[HN Gopher] Nvidia's RTX 5050 GPU starts at $249 with last-gen G...
       ___________________________________________________________________
        
       Nvidia's RTX 5050 GPU starts at $249 with last-gen GDDR6 VRAM
        
       Author : microsoftedging
       Score  : 43 points
       Date   : 2025-06-24 16:57 UTC (6 hours ago)
        
 (HTM) web link (www.theverge.com)
 (TXT) w3m dump (www.theverge.com)
        
       | gs17 wrote:
       | And it's 8GB of last-gen GDDR6 video memory, the exact same as
       | the $249 RTX 3050 from three years ago (same number of CUDA cores
       | too). Technically with inflation that's more per dollar, I guess,
       | but that's not super appealing.
        
         | bigyabai wrote:
         | If you're using less memory, it kinda stands to reason that you
         | can get more mileage out of less bandwidth. I'd be really upset
         | if this was a 16gb or 24gb card, but we've been using GDDR6 for
         | 8gb cards without issues for years now.
         | 
         | I agree that it's not super appealing, but Team Green has to
         | hit the low price points somehow. This feels more like a hedged
         | bet against Intel trying to muscle their way back into the
         | budget market.
        
           | ryao wrote:
           | AMD uses GDDR6 on their current generation cards:
           | 
           | https://www.techpowerup.com/gpu-specs/radeon-
           | rx-9070-xt.c422...
           | 
           | Intel does too.
        
         | pama wrote:
         | The performance figure in the link clearly says that it it's a
         | significant improvement to the 3050.
        
           | cogman10 wrote:
           | This is creative marketing from nVidia. Notice the "With DLSS
           | 4".
           | 
           | That's AI frame hallucination which the 5050 has.
           | 
           | Without the DLSS, the numbers from independent reviewers has
           | basically been exactly on par with the previous generations
           | (about 10% increase in performance).
        
           | sidewndr46 wrote:
           | The charts are from the Verge, not exactly known for their
           | integrity in regards to anything.
           | 
           | It's also with DLSS on, so you could just as easily have the
           | framerate be 100 FPS, 1000 FPS, or 10000 FPS. The GPU doesn't
           | actually have to render the frame in that case, it just has
           | to have a pixel buffer ready to offload to whatever hardware
           | sends it over the link to the display. Apparently some people
           | actually really like this, but it isn't rendering by any
           | reasonable definition.
        
             | pama wrote:
             | Maybe this is the better link:
             | 
             | https://www.nvidia.com/en-
             | us/geforce/news/rtx-5050-desktop-g...
        
               | colejohnson66 wrote:
               | Why would anyone trust Nvidia to not stretch the truth,
               | _especially_ with a press release? It 's been shown
               | multiple times they inflate their numbers.
        
               | cornstalks wrote:
               | This is a better link:
               | https://gamersnexus.net/gpus/nvidia-selling-lies-
               | rtx-5070-fo...
               | 
               | It's about the RTX 5070 but the criticisms still hold for
               | the RTX 5050 since Nvidia is still doing the same
               | shenanigans.
        
               | theyinwhy wrote:
               | Looking at 20 articles, they seem to be biased when it
               | comes to AMD vs Nvidia.
        
               | unaindz wrote:
               | I don't follow everything but they give flak to AMD too,
               | maybe in a different way.
        
           | kllrnohj wrote:
           | That's Nvidia's marketing slide and if you note the fine
           | print they are tested at different settings. The RTX 5050 is
           | using 4x frame gen which the 3050 isn't. Techpowerup has the
           | RTX 5050 as being 20% faster than the 3050 give or take,
           | which is certainly not enough to justify upgrading
        
         | happycube wrote:
         | This card might've been forgivable with 16GB or a $149 MSRP, so
         | of course they didn't do either...
        
         | ryao wrote:
         | Why did I not see any complaints when AMD used GDDR6 on their
         | current generation products:
         | 
         | https://www.techpowerup.com/gpu-specs/radeon-rx-9070-xt.c422...
        
           | unaindz wrote:
           | Because they announced double the VRAM for the medium end
           | card at around the same price. But there were complaints
           | anyway.
        
       | throitallaway wrote:
       | Can't wait to buy one for $550.
        
         | singhkays wrote:
         | I know you're being facetious but such is the market atm. I put
         | something together to track the absurdity of it all :)
         | 
         | https://gpuisfine.singhkays.com
        
           | mft_ wrote:
           | It's insane. I've been tracking a few used 4090s on EU ebay
           | this week, and the typical selling price is ~1950 EUR.
        
