• 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.

The most intersting bit is in this video from 16m 42 he states that RX Vega won't be able to use the drivers that the Frontier can.

This also beat the Titan XP (older one I think) with scores that are 1250 and 1050 in Cinebench.
 
Last edited:
nashathedog said:
The most intersting bit is in this video from 16m 42 he states that RX Vega won't be able to use the drivers that the Frontier can.

This also beat the Titan XP (older one I think) with scores that are 1250 and 1050 in Cinebench.
Click to expand...

It's the current Titan Xp, the one AMD tested against back in CES.

 
nashathedog said:
That's good news.

We just need to see some gaming results now so we can make better guesses at how the RX performs. :D
Click to expand...

Yeah if you check the video I edited in at 0:46 they show Cinebench OpenGL, and it clearly shows Titan Xp, and it's nearly 20% slower than Vega.

Which is interesting, as AMD sucks at OpenGL performance.
 
Gerard said:
The length of the card is a bit confusing to say the least, one of the points of hbm was to reduce board size with all the memory being basically tight to the gpu. So why is this card the same length as a titan x? Have to wonder whats going on underneath the heatsink for it to be back to regular card sizes even with hbm. Suppose its possible its just that length to accommodate a regular heatsink without a lot of it hanging off the end like some iterations of the fury card.

Be interesting to see teardowns of the card to see whats going on there.

Nice to see they kept the gpu tachometer on it, hopefully it works better than the one on furyx.
Click to expand...
It could well be related to the required heatsink size for sufficient heat dissipation.
 
DragonQ said:
It could well be related to the required heatsink size for sufficient heat dissipation.
Click to expand...

They might even bring out another Nano and the selling point will be the much smaller cooler. HBM offers plenty of room for different sizes on cards this powerful. For me i would want a full size cooler for the extra cooling potential. I don't think any air cooled Fury non X came with the small shroud but it would have been easily done judging by the Nano which most users were pretty happy with.
 
N19h7m4r3 said:
Yeah if you check the video I edited in at 0:46 they show Cinebench OpenGL, and it clearly shows Titan Xp, and it's nearly 20% slower than Vega.

Which is interesting, as AMD sucks at OpenGL performance.
Click to expand...

It is likely using a completely new OpenGL driver. They are also likely optimising for OpenGL methods that are Nvidia usage heavy, since in OpenGL there are practically tens of API calls for doing the same thing. But its a case where not every api call for doing the same thing is properly supported.
 
Last edited:
Mauller said:
I expect it to be able to go higher, considering Polaris 20XTX can hit 1600 without modifications to the architecture for higher clocks.
Click to expand...

I agree, it makes no sense for the Watercooled version to need up to 75W more, and cost 600 more, if it's not going to have a decent increase in performance.
 
My main concern with vega now is the price, given that pretty much every GPU has gone up in price, even nvidia's...... 1080+ performance for £400 isn't exactly looking possible now :(
 
Mauller said:
I expect it to be able to go higher, considering Polaris 20XTX can hit 1600 without modifications to the architecture for higher clocks.
Click to expand...
This will depend on how good the architecture is. They might already be pushing the clocks to be close to Nvidia or it might be like 7970/50 where they went pretty low on the clocks so plenty of headroom. My guess is they are pushing the clocks and OC headroom will be pretty limited. Power usage looks to be pretty high out of the box which leads me to belive they are already near the max. Hopefully i am wrong but everything points to pushing clocks from the get go.
 
AMD are really dancing around the Driver questions in the PC World video.
The only person that mentions Professional Drivers for the Vega card is the PC World chap; and AMD only mention that the TitanXp doesn't use Professional drivers.

A US specialized vendor for workstation has this blurp in regards in to Frontier Edition as well.
sdQ8OdcuTuuZ-dTTFfbyBg.png


That once again explains why the little Radeon Pro WX 7100 is only around 5% slower than Frontier Edition in SpecViewPerf tests, while costing only $660.

I'm not sure where the FE sits really. You could probably buy the Radeon RX Vega AND WX 7100 for the price of the Frontier Edition, or a little less.
 
TheRealDeal said:
This will depend on how good the architecture is. They might already be pushing the clocks to be close to Nvidia or it might be like 7970/50 where they went pretty low on the clocks so plenty of headroom. My guess is they are pushing the clocks and OC headroom will be pretty limited. Power usage looks to be pretty high out of the box which leads me to belive they are already near the max. Hopefully i am wrong but everything points to pushing clocks from the get go.
Click to expand...

Yeah I agree, although the architecture this time is supposed to be modified for higher clocks, unlike Polaris which only gained higher clocks through Node improvements alone.
 
Status
Not open for further replies.
Back
Top Bottom