Man of Honour
As a side note. Look how much higher Nvidia put Pascal than Maxwell.... yeah right! lol
That slide is showing bandwidth available on professional cards using NVLink.
-
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
- Thread starter Doom112
- Start date
- Status
That slide is showing bandwidth available on professional cards using NVLink.
Funny you say that as I just put mine up for sale also
Would not call it dumping though, it is a great card but I just purchased a Fury Nano from AMDMatt to finally try Freesync out on my monitor
Enjoy that, I have found freesync to be pretty good, would not like to go back.
why wouldn't it compete with volta ? you do realize that pascal and volta are the same node, a revision of 16nm, this is not like the jump from 28nm to 16nm, so you will be left with minimal improvements from the node and maybe little architectural too, if there is any.Realistically. They can't be releasing Vega to compete with Pascal. It just doesn't make sense. It's too late. And seeing as how Pascal is basically just Maxwell+. It wouldn't be impressive at all.
I'm thinking purely from a business point of view AMD must be trying to position themselves here. Thats would be how to make the most money.
all volta with size up to 450mm² will be roughly the same performance as GP102, so if there is any more performance push it will come from die size, so ppl should'nt expect too much for the segments 100-500$ where vega competes, and AMD leaves 700-1000$ segment for Nvidia.
That doesn't make much sense, The P100 has 21.2 TFLOPS FP16 for Mezzanine, so if the Maxwell M40 was still 1:1 FP16 it would be 7 TFLOPs.
That's not x10 faster in FP16.
Soldato
- Joined
- 28 May 2007
- Posts
- 10,148
The 10x faster was a lot of Ceo maths as Jen put it.
10 times more efficient and not 10x faster
Yeah.. even the claim of x2 Peak Single Precision compute was just a lie.
Peak for Tesla M40 = 7 TFLOPS - 250W
Peak for P100 = 10.6 T FLOPS / 9.3 T FLOPS ( PCIe card) - 300W
When you say Nvidia are the “owners of the PC gaming market”, it reminds me of when Americans imply that the USA somehow owns freedom.
The market is worth around $35bn a year, if anyone owned it, it would be Valve and Blizzard et al. Riot still make $150m a month from League of Legends, which was released in 2009. So when you talk about Nvidia “owning” the PC gaming market, I don’t know WTF you’re talking about. Are you referring to the super-niche, high-end, non-workstation, gpu market?
Caporegime
AMD should just forget about all this new tech, and just concentrate on the here and now, like Nvidia, and also go toe to toe with them, bringing out cards around the same time, at around the same performance, or better, not cards that match performance we've already had for years, and surpassed, as thats the only way they'll get their share up, which is what they need, they need to be up with Nvidia on it, or ahead.
Once they do that, then they can concentrate on new tech, otherwise they are just ******* in the wind, just bringing out antiques, isn't going to do it.
Everytime a game comes out for our PCs (barring the odd one, here, there), its got Nvidia slapped on it, so to me, they own the market, our PC games, are done for Nvidia, as its their cards thats are in just about every PC a going, its console games that are done for AMD hardware, but thats a different market.
Unless we get AMDs share up with, and over Nvidias on our PCs, then nothing will change.
Once they do that, then they can concentrate on new tech, otherwise they are just ******* in the wind, just bringing out antiques, isn't going to do it.
Everytime a game comes out for our PCs (barring the odd one, here, there), its got Nvidia slapped on it, so to me, they own the market, our PC games, are done for Nvidia, as its their cards thats are in just about every PC a going, its console games that are done for AMD hardware, but thats a different market.
Unless we get AMDs share up with, and over Nvidias on our PCs, then nothing will change.
Last edited:
Thanks man
Looking forward to test it out. Many older titles I will be revisiting should run 60fps easily on the nano and benefiting from freesync
Soldato
It might well compete with Volta. I don't know. none of us do.
Im ust saying from a business sense that is a sensible output.
Soldato
11,353 posts and we're still none the wiser. This is shaping up to be one huge anti-climax.
Permabanned
vega RX 45% faster that vega FE in gaming...
Caporegime
Someone wasn't paying attention to nVidia's GPU Tech Conference - while mostly concerning itself with the professional cards there are loads of changes over Pascal:
https://devblogs.nvidia.com/parallelforall/inside-volta/
12FF is going into risk production next week I believe so if nVidia is planning on bringing consumer Volta to GeForce on a normal kind of timescale it will likely show up sooner rather than later - this also gives a potential 10% performance increase or 25% power reduction but through a mixture of area reduction and other techniques some designs can attain potentially upto 33% performance increase.
I strongly disagree.
Here and now was what got them to this situation, where even people far from tech know they don't want an AMD CPU or GPU for their gaming PC.
AMD tried here and now improvements with bulldozer and with iterations of GCN. SandyBridge happened and they had no answer. Maxwell happened and Fiji flopped. Pascal happened and they had no answer.
The long term picture presented on financial day makes much more sense. After 2-3 years in development they are now in position to compete everywhere and for a long while.
Associate
I said the difference between ((medium) and (high or ultra)), not between high and ultra
Associate
I am not sure that we can decide here and how and what AMD should do. In my eyes, you have to offer a product first and then if you want to innovate, you can. If all you have is a different product you can either be extremely successful if it's amazing or fall flat on your face if it isn't. There's nothing for AMD to fall back to if they fail with Vega. Not long now, hopefully, before we know for sure.
The opposite I think - Bulldozer they tried pushing more but slower cores trying to push an envisioned future where things were more parallel and less dependant on IPC, etc. and it just didn't happen while Intel concentrated on a CPU that primarily could run existing software very fast and secondary had an eye to the future - like with these new features on Vega they seem to just assume by putting them out there people will start using them and historically it hasn't proven to be the case in the majority of cases - same with GCN while nVidia concentrated on what worked best with DX11 AMD tried pushing an architecture more suited to advanced DX12 use but no one took up DX12 type APIs in any kind of hurry.
I hope I'm wrong but Vega is starting to look kind of a turkey with many of its potential bleeding edge improvements that could have really made it shine increasingly appearing to having some substantial caveats to actual real world leveraging of those features.
- Status