             | singhkays wrote:
             | huh..you might has well get a 5090 for that price. Seems
             | like it's dipping below MSRP there -
             | https://www.techpowerup.com/338256/nvidia-geforce-
             | rtx-5090-b...
        
               | mft_ wrote:
               | Indeed; there are new 5090s from legit-looking sellers in
               | eBay from around ~2300 EUR.
        
       | Aurornis wrote:
       | The emphasis on last-gen memory is misplaced. I don't care what
       | memory technology is used as long as the performance is good for
       | the price.
        
         | seiferteric wrote:
         | Ya, don't current gen Radeon cards use GDDR6 as well?
        
           | ryao wrote:
           | They do:
           | 
           | https://www.techpowerup.com/gpu-specs/radeon-
           | rx-9070-xt.c422...
        
       | Night_Thastus wrote:
       | It's complete garbage and not worth buying. It's so cut down it's
       | nearly useless outside of web browsing and very light games. The
       | price is also effectively lie, it's going to be hard to get it
       | for less than $300. Once we get some proper 3rd party test data
       | in I'd be shocked if it's 5% better than a 4050 in raster without
       | the use of fake frames.
        
         | jekwoooooe wrote:
         | Why is there so much hate against fake frames? I know you can't
         | tell the difference
        
           | kllrnohj wrote:
           | because you can tell the difference, they have quite a few
           | artifacts, and they make latency worse which is especially
           | problematic in the scenarios where you need the "performance"
           | offered by fake frames. At this price point it's that last
           | thing that's especially problematic. You may get 60fps in an
           | fps counter with dlss 4, but it'll _feel_ like 15-20fps and
           | not be very playable
        
           | jamesgeck0 wrote:
           | NVidia themselves have said that framegen shouldn't be used
           | if the card isn't hitting 60 FPS to start with because of the
           | latency it introduces. If the card is cut down enough that
           | it's struggling to hit 60 FPS in games, enabling framegen
           | will do more harm then good.
           | 
           | You can feel additional latency easily in competitive FPS or
           | high speed arcade racing games.
        
             | unaindz wrote:
             | You can feel less than 50-60 fps on a management game where
             | you only interact with the UI and move the camera around,
             | not game breaking but doesn't feel great. And I used to
             | play far cry 3 and CSGO at ~25 fps, I'm used to lack of
             | performance.
        
           | Night_Thastus wrote:
           | You absolutely can tell the difference. DLSS (upscale)
           | visually is massively different in some games. Sometimes it
           | works great, sometimes the result is very ugly. I've tested
           | with several of my favorites.
           | 
           | And generated frames are far worse than that. If you're
           | running at a very high base framerate (100+) then they can
           | look OK but the moment the frames get any further apart the
           | visual quality starts to tank.
        
           | orphea wrote:
           | This is why: https://news.ycombinator.com/item?id=44368785
           | 
           | Fake frames are cool tech but they are horribly mismarketed,
           | indistinguishable from scam.
        
           | toast0 wrote:
           | Fake frames have a big latency penalty, because you can't
           | generate a frame between X and Y until you have Y. At the
           | point that you have generated frame Y, however many frames
           | you insert give you that much additional latency, beyond
           | whatever your display adds.
           | 
           | I guess I can see some utility in situations where latency is
           | not a major factor, but IMHO, that pushes out most gaming.
        
         | pitaj wrote:
         | Agreed. Anybody buying this would be better off spending that
         | money on a used card like a 2070 or 3060, and might even save a
         | buck.
        
           | jamesgeck0 wrote:
           | Or an Intel Arc card. They aren't very high end, but they're
           | competitive at MSRP and I suspect they'll demolish this in
           | benchmarks.
        
       | lvl155 wrote:
       | Just buy AMD AI Max+ no?
        
         | anonym29 wrote:
         | Undoubtedly a better system, but the 395 variant with a full
         | 128GB of (soldered on) RAM, you're looking at ~$2k for the
         | system. Comparing that to a $250 dGPU (that arguably isn't even
         | worth that) is a very "apples to oranges" comparison.
        
