Zen is supposed to have +40% IPC more than Carrizo, which for most scenarios it would bring it to Haswell IPC, and in some scenarios it would bring it higher than that. I could also see some parts having integrated HBM, destroying everything Intel has to offer (unless they make up their mind and have eDram on Broadwell EP, like the desktop parts have).

In other news: AMD's "design wins" point to new console hardware in 2016 CEO Lisa T. Su has teased "three design wins" for the division responsible for those products. Su told the company's investors that the new products are expected to bring in lifetime revenue of $1.5 billion, spread over three to four years, starting in the second-half of calendar 2016. AMD's descriptions certainly fit with the rumours around the Nintendo NX, which has no fixed release date but is widely expected to launch later this year. Rumours of a new, more powerful version of the PlayStation 4 emerged last month, and the alleged system specifications that followed did contain AMD technology. There are no specific rumours regarding a new Xbox console, but it's worth noting that Phil Spencer used the company's Spring Showcase to describe the possibility of Xbox launching, "new hardware capability during a generation."

To the "First off", both are OC results. Neither is close to top OC as I post only things achievable by masses. - Then WTF!?! is that your comparison based on? - > Did you really stated that 2439 ST / 12937 MT is best case scenario 8-core FX chip can do? FX-8350 has best case 3000+ ST / 16000+ MT (Not that many will reach that) - > And even then FX-8350 costs less than 1/2 of i7-6700k (Some of us can actually do comparison based on price relevance. Instead on basing everything on ePen) The Rest of your comparison... I must say HH did not do i7-4770k in that review any justice. He used Stock 3.5/3.9GHz i7-4770k and it looked naturally weaker pitted against 4.0/4.2GHz base i7-6700k. And then only i7-6700k got OC results. I am not saying it OCs badly or worse than i7-4770k. But best of each is not saying that it is that drastic upgrade over i7-4770k. http://browser.primatelabs.com/geekbench3/compare/2941440?baseline=6165912 And finally, post was to show that there are still many users who decided not to upgrade for many reason like performance per $ of new system versus keeping their old system at no additional cost. And later part is to show how much AMD improved CPU cores since Bulldozer/Piledriver. Hell, here you can see worst ever measured score for Carrizo based Athlon X4 845 vs best ever measurement of Bulldozer FX-4100. Think about clock.

Both OCed? I see the 6600k is stock. Intel Core i5-6600K @ 3.50 GHz 1 processor, 4 cores vs AMD FX-8350 @ 4.44 GHz 1 processor, 4 cores, 8 threads And why would we ever compare best case scenarios? Look at this: http://browser.primatelabs.com/geekbench3/search?q=fx-9590 Many of these are even under 14k MT. 'Best scenarios' is a moot point. You can have a golden sample. Wait, what's the deal there? Is the frequency not recognized correctly? I doubt a 6700k has such a high fluctuation, look. http://browser.primatelabs.com/geekbench3/search?dir=desc&page=3&q=i7-6700k&sort=created_at I see a few have a low 12k while the highest is almost 28k. Regarding to your 4770k point, Skylake surely doesn't offer a lot in terms of clock-per-clock performance compared to broadwell but the whole point was the overclockability. Haswell seemed to be aimed to the mobile market since its inception, considering its low idle and low-load power consumption. Not to mention the integrated voltage regulator which increases temperature. This limits overclocking ability significantly. Skylake is the exact opposite. And I understand your clock-per-clock conception but it doesn't always pan out the way we imagine it. Think Pentium 4 vs Core 2. Core 2 made Pentium 4 look like a complete joke with its per clock performance but that doesn't mean it was easy bringing them to high frequencies. In fact, the most popular mainstream models ran at around a measly 2GHz. They still beat the **** out of the Pentiums tho so maybe this case isn't that relevant. My point is that in order to really be competitive (aka, competitive at the top-end), AMD needs to match not only Skylake's IPC, but also frequency. Which might actually prove difficult considering how slow Zen's development has been going on. I have doubts regarding AMD's capability in making efficient chips. Probably because I've seen way too many inefficient chips.

