![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
no where ooops... but shit did I have to dig to find where you quoted that from
The CPU team demo'd using the Titan Pascal when they compared to Intel and used Battlefield 1 max settings at 4k for the test, was the very last demo presented by Lisa that used Vega and the unreleased next Star Wars Battlefront game.Excuse me ?!?!? At the Zen event you saw Vega running BF1 !?!?!?!?
imagining or misleading, take your pick.Correct !
But some people here are imagining things... like Vega running BF1 side by side with the competition... and "looking" slower...
) ...but that is not enough for me to forget those words from Joe Macri...), words that got me excited........ a few weeks later when the reviews came out was in fury because how pissed I was...
You can call me a AMD "fanboy" ( I would not be offended
"...we built this for the overclockers, for the gamers... we got a 6 fase, 400 amps of power... I mean it's amazing... you gonna be able to overclock this thing like no tomorrow with the cooling solution... this is an overclockers dream..."
Words that I will never forget (well maybe I will if Vega surprises me
What pissed me off was not the fact that the Fury X basically did not overclocked... what pissed me off was the delusional misleading shameless lie !
And that's why I am still rocking my 290X twins and never bought a couple of Fury X.
I think you meant 56 ROP on the GTX 970.
Guess you're still stuck in timewave zero? Have you ever bought and sold a high end AMD card? I have and got pretty damn good dosh back selling at the right time multiple times. No worse or different than if I'd gone nvidia. At the moment AMD are not really competitive in high end with price drops, so of course the prices will be low.
Lol.. well, here's the roadmap people seem to forget from 2014;
Basically - "Where is the beef?"
I will suggest that question will be asked tomorrow, a lot.
I think that the successful volume of moderately price GDDR5X had a lot to do with it as well. And conversely, issues with supply and demand on HBM after AMD went all in.
The Failure tech lab video late q3 last year they stated they were sampling/testing 10nm. Nothing taped out for production, don't know why that FUD comment early in this thread got so many 'likes'..
Guess you're still stuck in timewave zero? Have you ever bought and sold a high end AMD card? I have and got pretty damn good dosh back selling at the right time multiple times. No worse or different than if I'd gone nvidia. At the moment AMD are not really competitive in high end with price drops, so of course the prices will be low.
Lol.. well, here's the roadmap people seem to forget from 2014;
If AMD cards were so uncompetitive and useless, then why did they push Volta back and go for Paxwell? It's pretty clear by spacing they were expecting 2016. Just like the AMD Vega roadmap, 2017. Which seems to be staying in place funnily enough.
And I've seen in a few recent benches the ancient, thermal beast Fury-X sticking with 1070-1080 and on a fraction of the VRAM. Not all the time but it can happen. Not bad for an ancient old design, however it's large die vs node shrink...
That dual thing they made when the 290X2 was kicking ass. I remember seeing it at 3k lol, GTX titan duo-Z black rape edition. Oh, no doubt the fanboys will scream 'TEH CUDA COREZ', as if any of the gamers had any use of them.
And people saying you can't gouge on 'luxury' goods.. fucking hell. You can gouge on anything. GPUs are not bought as traditional luxury goods, they are primarily functional. Is a television a luxury good too now? They can cost much more and can look good sitting off. What about a budget car, is that also a 'luxury good' because it costs more than a GPU? Fucks sake, every slimey damn excuse in the book to paint Nvidia as the faultless white knight.
I remember this data, it was shown by a few here to be dubious the many times it's been bought up. From memory the sample rate was far higher for Nvidia, which skews things already, plus the aforementioned usage of each card and missing sample models. I've had failures from both sides. But more from Nvidia if you count the doomed 8800s. Even I have mined on my 7970 for a while, running it at 1.3GHz (nearly 290X speed) 1.25VDDC, along with heavy stints of Crysis2@1440/60fps. The 7970 failed around the same time as a failing PSU while I was overseas, jury is still out but I'm prepared to take blame for electromigration. Thing is the 8800GTX was BGA failure, not abuse, was stock and hardly ever loaded as it was used for VJ production. Another friend also had one die. TLDR: There have been more than a few hardware issues on both sides of the fence in the last 10-15 years. It happens, just like car manufacturers and the rest have issues from time to time.
To say one brand is significantly more failure prone than another is BS, it's bad for business and is corrected quickly. They both drop the ball from time to time. I remember the Flounder Edition and its missing FETs, plus other BS they pulled. It was a cut rate card built to the absolute minimum quality passable as 'premiurrum', sold to gouge the fuck out of everyone and fuck their AIB partners.
Oh yeah, while we're at it, as it's a pertinent issue to AMD marketing fluff this time around - what about the huge, global motherboard failure from the RX480 and its huge current draw? I can't even see the sunsets due to the fires! Some of you guys hyped the shit out of that and blew it way out of proportion. Just like WCCF tech riding the hype train dick for everything - we all know they will bury that Vega 4x claim into the ground, like a dog with a bone.
We saw the same crap with No Mans Sky. Sure it should've been more polished upon release, but the majority of the zombies following it hyped the fucking shit out of it, taking things out of context. I was following it for a year or more prior and saw this happen as we got close to the first 'release date'. A decent group of us sat back, played it and enjoyed it for what it is, which was almost entirely what was promised with very few exceptions. The patches made it better. I guess that's a little like AMD drivers in past - they matured with age.
edit: Enough bitching - a solution you ask? I am involved in many ways including sales/marketing in a high end orientated company, just like our products, we simply over deliver and avoid any tech hype marketing with figures that can be speculated out of control, like we see every damn time a new card comes out.
Keep it to a minimum until it's out, then let the minimum specs out with minimum info needed. Give more detailed info to those who are interested or ask. Over deliver to the public.
AMD could benefit from a similar approach when they truly do have something worth hyping (but not hyping out of proportion!). They were doing so well to keep Vega quiet until recently and would have done well to not release any of those figures we saw yesterday 'leak'.
Basically, AMD - please stop being caught out on a 'swamp gas and venus intersecting gives you 4x perf/watt' BS claim. I'm almost calling it now, it'll be some ridiculous justification. Even in the early days, I can't remember a node giving close to 4x perf/watt in real world use. Perhaps it'll be 'under VR in x Engine with AMD ultra quadocto-projection® and mirroring an output to a 2nd screen' etc.... weird incredibly uncommon scenario.
Well, I sold my 980Ti. Already own a Freesync monitor and waiting for Vega to arrive, 1st day buyer. Zen so far not interesting to me, but if it drags 8 core 16 thread prices down, who would complain?
I won't complain, but it's way too late from my perspective. I was considering an upgrade in early 2016, had to wait for mid 2016 in the end. Waiting another 9 months to a year with uncertain results wasn't really an appealing proposition.
And as I said the Tegra releases much later than the dGPU it is based upon, along with commitments to HPC projects that require the hardware to start to be implented in 2017.
