• 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.

AMD VEGA confirmed for 2017 H1

Status
Not open for further replies.
Do you honestly think any money from Ryzen has had an affect on Vega's R&D?

It's still a product from 2-3 years of development, long before Ryzen was even tapped out.

Of course Ryzen resources didn't impact Vega R&D (as in money came in too late for that), but nobody really cared: AMD the past 2 years was all about Ryzen. It was in "deliver a good Ryzen or go bankrupt" mode. Seeing as how they've delivered - and at a very high standard - they can start putting in more R&D.

But if you ask me, what they need to do is put money toward deep collaborations with game studios to push DX12/Vulkan (like they are doing with Bethesda).

Everyone compares AMD cards to Nvidia cards and thinks 'AMD cards are so much worse, they consume more and provide less FPS'. But what they don't realize is that the game is rigged because of the playing field: it's not that they're so much worse, it's that they're competing in a different category.

Think of it like this: you build an off-road rally car, whereas I build a flat-road race car and we pit them against each other in a formula one circuit. Of course the race car will win! If we go to a dirt track though, the tables WILL turn.

If you look at what the RX480 is capable of in Doom/Vulkan, suddenly perf/watt isn't that bad now is it? Those extra GFLOPs are put to work and the 480 pulls ahead of the 1060. AMD need to get as many studios lined up and bring them on board. Basically sponsor properly optimized, pristine DX12/Vulkan implementations.

It's either that, or they need to give up their architecture and optimize for DX11.
 
Of course Ryzen resources didn't impact Vega R&D (as in money came in too late for that), but nobody really cared: AMD the past 2 years was all about Ryzen. It was in "deliver a good Ryzen or go bankrupt" mode. Seeing as how they've delivered - and at a very high standard - they can start putting in more R&D.

But if you ask me, what they need to do is put money toward deep collaborations with game studios to push DX12/Vulkan (like they are doing with Bethesda).

Everyone compares AMD cards to Nvidia cards and thinks 'AMD cards are so much worse, they consume more and provide less FPS'. But what they don't realize is that the game is rigged because of the playing field: it's not that they're so much worse, it's that they're competing in a different category.

Think of it like this: you build an off-road rally car, whereas I build a flat-road race car and we pit them against each other in a formula one circuit. Of course the race car will win! If we go to a dirt track though, the tables WILL turn.

If you look at what the RX480 is capable of in Doom/Vulkan, suddenly perf/watt isn't that bad now is it? Those extra GFLOPs are put to work and the 480 pulls ahead of the 1060. AMD need to get as many studios lined up and bring them on board. Basically sponsor properly optimized, pristine DX12/Vulkan implementations.

It's either that, or they need to give up their architecture and optimize for DX11.

i'm pretty sure it's a bit more complicated than just sponsor some implementations. They need to WORK with the people (just as they did with tomb rider, deus ex and supposedly prey). That's why nVidia is doing so well, they are not specifically sponsoring with money companies, they work with them.
 
Given this is a non-conventional card, I predict teething issues with games.
It wont give its best on day 1. Not sure if that means bad value at first or a bargain from future development potential. But the advance hardware is handled by AMD themselves, it shouldnt be causing a negative with prior games, I dont see why it wont work fine but fine tuning is another matter.

Basically sponsor properly optimized, pristine DX12/Vulkan implementations.

It's either that, or they need to give up their architecture and optimize for DX11.

The other option is to make your hardware as easily accessible as possible. I dont agree with all this back room sponsorship and partisanship, this game is made for nvidia this is amd.
A proper market would have both companies competing, each game and its developers able to achieve the best results from genuine personal innovation in optimising not unwritten specialist knowledge.


Sterling value is more positive now, so long as it stays above December prices, 1.31 to 1.34 seems a reasonable target. If they are releasing June, this election is decent timing to help prices
 
Last edited:
i'm pretty sure it's a bit more complicated than just sponsor some implementations. They need to WORK with the people (just as they did with tomb rider, deus ex and supposedly prey). That's why nVidia is doing so well, they are not specifically sponsoring with money companies, they work with them.
Agreed! I think its a little more deep than suggested and developers optimising for various strengths on cards is a lot less obvious than in the past of the days of 3DFX and early Geforce cards.
 
i'm pretty sure it's a bit more complicated than just sponsor some implementations. They need to WORK with the people (just as they did with tomb rider, deus ex and supposedly prey). That's why nVidia is doing so well, they are not specifically sponsoring with money companies, they work with them.

