-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Poll: ** The AMD VEGA Thread **
- Thread starter TNA
- Start date
- Status
I'd say best case is closer to 15% above Ti, while sticking within the realms of reason at least. I'd say matching the Ti is the minimum required from Vega, anything less and it's an instant loss for AMD. Obviously BEST case would be 1000000000000000000000000000000x faster but that's definitely not going to happen.
About a web store with made-up specs trying to get pre-order money out of customers? More third-hand click-bait I'm afraid.
Technically no but the 12.5TF should give some indication.Do we know for sure Greg lmao.
Correct. Its from a component distributor of workstation graphics and HPC solutions
https://exxactcorp.com/index.php/product/prod_detail/2565
Hold up, doesn't the Titan Xp use GDDR5X? That site lists GDDR5, and Half Precision at 190 GFLOPs, the Titan Xp should be doing Half Precision at 12 TFLOPS, same as Single Precision.
Looks like they either can't get their info correct, or it's mostly all made up.
As I pointed out before, the Radeon Pro WX 7100 ( 5.7 TFLOPs ) is matching the Vega FE results they've posted. Hmm.
Well there we have it, the Vega Frontier Edtion can barely beat the Radeon Pro WX7100 ( £660 ) aka RX 480.
Vega FE
Catia = 135.75
CREO = 83.94
SolidWorks = 114.88
Cinebench OpenGL = 183.28
Radeon Pro WX 7100 ( £660)
Catia = 104.34
CREO = 82.08
SolidWorks = 111.55
Cinebench OpenGL = 197.11
Titan Xp
Catia = 107.29
CREO = 65.2
SolidWorks = 67.75
Cinebench OpenGL =169.72
What was the source of this info again? Based on the above, I'd say it can be safely ignored...
The source is an american online seller that sells workstation cards. They removed the price it seems, as it was $1200 listed earlier.
https://exxactcorp.com/index.php/product/prod_detail/2565
I think it's very suspect that the FE can barely beat a Polaris 10 card.
Yeah, it makes little sense really. No way the FE card was run using Radeon Pro drivers for those scores. Not if the Polaris WX 7100 can match it, and hell beat it in OpenGL.
Exactly. Obvious that Vega will rip a rx480 to bits. Why is that even being brought up.
So is the general conclusion that both cards were running gaming drivers?
Also can you run gaming drivers and professional drivers at the same time on a Titan Xp?
If Vega is running gaming drivers why is everyone comparing it cards running professional drivers? (Doesn't really make sense to me to do that.)
Also can you run gaming drivers and professional drivers at the same time on a Titan Xp?
If Vega is running gaming drivers why is everyone comparing it cards running professional drivers? (Doesn't really make sense to me to do that.)
Caporegime
- Joined
- 8 Jul 2003
- Posts
- 30,063
- Location
- In a house
Vega is their Tesla card, its nothing like the Titan, as thats a gaming card, AMD don't do gaming cards.
Last edited:
Soldato
- Joined
- 26 Sep 2013
- Posts
- 10,756
- Location
- West End, Southampton
fixed it for you mate.
Lets not be too hasty in dismissing all the results as fake!
At least some of the results are genuine.
Posted earlier by Chuk_Chuk
The blog published on 16/05/17 was written by Raja and corroborates the Catia, CREO & SolidWorks results.
The drivers listed as being used were AMD graphics driver 17.20 & NVIDIA graphics driver 382.05. Does that mean they were the gaming drivers?
.jpg)
Under performing!!!!!
Well there we have it, the Vega Frontier Edtion can barely beat the Radeon Pro WX7100 ( £660 ) aka RX 480.
Vega FE
Catia = 135.75
CREO = 83.94
SolidWorks = 114.88
Radeon Pro WX 7100 ( £660)
Catia = 104.34
CREO = 82.08
SolidWorks = 111.55
Titan Xp
Catia = 107.29
CREO = 65.2
SolidWorks = 67.75
At least some of the results are genuine.
Posted earlier by Chuk_Chuk
The blog published on 16/05/17 was written by Raja and corroborates the Catia, CREO & SolidWorks results.
The drivers listed as being used were AMD graphics driver 17.20 & NVIDIA graphics driver 382.05. Does that mean they were the gaming drivers?
Associate
- Joined
- 27 Dec 2014
- Posts
- 1,686
- Location
- Southampton
oh dear, so it got to this kind of jokes ...!
Soldato
- Joined
- 26 Oct 2013
- Posts
- 4,112
- Location
- Scotland
And Vega is a Street Fighter 2 character? Come on, stop being ridiculous.
neither of them run 'professional' drivers, thats the quadro type of market, hence why the p4000 is quicker in professional workflow (as the drivers are optimised for it)
technically you can run quadro drivers on the titan with a bios flash. I'm not sure how amdwill handle the drivers for this card, but currently as it is, it's not running their pro software.
well it performs nothing like a quadro in professional level, it's not priced like a professional card.
it's price and performance are just like a titan...
Not remember Bulldozer?
Soldato
- Joined
- 13 Jul 2004
- Posts
- 20,341
- Location
- Stanley Hotel, Colorado
If somehow people ended up thinking two cards are needed for 60fps previously then on the same theme its possible the noise on Frontier data just ends up confusing things.
We'll not know really till a normal reviewer gets a card and puts into his own rig and gives their own impression hopefully with some insight and experience on development, that will be 1 week before release ?
Its a big open world game 70 gigs or so, any card can be setup to handle frames singularly at steady FPS (in fact the game limits options by spec) but I dont think any card can handle vast sudden changes. If every single model is thrown at the gpu simultaneously theres never enough memory, the latency to call from disk, etc becomes a factor in FPS shown. Its slight normally (they design the game not to stress like this of course) but Im saying I think this drag on fps is always in play.
With new data being called constantly even on a 8 gig card thats where I think every card is being helped by HBCC. On the best card maybe its a 5% effect perhaps and with less memory or buffer its its going to be more obvious but its always going to be a factor when the game is so large and diverse. GTA isnt a one off in this vast challenge to cards. If games are being limited by this dynamic it'll be great to see them no longer in fear of it, though I agree costs is a large part of how games are made anyway.
Part of the fps will always be on even the best cards not the GPU process exactly but some delay in infrastructure to call textures.
There isnt any card in gta that can teleport from one part of the map to the other and not suffer lag from this effect. But teleporting in gta isnt supposed to be a thing, texture sourcing is speed limited by vehicles in theory. I'm just calling out something more obvious to me currently but I think applies to everyone.
If it turns out HBCC is a fantasy feature, more theory then anything they managed to implement live it will be maybe the most disappointing thing.
If on the other extreme they are able to seriously sell a 4gb card that performs just as well, that'd be great. I reckon thats the only way you see a sub £500 card before xmas, by surprise despite featuring it nobody seems to believe it'll work
Another example I saw was in CSGO. They designed 1 new map to have no LOD. Every model all over the map is max detail always, even out of sight the gpu was being asked to process it. Sure if your GPU is the top spec you dont notice so much and CSGO isnt a hard game to run at all but all the same people want the best fps or at least the most consistent not dropping in parts. But if you get intelligent infrastructure design around the GPU that somehow can help with stressful processing, that seems a feature that'll help all cards and reflect in frame rates shown. I'm quite hopeful they arent reliant on the game design to show some improvement in min fps
- Status