Originally Posted by Mookster
I'm not particularly moved by the fact that you've been benchmarking AMD cards since 2011.
And, no, those are not what I'd call criteria for an educated guess. In fact, that "It's AMD" attitude is exactly what resulted in my "antagonistic" response.
Did you know whether it would be produced on GloFlo's, TSMC's, or Samsungs node? No. Did you know the differences between those nodes before the release? No. Do you now? No, no one does. Did you know the die size before release? No. Did you know anything substantial about this release? No, you didn't.
You really summed it up for me. You based your opinion on Fury X's launch and "it's AMD". It doesn't matter if you're right or wrong, those are just bad reasons to hold a strong opinion either way. You have nothing to be sitting here bragging about. All this "I called it" talk is weak, and your "high rank" at HWBOT doesn't change that. Die size, fab process, and provider are just as relevant to the performance of the final product as AMD's ability to design the chip. Architecture performance has been so close between NV and AMD, outside of driver enhancements and proprietary garbage, that those few factors matter more than almost anything else these days.
So, no. I wouldn't say my attitude is any more "antagonistic" than your opinion. Your statements warrant a strong response so I gave you one.
Your judgment about whether or not my opinion is educated is irrelevant, and so is the specifics of how I arrived at my conclusion, because I was still right. Just because my assumptions don't make sense to you, or aren't "in-depth" enough to you, doesn't discredit them or my way of thinking. My conclusion was still mostly accurate.
And actually, contrary to your assumptions, I actually did know that it would be produced on a combined GloFlo/Samsung 14nm node. I also knew the die size would be roughly 232mm2, this has been known for some time, surely it was less than 250mm2, and it's obviously a midrange chip with a smaller die. This has been common knowledge for months. Do I know the differences? Does it matter? I was still correct. More importantly, I've stated in other posts that I believe that the weak performance of the card are due to it having less ROPs and TMUs- which I also expected. The Fury X only had 64 ROPs while the 980ti had 96, and on release I knew this was a reason for it's poor performance at 1080p. So, I figured that the RX 480/Polaris would likely have 32 ROPs and probably underperform compared to the 390/290.
"It being AMD" is actually a valid reason to be suspicious, I thought everyone knew about Bulldozer at this point
We also get tons of delays for the 14nm node, the Fury and Fury X were turds, the company itself is failing and so on. Personally, I don't know how you can even hold this against me when I've stated I've used nothing but AMD graphics cards since 2008. I'm certainly an AMD supporter, and not an Nvidia troll or something. I'm not blind to the facts though and the companies' history, so obviously I have caution and am diffident in my expectations.
I don't know how these are "bad reasons", I'm sorry I'm not an electrical engineer at a fabrication facility, christ some people think you can't have an opinion or make statements unless you have a degree in this stuff.
The point about HWBOT is simple- where's your results? Where's your cred? If you know so much, then do something
and prove you're actually good at this- building computers- instead of being full of hot air. I've seen your rigs, they're seriously outdated and they look like crap
So good luck on this front.
With that sir, have a good day. You will remain in my blocklist and receive no further responses.