1
Amd's Vega Flagship Video Card To Arrive In October?
Started by Rykiel, May 12 2016 08:25 PM
13 replies to this topic
#1
Posted 12 May 2016 - 08:25 PM
http://videocardz.co...ward-to-october
Keep in mind that it IS VideoCardz (well-known clickbait site), so take it with a BIG grain of salt, and (of course) YMMV, but if true, Q4 this year/Q1 2017 might just be the "sweet spot" in terms of a new video card purchase from either camp ...
Keep in mind that it IS VideoCardz (well-known clickbait site), so take it with a BIG grain of salt, and (of course) YMMV, but if true, Q4 this year/Q1 2017 might just be the "sweet spot" in terms of a new video card purchase from either camp ...
#2
Posted 13 May 2016 - 07:26 AM
They're too late. Again.
AMD has a node advantage technically, 14nm for Polaris and Vega vs 16nm for Pascal, though I'm not sure how much of an actual advantage that will turn out to be. The main advantage would be slightly less power usage, but AMD was severely handicapped in that department already so it might only make up for lost ground. We'll have to see.
My thought here is that they will launch their big guns and it will either barely do better or be at parity with the 1080, and Nvidia will rain on their parade a few months down the road with an HBM2 1080Ti card.
The best case scenario for AMD is that Vega is an absolute monster and really takes advantage of HBM/HBM2, in which case Nvidia will play catch-up with an HBM2 1080Ti.
AMD has a node advantage technically, 14nm for Polaris and Vega vs 16nm for Pascal, though I'm not sure how much of an actual advantage that will turn out to be. The main advantage would be slightly less power usage, but AMD was severely handicapped in that department already so it might only make up for lost ground. We'll have to see.
My thought here is that they will launch their big guns and it will either barely do better or be at parity with the 1080, and Nvidia will rain on their parade a few months down the road with an HBM2 1080Ti card.
The best case scenario for AMD is that Vega is an absolute monster and really takes advantage of HBM/HBM2, in which case Nvidia will play catch-up with an HBM2 1080Ti.
#4
#5
Posted 13 May 2016 - 07:48 PM
According the AMD their TDP for Polaris and Vega was supposed to be equal to or below the current offerings from Intel and NVidia. But that was just "what AMD says" and there is no data to support the claims yet.
However, Pascal seems to be a major jump in performance compared to Maxwell. From the benchmarks the $599 MSRP 1080 is supposed to put the Titan X to shame, and the 1070 may as outperform as well. Makes me wish I didn't buy my 970. But on a brighter note, 970 prices are already dropping.
http://arstechnica.c...icing-revealed/
However, Pascal seems to be a major jump in performance compared to Maxwell. From the benchmarks the $599 MSRP 1080 is supposed to put the Titan X to shame, and the 1070 may as outperform as well. Makes me wish I didn't buy my 970. But on a brighter note, 970 prices are already dropping.
http://arstechnica.c...icing-revealed/
#6
Posted 14 May 2016 - 11:12 AM
Goose, on 13 May 2016 - 12:09 PM, said:
That was debunked -- AMD actually released a statement clarifying the situation: http://www.eteknix.c...gative-rumours/
#7
Posted 14 May 2016 - 11:17 AM
xWiredx, on 13 May 2016 - 07:26 AM, said:
AMD has a node advantage technically, 14nm for Polaris and Vega vs 16nm for Pascal,
The best case scenario for AMD is that Vega is an absolute monster and really takes advantage of HBM/HBM2, in which case Nvidia will play catch-up with an HBM2 1080Ti.
The best case scenario for AMD is that Vega is an absolute monster and really takes advantage of HBM/HBM2, in which case Nvidia will play catch-up with an HBM2 1080Ti.
My understanding is that the 16nm process that NVIDIA (via TSMC) is using is actually slightly superior to the 14nm process of AMD (via Samsung --> Global Foundries).
But yes, if this rumor's true, AMD was caught off-guard by just how good Pascal was, and is hoping to snatch a win in the high-end market with Vega before Big Polaris hits.
#8
Posted 14 May 2016 - 12:02 PM
Rykiel, on 14 May 2016 - 11:17 AM, said:
My understanding is that the 16nm process that NVIDIA (via TSMC) is using is actually slightly superior to the 14nm process of AMD (via Samsung --> Global Foundries).
Maybe, maybe not.
Nvidia did manage some ridiculous clocks on the pascal gpu's. Amd will probably go for a higher transistor density and lower clocks.
#9
Posted 14 May 2016 - 12:05 PM
oh man i cant wait for pascal to come out! i imagine the 9 series dropping in price!
#10
Posted 14 May 2016 - 04:24 PM
Aleksandr Sergeyevich Kerensky, on 14 May 2016 - 12:05 PM, said:
oh man i cant wait for pascal to come out! i imagine the 9 series dropping in price!
What does everyone think they're gon'a see? 40% off? 60?
Turns out I know quite a few people whom just found a burning desire to get a new card, and they all have a bad time with the concept of keeping their powder dry 'till we find out just how impressive the 1060 and 1060Ti turn out.
#13
Posted 25 May 2016 - 07:40 PM
These so-called "benchmarks" for Polaris (via VideoCardz) make me wonder what Vega's performance will be -- big win in the price/performance dept, or the performance crown-winner, or a big bust? Mobile parts, or desktop (or both)? It's all in AMD's ballpark right now.
Benchmark link: http://videocardz.co...rk11-benchmarks
Benchmark link: http://videocardz.co...rk11-benchmarks
#14
Posted 30 June 2016 - 05:19 PM
October/school purchase season "release date for Vega" rumors still floating around in the wake of the RX 480 release. Not sure what's to happen, but there is a big hole between the RX 480 and the GTX 1080 that needs to be filled up. Possibly Vega 11, with Vega 10 being the HBM2 user? Or is AMD simply going to cut Fury/Nano/Fury X pricing until Vega hits the scene?
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users