresting at 130W TGP. The RTX A2000 is the rival to beat here, in efficiency, for the HTPC/homelab users. Some folks do not undertand that some people (like me) do NOT want a mere 30% more performance at the cost of almost doubling the total power consumption of a card that will be powered 24/7
with an RX480 8GB cause it was about same performance, same price, but had 8GB instead of 3GB, but that and the 6GB sold here in big numbers as well.
Curious on the 5050 though, got my RX660 for $235CAD ($172USD) a few years ago, be interesting how the 5050 compares at $342cad.
Still be nice if there was some cheap modern cards for people that don't game and don't have igpu
We should be demanding lower power usage and great performance.
dGPU's for the win.They're not arguing prices, they're arguing specs, so geography doesn't really factor in. Fair thought though.