• 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 **

On or off the hype train?

  • (off) Train has derailed

    Votes: 207 39.2%

  • (on) Overcrowding, standing room only

    Votes: 100 18.9%

  • (never ever got on) Chinese escalator

    Votes: 221 41.9%
  • Total voters
    528
Status
Not open for further replies.
B12B6 said:
I dont really understand what you are tying to say when I already wrote 2 times the point of coming real multigpu chips is to them not to have load frame for both gpus. The whole point is to get them use same memory and load the frame only once and have the gpus both share that. If they can share the memory we dont need double the memory, nor the bandwich.
Click to expand...

you are saying FRAME, yes, but you are ignoring textures, if textures are not in VRAM they need to be loaded from somewhere else and you haven't actually stated a reasonable solution that isn't latency intensive, if they are sharing memory they are sharing bandwidth and increasing latency (because sharing memory would need an off chip controller to prevent other issues)

at present a GPU loads all or nearly all of a level's textures to VRAM so that they are ever present and easily accessible without latency - splitting a single frame to smaller chunks does not solve this issue

basically, you appear not to have a clue what you are talking about and no one has actually stated how they intend to solve these issues

I'm not saying these issues are not at all solvable, but the actual performance increase the solution represents remains to be seen as to whether it is really worth it over a monolithic design

saying "because CPU" isn't an answer as the challenges the two face are fundamentally different

when mantle/dx12/vulkan first started appearing there were statements that they would allow GPU's in crossfire/SLI to share VRAM, but for the same reasons many of us with a little bit of knowledge questioned this and pointed out the same reasons as to why it wouldn't save much or offer much of an advantage and to date no one has even tried (with DICE devs even pointing out the same issue as to why they would not even try it).
 
Sargatanas2511 said:
That's the limitation they're removing, the limitation of single monolithic die size. They could possibly create multiple smaller chips that equate to a much larger than possible current single die. This will bring cost reductions and performance benefits.

I understand it won't be a linear performance increase even if they get it working well, but theoretically it could limit or eliminate the barrier of die size/cost/failure rate which the industry has been at for the last 5 years or so.
Click to expand...

Nvidia is also working on doing something similar so I am not sure why they are so annoyed if AMD is also looking at similar technology.
 
AthlonXP1800 said:
https://videocardz.com/71776/amd-is-trying-to-sell-radeon-rx-vega-64-at-a-higher-price
Click to expand...

I've come to the conclusion that AMD are trying to drive up GPU prices, coz that 599$ "pack price" nonsense is not all that dissimilar to the founders edition tax, we won't be able to find 499$ prices (which is already too expensive imo) at release, allowing the retailers to do the rest of the price gauging because the mining craze allows that at the moment.
Imo they are trying to "make" Nvidia bring up their prices, coz if they can't gain market share they can at least try to make more profit on the little market share they have.
 
JediFragger said:
I think that eVGA 1080 Hybrid is going to prove a hell of a bargain come Monday...
Click to expand...

Bargain is a relative term - we had aftermarket R9 290 cards for under £200 a few years ago,and a mate just managed to get a brand new MSI GTX1070 from the European part of a major retailer(was posted on HUKD and they had to wait two months for it) for £280ish after cashback.
 
CAT-THE-FIFTH said:
Bargain is a relative term - we had aftermarket R9 290 cards for under £200 a few years ago,and a mate just managed to get a brand new MSI GTX1070 from the European part of a major retailer(was posted on HUKD and they had to wait two months for it) for £280ish after cashback.
Click to expand...

Yeah those 290's were a billy-bargain considering their longevity!!! :cool:

Used to love my pair of Sapphire Tri-X 290's, when Crossfire was in its hayday :)
 
This price!!!!

AMD.PNG
 
orbitalwalsh said:
https://videocardz.com/71771/final-radeon-rx-vega-64-3dmark-performance

***not sure if been posted last page
Click to expand...

It's worth noting that when the HD7970 launched it was only ~8% ahead of the GTX580, and then when the 600 series launched the GTX670 was able to match it, but following the driver/BIOS updates it stomped those cards and beat the GTX680/770.

I'm not saying the same is sure to happy here, just pointing out that prerelease performance leaks can be very very far off the final mark.


Ace Modder said:
This price!!!!
Click to expand...

That's the FE, the Titan XP rival.
 
B12B6 said:
The whole point is to provide unified memory for all the gpus where the dont have to load whole screen. Multiple smaller gpus wont need uber high bandwhich then. Braking the screen in multiple parts helps them to get max utilization to all shader cores. I dont know why AMD cards wont have more than 4 shader engines, maeby its the architecture but other reason is it wont benefit in compute workloads. Seems you edited you post, yes memory problem needs to be adressed by the board. But DX12 Vulkan brings usage of multiple gpus on another level.
Click to expand...
CAT-THE-FIFTH said:
Nvidia is also working on doing something similar so I am not sure why they are so annoyed if AMD is also looking at similar technology.
Click to expand...


Who is annoyed at AMD?


The only annoying think is clickbait articles trying to get ignorant AMD diehards to believe that Navi is going to be some miracle that cures cancer.

Navi will be a monolithic die.
When MCM gous come to fruition beyond 7nm, the goal is purely to reduce manufacturing costs since the process gets so expensive.
What AMD have publicly talked about is simply called crossfire, and getting developers mutiple GPUs. This is self-evident by the fact AMD just aren't bothering making high-end compeition and instead want developers to better code crossfire compatible games
 
D.P. said:
Who is annoyed at AMD?


The only annoying think is clickbait articles trying to get ignorant AMD diehards to believe that Navi is going to be some miracle that cures cancer.

Navi will be a monolithic die.
When MCM gous come to fruition beyond 7nm, the goal is purely to reduce manufacturing costs since the process gets so expensive.
What AMD have publicly talked about is simply called crossfire, and getting developers mutiple GPUs. This is self-evident by the fact AMD just aren't bothering making high-end compeition and instead want developers to better code crossfire compatible games
Click to expand...

Because everytime people talk about AMD going to multiple chips in AMD THREADS,the same people get annoyed saying it will be rubbish or bad or will never happen whilst ignoring the fact Nvidia is doing the same,and yet never do that in Nvidia threads saying the same thing.

So it is hypocrisy really that the same people have not been moaning in Nvidia threads about Nvidia looking at the same thing.

All they are doing is trying to thread thrash AMD threads.

Whether Navi does it not is not relevant,unfortunately for certain people,its the way things will go whether they like it or not,so they will need to deal with it.
 
Status
Not open for further replies.
Back
Top Bottom