That's precisely what I meant and why I gave the recent Bethesda collaboration as an example.

Actually working with them is a form of deep sponsoring. You don't just give some money, you give actual man-hours of developers that are paid by AMD yet are coding to analyse and improve the performance of the game...
 
my bad then.
yes you're right,even giving the knowledge is very important in this stuff. Documentation is nice and crap, but hands on is the bread and butter.
 
my bad then.
yes you're right,even giving the knowledge is very important in this stuff. Documentation is nice and crap, but hands on is the bread and butter.

In the end, AMD need to prove that it's worth it. Can they build not one, but 5 examples - AAA titles - that perform like Doom? Where the RX480 is consuming 20% more but is also providing 30% more FPS? (NOOMA - numbers out of my ...)

That's how you prove your approach is superior and going DX12 is worth it. Because if you're getting the same performance as DX11 and Nvidia then you haven't really gone that extra mile now, have you?
 
Last edited:
In the end, AMD need to prove that it's worth it. Can they build not one, but 5 examples - AAA titles - that perform like Doom? Where the RX480 is consuming 20% more but is also providing 30% more FPS (NOOMA - numbers out of my ...)?

That's how you prove your approach is superior and going DX12 is worth it. Because if you're getting the same performance as DX11 and Nvidia than you haven't really gone that extra mile now, have you?

Well they are really together now with Bethesda. AMD will develop a lit of effects for tgeir games.
 
Just a few info if it wasn't posted yet.
There well be some Vega topics in May, but hard launch probably in Jun.
AMD first ships the professional cards and they are clocked at 1525MHz (LiquidSky is building on Vega GPUs).

Seven AMD Vega GPU IDs have appeared in the latest Linux driver release:
https://www.pcgamesn.com/amd/amd-vega-gpu-linux

Interesting, it mentions only having VEGA 10 IDs and no VEGA 11. So looking at a single die variant in the near future.

AMD is already leasing time on their MI25 based platform, right?
 
the thing is amd knew they where driving on 2 fronts and if they had too win they had to show force .. so the combined effort would be a good cpu (which they did ) and a good gpu (case still open)
I'm going to say that they will bring a dam good gpu out to be as competitive as the ryzen is to intel .. at a lower cost ..
this is the only way they can win 2 fronts and still stay in both games until the next revision
 
UoiaXDV.png


Any day/month now, last possible time countdown; anything before then, AMD released it early :p

since AMD didn't release vega with 500series in april, it will 100% be launched at computex taipei around 29th of May, + AIBs will have their cards there also, so there isn't much surprise left really.
plus it's better to have no info to manage the hype down for ppl to brace for reality.
 
since AMD didn't release vega with 500series in april, it will 100% be launched at computex taipei around 29th of May, + AIBs will have their cards there also, so there isn't much surprise left really.
plus it's better to have no info to manage the hype down for ppl to brace for reality.
I think every base has been covered in this thread. We have had:

It is gonna be fast
It is gonna be slow
It is gonna be average
It is gonna be cheap
It is gonna be expensive
It is gonna be average

Only thing missing is Floppers take on Vega :D
 
I think every base has been covered in this thread. We have had:

It is gonna be fast
It is gonna be slow
It is gonna be average
It is gonna be cheap
It is gonna be expensive
It is gonna be average

Only thing missing is Floppers take on Vega :D
Flopper is lurking in the cpu, ryzen thread and quite subdued by his standards
 
Just a few info if it wasn't posted yet.
There well be some Vega topics in May, but hard launch probably in Jun.
AMD first ships the professional cards and they are clocked at 1525MHz (LiquidSky is building on Vega GPUs).

Seven AMD Vega GPU IDs have appeared in the latest Linux driver release:
https://www.pcgamesn.com/amd/amd-vega-gpu-linux

Posted that a while ago when phoronix reported it :p

https://www.phoronix.com/scan.php?page=news_item&px=AMDGPU-Vega-10-Support
 
Status
Not open for further replies.
Back
Top Bottom