Here it is: Radeon RX Vega reference card shows up on photo

Published by

Click here to post a comment for Here it is: Radeon RX Vega reference card shows up on photo on our message forum
#5455720
data/avatar/default/avatar29.webp
Wonder is they were provided with an updated driver as well.
#5455726
data/avatar/default/avatar09.webp
LOL, a blind test. Just LOL. I remember Pepsi vs Coca-Cola blind tests... You can't beat technology by taste preferences.
#5455734
https://forums.guru3d.com/data/avatars/m/248/248627.jpg
Doing blind tests is just a bad idea i get this feeling they know they have a turd of a product.
#5455750
https://forums.guru3d.com/data/avatars/m/220/220755.jpg
well, AMD doesn't want to say "ok guys we couldn't reach 1080Ti performance level, but hey! we still have a nice product, if you are planning to buy a variable refresh rate monitor setup with this card you will expend some less bucks" of course they have a great product but it is very delayed and the expectations on it were just huge.. i hope the next iteration of flagship cards of AMD take the crown of performance again (at least for some months)
#5455754
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
Doing blind tests is just a bad idea i get this feeling they know they have a turd of a product.
Why? The sole purpose is to prove that people's expectations are skewed. AMD's point (or at least what they're gambling) is despite the numbers, you're not going to notice a difference, so why not go for them and pay less? Blind tests are the best tests. It removes bias. In blind tests, Pepsi wins over Coke. Cheap wine wins over expensive. More qualified people get a job regardless of their race, sex, ethnicity, religion, etc.
#5455758
data/avatar/default/avatar38.webp
well, AMD doesn't want to say "ok guys we couldn't reach 1080Ti performance level, but hey! we still have a nice product, if you are planning to buy a variable refresh rate monitor setup with this card you will expend some less bucks" of course they have a great product but it is very delayed and the expectations on it were just huge.. i hope the next iteration of flagship cards of AMD take the crown of performance again (at least for some months)
Umm sorry what? Asking for better performance then year and half old card is huge expectations? No one asked for Volta killer but Vega should have been able to defeat 1080 Ti comprehensively by now. :bang::bang::bang: Vega should have been ok if it was released last year mid summer not now when Ti will probably own it and not to mention volta is just around the corner which will take the performance gap to even more proportions.
#5455764
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
Umm sorry what? Asking for better performance then year and half old card is huge expectations? No one asked for Volta killer but Vega should have been able to defeat 1080 Ti comprehensively by now. :bang::bang::bang: Vega should have been ok if it was released last year mid summer not now when Ti will probably own it and not to mention volta is just around the corner which will take the performance gap to even more proportions.
I think people tend to take for granted that Pascal was one of the best architectures Nvidia has ever released. Even calling it a "big improvement" over Maxwell is an understatement. AMD was not prepared for it. If Nvidia released a more typical incremental upgrade over Maxwell, AMD's situation would have looked a lot less grim. Meanwhile, Nvidia did so much of a good job with Pascal that they have the funding to yet again go beyond just a typical incremental upgrade (with Volta). Nvidia's position with Pascal is like what Intel's position was with Sandy Bridge - AMD has no choice but to ditch GCN if they want to compete. But again, they weren't expecting this, so they thought they could hold onto GCN just a little bit longer.
#5455766
https://forums.guru3d.com/data/avatars/m/258/258664.jpg
Looks like a huge card. What happend to the space saving qualities of HBM2? I'm more curious about their water cooled cards tbh, if they can save some space on the PCB / cooler side with that, it could be a great thing to have a "smaller" version of the card for small form factor enthusiasts, like the Nano.
#5455771
https://forums.guru3d.com/data/avatars/m/209/209146.jpg
Looks like a huge card. What happend to the space saving qualities of HBM2? I'm more curious about their water cooled cards tbh, if they can save some space on the PCB / cooler side with that, it could be a great thing to have a "smaller" version of the card for small form factor enthusiasts, like the Nano.
If it's like the FE version then there's a secondary part to the water cooling component which is why that card is a bit longer too but there's also additional electronics even if the VRAM is now stacked on the GPU die. https://www.pcper.com/files/imagecache/article_max_width/review/2017-07-17/IMG_4733.JPG From: http://www.neogaf.com/forum/showthread.php?p=244514952#post244514952
#5455772
https://forums.guru3d.com/data/avatars/m/16/16662.jpg