       | Kon-Peki wrote:
       | This is not going to go well:
       | 
       | > x50-class GeForce GPUs are among the most popular in the world,
       | second only to the x60-class on Steam. Their price point and
       | power profile are especially popular:
       | 
       | > For anyone upgrading an older x50-class system
       | 
       | > Each GeForce RTX 5050 graphics card is powered by a single PCIe
       | 8-pin cable, drawing a maximum of 130 Watts at stock speeds,
       | making it great for systems with power supplies delivering as
       | little as 550 Watts.
       | 
       | The 1050, 2050 and 3050 were all bus-powered cards. I doubt 95%
       | these systems even have the cable coming from their power supply.
       | Imagine all the poor saps that excitedly swap out their old card
       | for this, and... nothing works.
       | 
       | Source link: https://www.nvidia.com/en-
       | us/geforce/news/rtx-5050-desktop-g...
        
         | LorenDB wrote:
         | Hey, at least it's not the infamous 12-pin connector.
         | 
         | (Full disclosure, I have a 9070 XT with a 12-pin connector. No
         | fires yet, though.)
        
           | Kon-Peki wrote:
           | Well, maybe a molex-to-8pin-PCIe cable comes in the box!
        
         | reginald78 wrote:
         | To bolster this, after the 750ti the 50 products have had
         | pretty lame price to performance compared to the next step up,
         | but have remained quite popular. Most people seem to argue that
         | the lack of additional power is their main advantage and why
         | they are popular.
         | 
         | I personally think people remember being happy with the 750ti
         | and just keep buying those cards.
        
         | toast0 wrote:
         | I've got at 1650 Super; it's not bus-powered either. I think
         | it's got a 6-pin, but often you can plug a 6-pin into an 8-pin
         | board and it'll just run a lower current limit (this might not
         | be accurate --- a lot of internet comments say 8 pin boards
         | will detect a 6-pin connector and refuse to work). A whole lot
         | of modern computing gets ~ 90% of the performance with 50% of
         | the power; so if using a 6-pin lead drops power to 50%, you
         | would still get most of it.
         | 
         | I've got a ~ 2006 380W power supply hanging out near my desk
         | and it's got a 6-pin pci-e cable; I really don't think people
         | won't have at least that, certainly not 95% of systems with a
         | pci-e x16 slot.
        
       | neepi wrote:
       | Just bought a 16Gb 5060 Ti for twice that. I don't feel
       | disappointed.
        
         | happycube wrote:
         | Yeah, moreso since that's the highest nvidia card this gen
         | available with a sane power connector.
        
           | sidewndr46 wrote:
           | 12VHPWR has to be one of the weirdest industry decisions I've
           | seen in a while. So far I thought I had been able to avoid
           | it, but recently bought a power supply that uses it on the
           | modular cable connector.
           | 
           | But it isn't really that uncommon either, I had a suzuki
           | motorcycle that used a connector with 15 amp pins to handle
           | 30 amps of current on one pin. I eventually concluded the
           | only reason that connector was in the harness was to ease
           | assembly and just cut it out entirely and soldered the
           | junction together.
        
           | neepi wrote:
           | Oh yeah totally agree with that one. I hate Molex connectors
           | but the new one is just stupid.
        
       | LorenDB wrote:
       | I can't believe that nobody has yet mentioned the Intel Arc
       | Battlemage B580. Same $250 MSRP (which has inflated, but every
       | other GPU is inflated too, and the 5050 will probably inflate as
       | well), but has 12 GB of VRAM and bats just below a 4060 Ti 16
       | GB[0].
       | 
       | [0]: https://www.tomshardware.com/reviews/gpu-hierarchy,4388.html
        
         | jamesgeck0 wrote:
         | I've been pretty happy with my Arc A770 LE (16 GB). The drivers
         | were rough at launch, but they've gotten much better, and at
         | the time it was the best performance $350 could buy.
        
           | prossercj wrote:
           | How is it for gaming? Had any compatibility issues?
        
