• 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.
Now this is where I see where AMD could falter. They generally produce very good hardware but implement tech that needs to be taken onboard and written for by the devs and with a 70/30% split between Green and Red, it is obvious that the devs are led by marketshare even when the actual tech is as good as or sometimes better than the market leader...it is not written for or used.

This has been the case since the 290 with Async Shaders etc...

Hopefully AMD can get Vega right and be more competitive, so as to get the tech leap-frogging going again as that is what the GFX industry needs now to keep it going forward and to hopefully get the prices down a bit.
:)

Think about what you're saying for a moment. If no one created new tech that had to be utilised by the dev to get improvements..... we'd still have DX8 hardware.

The ONLY, and I really mean that, the only way for graphics tech to actually keep moving forwards is to support new hardware features before the software is ready. The industry has 100% of the time always followed hardware with software, never the other way around. Devs will not waste time(which is equal to money) on adding features and writing code to utilise hardware that does not yet exist. Not least because if there is a promise that once a dev adds code that hardware guys will add hardware to utilise it, the hardware can still show up with changes and differences that necessitate rewriting the code anyway.

Hardware first, you provide new features then devs actually see an instant real world benefit to spending money writing code for it and they also have final hardware with a guide on how to code effectively to use that feature provided by the hardware company. AMD is one of the big reasons graphics tech is moving forwards, from hardware to software.

Nvidia does the almost opposite. The wait for AMD to add hardware, wait for devs to start writing the software, then add the hardware in the next gen thing. they spend less coming up with new features and more adding in a already used feature.

Yes as an AMD user it means slightly slower hardware on launch, but it also means better hardware, better performance over time and better software and graphics in general as they push companies to move graphics tech forward.
 
To be honest not that much different than Nvidia and their GTX1080. :P
Took almost 5 months for stock to be available, for those who didn't pre-ordered :rolleyes:

Quite different - even if you had trouble getting hands on stock - there was actually stock going out, some people were getting their hands on actual cards, and for the most part the less desirable models has some stock even if the desirable models had a long waiting list.
 
I wonder if one of the unannounced features will be AMD's equivalent to 'Simultaneous Multi Projection' it is a very good bit of tech for the situations that can use it.
 
Nvidia does the almost opposite. The wait for AMD to add hardware, wait for devs to start writing the software, then add the hardware in the next gen thing. they spend less coming up with new features and more adding in a already used feature.

Yes as an AMD user it means slightly slower hardware on launch, but it also means better hardware, better performance over time and better software and graphics in general as they push companies to move graphics tech forward.

That is almost breathtaking distortion of the truth.
 
That is almost breathtaking distortion of the truth.

Nvidia are the reason why we had DX10 and DX10.1 at the same time.

ATI was ready to launch full DX10 capable hardware at the time but nvidia's hardware was not ready. So they pleaded to Microsoft to alter the DX10 spec. Which is why DX10.1 was born which was the original DX10 spec.


Recently Nvidia have been ahead with a few features found in DX12.1, some provide a nice performance boost such as hardware based 'order independent transparency' although it is nothing new, the Dreamcast had OIT's implemented in hardware.

SMP is also another nice new tech from nvidia, but i don't think it will see a great deal of adoption yet till AMD release an equivalent.
 
Microsoft release DX12 Debugging tool

Interesting, Microsoft have released a DX12 debugging tool to help devs implement and code DX12 better. I am guessing this may be in part to the negative feeling a lot of DX12 has garnered, with a lot of enthusiasts.

https://blogs.msdn.microsoft.com/pix/2017/01/17/introducing-pix-on-windows-beta/

The above is a link to the MSDN article for the tool. Will be interesting to see if DX12 gets a bit of a boost with devs being able to use the tool to help troubleshoot etc.

It looks to me like Microsoft is really trying to push DX12 forward.

Wonder if AMD Vega cards have any additional DX12 feature support that the previous cards do not?
 
Yes as an AMD user it means slightly slower hardware on launch, but it also means better hardware, better performance over time and better software and graphics in general as they push companies to move graphics tech forward.

^ and that is the current problem for Amd, Not many people would want a slower card and be a guinea pig for amd's 'ambitious' tech.

Give me a card with performance now ( not a year from now ) and performs well in the future with the bonus of additional hardware/software features.
 
^ and that is the current problem for Amd, Not many people would want a slower card and be a guinea pig for amd's 'ambitious' tech.

Give me a card with performance now ( not a year from now ) and performs well in the future with the bonus of additional hardware/software features.

I do believe that will change this time where we are actually going to see something that competes at same speed now and will get faster as devs learn to use the new abilities the cards have over time.

So right when Vega drops I do feel this is a good selling point.
 
I do believe that will change this time where we are actually going to see something that competes at same speed now and will get faster as devs learn to use the new abilities the cards have over time.

So right when Vega drops I do feel this is a good selling point.

I am optimistic, and luckily a patient person.
 