Administrator
Looks like a huge card. What happend to the space saving qualities of HBM2?
Likely more radiator surface area is needed for cooling.
#5455781
data/avatar/default/avatar11.webp
Looks like a huge card. What happend to the space saving qualities of HBM2? I'm more curious about their water cooled cards tbh, if they can save some space on the PCB / cooler side with that, it could be a great thing to have a "smaller" version of the card for small form factor enthusiasts, like the Nano.
Ask Nvidia. Their $6,000 Quadro GP100 is just as big as their other GPUs. http://www.pny.com/nvidia-quadro-gp100 I doubt youll see a low TDP version thats tiny until they redo some of the circuit layout and put it on a low power optimized process with less leakage. They might not do that with this particular chip.
#5455798
https://forums.guru3d.com/data/avatars/m/270/270008.jpg
Why? The sole purpose is to prove that people's expectations are skewed. AMD's point (or at least what they're gambling) is despite the numbers, you're not going to notice a difference, so why not go for them and pay less? Blind tests are the best tests. It removes bias. In blind tests, Pepsi wins over Coke. Cheap wine wins over expensive. More qualified people get a job regardless of their race, sex, ethnicity, religion, etc.
Yeah but a very large part of it isn't something subjective like your other points. Sure you can blind test playing game x and its fine but we all will need performance numbers to do the value judgment. Also its not a static purchase playing only one game, new more demanding games come out. Sure you could blind test each new game but more so one would want to see this card has headroom for the future as well. Sites like this and many others will have hard data behind there testing so no worries from my end.
#5455799
https://forums.guru3d.com/data/avatars/m/80/80129.jpg
Ask Nvidia. Their $6,000 Quadro GP100 is just as big as their other GPUs. http://www.pny.com/nvidia-quadro-gp100 I doubt youll see a low TDP version thats tiny until they redo some of the circuit layout and put it on a low power optimized process with less leakage. They might not do that with this particular chip.
I don't think there is a big push for smaller workstation GPU's. Like a consumer might want to shove a gaming card into a miniITX case or something, but a workstation user probably couldn't care less - would rather have better cooling/performance.
#5455801
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
Yeah but a very large part of it isn't something subjective like your other points. Sure you can blind test playing game x and its fine but we all will need performance numbers to do the value judgment.
But that's exactly my point - if you are perfectly satisfied with a [cheaper] product, what difference do the numbers make? Most of the time, PC enthusiasts just get in a pissing contest with numbers that (in most cases) change literally nothing about their experience. People care more about bragging rights than value. Companies like Nvidia know this, which is why the Titans are so woefully overpriced. All that being said, I wouldn't be in the market for a Vega GPU. Without seeing benchmarks, I already know it's not good enough for 4K, and a GTX 1070 makes for a better value in terms of 2K. There are other appropriate reasons to not want Vega, such as thermals and potentially noise, but very rarely does the FPS actually matter in playability. Sure, if you're competitive in shooters then every frame counts, but for just about everything else, people need to get off their high horse. As much as you want to be noticed, nobody cares about your performance (not you specifically, just people in general). It amazes me how many people care about streaming or recording their gameplay, as though anyone is going to watch it.
#5455813
data/avatar/default/avatar34.webp
I don't think there is a big push for smaller workstation GPU's. Like a consumer might want to shove a gaming card into a miniITX case or something, but a workstation user probably couldn't care less - would rather have better cooling/performance.
Im just saying that the HBM may make the GPU and RAM have a smaller footprint, but the cooler has to be quite large since its stilla high TDP GPU. If they made a "mobile version" style RX Vega with low power in mind you may be able to have a compact cooler, but this is not that kind of chip.
#5455818
https://forums.guru3d.com/data/avatars/m/175/175780.jpg
The blind test is just outright stupid if you ask me, i dont believe that its good way to showcase a new graphic card at all, i think most people will still ask the question "so what is in fact fastest / most efficient " etc. Anyway, hope for a nice product, would happily exchange my GTX 1070 for an equal performing AMD product.
#5455821
https://forums.guru3d.com/data/avatars/m/132/132389.jpg
Anyone know an estimate/rumour of what it might cost?
#5455824
https://forums.guru3d.com/data/avatars/m/63/63215.jpg
Lol, blind test. Next they'll do a blind test for cpus XD

Next Page