             | KronisLV wrote:
             | I had both an A580 (not an A770, but at least something
             | from that generation) and then later a B580, at one point
             | even both in the same computer, side by side, when I wanted
             | to use one for games and the other for encoding:
             | 
             | https://blog.kronis.dev/blog/what-is-ruining-dual-gpu-
             | setups
             | 
             | https://blog.kronis.dev/blog/more-pc-shenanigans-my-setup-
             | un...
             | 
             | https://blog.kronis.dev/blog/two-intel-arc-gpus-in-one-pc-
             | wo...
             | 
             | When paired with a worse CPU like a Ryzen 5 4500, the
             | experience won't always be good (despite no monitoring
             | software actually showing that the CPU is a bottleneck).
             | 
             | When paired with a better CPU (I got a Ryzen 7 5800X to
             | replace it, eventually with an AIO cause the temperatures
             | were too high under full load anyways), either of them are
             | pretty okay.
             | 
             | In a single GPU setup either of them run most games okay,
             | not that many compatibility or stability issues, even in
             | older indie titles, though I've had some like STALCRAFT: X
             | complain about running on an integrated GPU (Intel being
             | detected as such). Most software also works, unless you
             | want to run LLMs locally, where Nvidia will have more of an
             | advantage and you'd go off the beaten path. Most annoying
             | I've had were some stability issues near the launch of each
             | card, for example running the B580 with their Boost
             | functionality on in their graphics software sometimes
             | crashed in Delta Force, no longer seems to be an issue.
             | 
             | Temperature and power draw seem fine. Their XeSS upscaling
             | is actually really good (I use it on top of native
             | resolution in War Thunder as fancy AA), their frame
             | generation feels like it has more latency than FSR but also
             | better quality, might be subjective, but it's not even
             | supported in that many games in the first place. Their
             | video encoders are pretty nice, but sometimes get
             | overloaded in intensive games instead of prioritizing the
             | encoding over game framerate (which is stupid). Video
             | editing software like DaVinci Resolve also seems okay.
             | 
             | The games that run badly are typically Unreal Engine 5
             | titles, such as S.T.A.L.K.E.R. 2 and The Forever Winter,
             | where they use expensive rendering techniques and to get at
             | least 30 FPS you have to turn the graphics way down, to the
             | point where the games still run like crap and end up
             | looking worse than something from 5 years ago. Those were
             | even worse on the A series cards, but with the B series
             | ones become at least barely playable.
             | 
             | In a dual GPU setup, nothing works that well, neither in
             | Windows 11, nor Windows 10, neither with the A580 + B580,
             | nor my old RX 580 + B580: system instability, some games
             | ignoring the Intel GPU preference being set when an AMD one
             | is available, low framerates when a video is playing on a
             | secondary monitor (I have 4 in total), the inability to
             | play games on the B580 and do encoding on the A580 due to
             | either just OBS or also the hardware not having proper
             | support for that (e.g. can't pick which GPU to do encode
             | on, like you can with Nvidia ones, my attempts at patching
             | OBS to do that failed, couldn't get a video frame from one
             | GPU to the other). I moved back to running just the B580 in
             | my PC.
             | 
             | For MSRP, I'd say that the Intel Arc B580 is actually a
             | good option, perhaps better than all A series cards. But
             | the more expensive it gets, the more attractive
             | alternatives from AMD and Nvidia become. Personally
             | wouldn't get an A770 unless needed the VRAM or the price
             | was really good.
        
         | some_random wrote:
         | I have to assume things are better to some degree but last I
         | looked at Intel's offering the support was still unacceptably
         | bad. That said, I really hope they can get things to a good
         | state because there needs to be more competition at this price
         | point.
        
           | kllrnohj wrote:
           | The support is still worse, but you're getting a big discount
           | on the hardware by comparison. So it kinda evens out at this
           | price point where you're deciding between either having every
           | game run badly or most, but not all, games running decently
        
       | speed_spread wrote:
       | For such insanity, it should be labeled the RTX 5150
        
       | gunalx wrote:
       | The 50 line is the new gt-30 tier.
        
       | dunno7456 wrote:
       | It will take Nvidia 10 years to release the firmware for PMU and
       | then they will cancel it because it's "too old". Just like the
       | they did with pascal, P520 and other perfectly working hardware
       | that are barely usable to this day.
        
       | drcongo wrote:
       | I'm still waiting for my DGX Spark. Starting to wonder if Nvidia
       | have hired Musk for their PR to promise things they'll never
       | deliver.
        
       | yrcyrc wrote:
       | Not following much hardware news. What can I do with this or the
       | Intel Arc? Play games? Run AI workload? Genuine question
        
         | LorenDB wrote:
         | Both. AI is dependent on available VRAM, so the B580 will run
         | some larger models than the 5050.
        
       | eighthourblink wrote:
       | Coming from a 2060 Super, would this be a good upgrade? I dont
       | really play newer high demand games, but i do enjoy my emulation.
       | Currently on 2060 super and dont really have any issues with
       | emulation.Ryzen 5 3600X / linux (of course :))
        
       ___________________________________________________________________
       (page generated 2025-06-24 23:01 UTC)