I am assuming that some of the unannounced stuff will be dx12.1 stuff like hardware based order independent transparency etc.
There will probably be something we didn't know we needed, and don't actual need just yet :D
Wonder if AMD Vega cards have any additional DX12 feature support that the previous cards do not?

I would hope so.
 
Last edited:
^ and that is the current problem for Amd, Not many people would want a slower card and be a guinea pig for amd's 'ambitious' tech.

Give me a card with performance now ( not a year from now ) and performs well in the future with the bonus of additional hardware/software features.

Its not like they perform badly. They just have to be cheaper then what Nvidia offer otherwise there's no point buying it. If performance is good enough it doesn't have to be the fastest basically. That's why I got a Fury pro I looked at the benchies and drew the conclusion that it met my criteria and I can save 100 quid.

I get what you're saying though and many people do just want the fastest one just because it is the fastest.
 
Last edited:
^ and that is the current problem for Amd, Not many people would want a slower card and be a guinea pig for amd's 'ambitious' tech.

Give me a card with performance now ( not a year from now ) and performs well in the future with the bonus of additional hardware/software features.

I dunno. I prefer a card that performs well now and gets faster in the future, as opposed to one that performs well now and gets slower when the vendor wants to sell something new and gimps your drivers.

Forward looking tech means a product with a longer life, which is good for the consumer.
 
I dunno. I prefer a card that performs well now and gets faster in the future, as opposed to one that performs well now and gets slower when the vendor wants to sell something new and gimps your drivers.

Forward looking tech means a product with a longer life, which is good for the consumer.

Yep it's not like AMD cards are slow on release like some make out. The gtx780 was more expensive compared to the 290. While it was slightly faster it was not by much. These days it's a good bit slower and has 1gb of Vram less. Same goes for the 290x v the 780ti.
 
I dunno. I prefer a card that performs well now and gets faster in the future, as opposed to one that performs well now and gets slower when the vendor wants to sell something new and gimps your drivers.

Forward looking tech means a product with a longer life, which is good for the consumer.

Performance does not get slower, it just doesn't improve as much as amd which tbh is not surprising as nvidia released 3 gen cards ( 780/980/1080 to amd's 290/fury/ and mid range 480. Not to mention the titan and ti cards.
 
Yep it's not like AMD cards are slow on release like some make out. The gtx780 was more expensive compared to the 290. While it was slightly faster it was not by much. These days it's a good bit slower and has 1gb of Vram less. Same goes for the 290x v the 780ti.

The 780 wasn't more expensive when i bought it and was faster in every game "I"played at the time, 2 years of a performance advantage vs 1 year of the 290 advantage.

So in hindsight i would have still bought the 780, likewise if vega performs faster than the 1080 for similar money then i would consider that.
 
Last edited:
The 780 wasn't more expensive when i bought it and was faster in every game "I"played at the time, 2 years of a performance advantage vs 1 year of the 290 advantage.

The 780 was always more expensive than the 290 and even very early on there was little difference between them, the 290X beat the Titan-X on lauch, the 290 was not far behind.

You could get 290's for sub £300 just a couple months after it launched, mine was one of the midrange ones at £320, the 780 was never anything like that cheap until it was EOL.
 
Last edited:
Yep it's not like AMD cards are slow on release like some make out. The gtx780 was more expensive compared to the 290. While it was slightly faster it was not by much. These days it's a good bit slower and has 1gb of Vram less. Same goes for the 290x v the 780ti.

My 780GHz was more expensive than the 290 - but cheaper than most of the 290X and on release in most games it beat out the 290X in some cases by a reasonable amount and was mostly ahead of the 290X OC models and continued to do so for a good 2 years until the 290 cards started to come upto parity by which time all those cards are old news really - they've only been saved a bit by the length of time we had been on 28nm, AMD's failure to push the performance envelop and nVidia trickling out performance up ticks. In any other generation those cards would have been completely irrelevant by now late performance bump or not.
 
Nvidia are quicker to stop optimising for a range than AMD, That may all change though, If the Vega architecture is noticeably different from GCN it's possible we'll see a similar situation over the next few years, But whether we do or not there's no denying that Nvidia are a lot more pro-active when it comes to giving users a nudge onto the latest and greatest range, For example, Maxwell was promoted too have Async compute which was part of the push from Kepler but it never got it and became part of the reason why Pascals doing better in certain titles and it's part of the reason to move on from Maxwell, From what I read Pascal isn't completely ready for the new api's which gives them a way to push from Pascal to whatever's next. And that's the thing they are drip feeding the tech in order to ensure people keep buying new gpu's because at the end of the day there's going to be a point where we won't need yet another gpu because what we have provides enough performance, Nvidia know that which is why they're diversifying as a company, but for now they still have plenty of ways to make the next gen have something over the last so we feel the need to move on. They're just thinking ahead, They're a business and they need to think like that.
 
Last edited:
Status
Not open for further replies.
Back
Top Bottom