^ You can be such a dummy sometimes. But we will get there one day, right? geekbench is famous for their ability to read only stock clocks in so many cases. But you have already all cards in your hands. You brought them with you. Not that you needed more than your own link with i7-6700k benchmark showing ST geekbench scores on different clocks. - showing that 4.0GHz gives ST score: 3887 - showing that 4.9GHz gives ST score: 4586 What's again ST score in comparison I posted in #10? OC Skylake vs random OC FX-8350 system What? 4695 ST score? No way, that's not OC, right? I love the way you ask to confirm obvious thing 2 times. Sometimes three times. ...When it only takes few seconds of your attention to realize, what's what.

Hopefully but the ability for intel to counter is pretty easy, even if AMD will win most of the benchmarks, it will take years of that success before many will shift to AMD. I know i will stay with mine cpu even if zen is good as they make it to be, no reason to upgrade, plus PC games are stale and very bad in the recent month, all those upgrades don't justify themselves anymore.

Hopefully but the ability for intel to counter is pretty easy, even if AMD will win most of the benchmarks, it will take years of that success before many will shift to AMD. I know i will stay with mine cpu even if zen is good as they make it to be, no reason to upgrade, plus PC games are stale and very bad in the recent month, all those upgrades don't justify themselves anymore.

Thing is... Do we need AMD to dominate any market? No, we need AMD to be profitable again, to make even better technologies. And for that they need to gain like 10 more % in GPU/CPU market, 5% in server market. That's very doable.

^ You can be such a dummy sometimes. But we will get there one day, right? geekbench is famous for their ability to read only stock clocks in so many cases. But you have already all cards in your hands. You brought them with you. Not that you needed more than your own link with i7-6700k benchmark showing ST geekbench scores on different clocks. - showing that 4.0GHz gives ST score: 3887 - showing that 4.9GHz gives ST score: 4586 What's again ST score in comparison I posted in #10? OC Skylake vs random OC FX-8350 system What? 4695 ST score? No way, that's not OC, right? I love the way you ask to confirm obvious thing 2 times. Sometimes three times. ...When it only takes few seconds of your attention to realize, what's what.

Yeah I didn't pay attention as I am not familiar with geekbench. That gives you a good reason to be this salty? Once again you proved to me that it's worthless trying to have a decent conversation with you. Oh and.. You completely failed to see the point I was making. You have a good day now.

Thing is... Do we need AMD to dominate any market? No, we need AMD to be profitable again, to make even better technologies. And for that they need to gain like 10 more % in GPU/CPU market, 5% in server market. That's very doable.

hope so, nvidia gets very cocky when left unchecked, intel does not speed up CPU power, although seriously they probably cursing themselves for the success of sandy bridge, most of SB users won't upgrade due to cpu but only due to old tech, even so very few will do it. Yeah we need AMD back in the game, i wanna see Polaris and vega murder pascal.

hope so, nvidia gets very cocky when left unchecked, intel does not speed up CPU power, although seriously they probably cursing themselves for the success of sandy bridge, most of SB users won't upgrade due to cpu but only due to old tech, even so very few will do it. Yeah we need AMD back in the game, i wanna see Polaris and vega murder pascal.

I do not think that Polaris can murder Pascal. We know, that Pascal can clock at least as high as it did on 28nm (I presume it can do higher clock than 1400 base). We do not know how well will Polaris clock on 14nm. That's biggest variable. @xIcarus: It is very rare to see you engage in conversation with me. What you usually do is accusation of false information provided, being wrong, ... You do it to other members around too, so It looks like as your style. Unfortunately when you engage in that way, you are wrong party in 9 out of 10 cases. And as result you waste other people time.

Quote "•Total debt at the end of the quarter was $2.24 billion" Gulp! At least the total debt is stable atm. We should all hope that AMD does really well this year, but, to get to profitability? Can't see it happening for at least 3yrs at least and that's with strong products and sales year after year. Intel and Nvidia have capital to burn, they're going to come out all guns blazing if there's even a hint of them losing market-share. This is going to be a battle of survival for AMD.

I'm just excited to see real high end parts without that waste of a igpu that don't cost an arm and a leg.

@xIcarus: It is very rare to see you engage in conversation with me. What you usually do is accusation of false information provided, being wrong, ... You do it to other members around too, so It looks like as your style. Unfortunately when you engage in that way, you are wrong party in 9 out of 10 cases. And as result you waste other people time.

Hey it's a good thing you're keeping tabs on me. Yeah not really, we both know you pulled those remarks out of your a$. I really don't get why you're such a hateful individual. Anyway, I won't pollute this topic any further regarding your petty attacks about what I did or didn't do. Good day.