It's assumption but it's AMD saying it not me,kind of like Intel saying IB will have 10x the igp of SB while faking a video of it.
proof is in the pudding, but for those that are already convinced that AMD will fail or Intel can't be beat
proof isn't enough to open their eyes they can't see reality.
Thats the thing. Its AMD saying it. Even Intel cannot be trusted 100%. I think IB will push about 10-15% vs SB, but thats because thats what has been seen in the past with a die shrink, plus Intel is still using the same base arch since Core.
The reason I don't see Trinity doing anything amazing is because its based on PD which in order to be completley amazing would have to not only rework the arch completley, but gain at least 30% in performance to match Intel.
And its people with attitudes like yours, who trusted everything AMD stated pre-BD release instead of looking at available info. I stated many times before BD was released that the closer we got to launch, the more realistic the numbers from the "rumors" would be. It always is like that. With Intel, AMD any of them. Of course everyone with the mind set then set that all aside and were even more dissapointed, or just left the forums all together.
As for IB GPU, I have never seen Intel claim it doing 10x that of SB. But it does look, based on the improvements (double the EUs, improved performance of each EU) that it will at least be near double the performance of HD3K. Of course it will still not be what I ever use, except maybe a HTPC, but I am also more interested in QS 2.0. I do also remember a bit ago that you stated there is no reason why QS should be faster than anything AMD could come up with. We again have proof as even CUDA could not outperform QS and QS 2.0 will push higher quality while being 2x faster.
The main difference between Intel and AMD is R&D. As I have said before, Intel spends more money on R&D than a lot of companies put together. They start research for new technologies years before its release, such as the Tri Gates which started in 2001 or HK/MG which also started about 10 years before its release.
AMD doesn't have the same R&D budget and for process they now will rely on whatever another company comes out with. They may not get Tri Gates until 14nm which by then, Intel might have another trick up their sleeve.
Its all about logic. Intel has seen AMDs strong side so they brought their A game. Now AMD needs to bring theirs and I don't think their new strategy will do it. That is personal opinion.
He should be able to show some links then right?
Yes. I agree. But the thing is that its just a rumor. Most of the stuff I could post has no way of being proven 100% correct, hence the "rumor" part I added. I always take rumors with a large grain of salt. Even though I buy AMD GPUs, I wasn't sure the new arch would perform that great. It does, and of course AMD throws the "arm and a leg" price tag since it does outperform a GTX580.
Just remember this, most slids are marketing. There is a reason why the slides at first have no real benchmarks, then move to synthetic then finally close to launch only include select benchmarks or scenarios that make them look better. They will always cherry pick and thats the way to sell a product. I can't even tell you how many people bought a AM3+ mobo just to move to BD then see its results and either move to Intel, get a Phenom II or stick with what they had.
I say we will see. Thats because we will. I honestly don't think Trinity will do anything major. It will replace Llano so AMD will have only BD based systems and they can fully can Stars. I think its a mistake, like others, and think they should shrink that down to 32nm for consumers and leave BD in the HPC/Server market instead.
Facebook
Twitter
Instagram