@waltc3 and how many cpus arcs did amd release since A64? how many of those equalled intel in performance? right. they cried wolf so many times, im done believing anything from amd, until i've seen guru/thg and others testing a zen. @all even if it really performs really good (ipc), still doesnt mean we will have a "mature" platform as i could get (most likely will) right now from intel (5820k/6800k), where i know i won't have any "headaches" in the next 4-5 month, because of new and untested hardware/drivers.. and since there wasn't an accidental leak, i doubt its a "performer", but a "budget" cpu (don't mean that in a bad way). we had almost complete specs around june/july for a cpu released 3 month later (FX83xx). Where is all the detailed (official amd) info on zen (besides core count on the big one), if its sooo good? (clocks/voltage/chipset etc), since release is in less than 2 month. and my 3770k is my first intel rig ever (amd since 2001), so don't start with the fanboy crap.

My only concern is that motherboard manufacturers will cheap out on parts. Then we would end up with boards that will blow up and AMD will fully take the blame even though it would not be their fault. (Or the CPU will throttle like made and make it look like their CPU's are slow and once again, AMD would take the blame.):bang:

And this is why you should always buy a motherboard based on overall build quality and power delivery. There's plenty of companies which boast superior build quality on certain boards, from Gigabyte (at least they used to) to ASRock. There's always going to be a derp that buys a potato quality part and blames the issues on something else. If you buy an Asus Rampage Hero Dildo Edition board with extra sparkles that has a 4 phase power delivery system made out of that's on you. Speaking of which, what's up with Asus and always having far less power delivery phases than every other manufacturer?

Speaking of which, what's up with Asus and always having far less power delivery phases than every other manufacturer?

if its 2 p less than the usual max (8+2/2), but better quality on the components, you wont notice that on pwr delivery. at least i dont remember a review having trouble ocing 58xx/69x on asus boards. http://sinhardware.com/index.php/vrm-list seems the latest asrock x99 (gaming/overclocking series) are better than (most) asus...

While AMD is still tring to make CPUs to compete with i7 ( and xeon ? ) some of us have LONG bought Intel ES 14nm (and 22nm) xeon 14/20core chips for 300-600usd and have made that money up x100. If we waited for AMD we would be homeless now. For everything else that needs single thread performance just use and i5-6600k overclocked or even that i3. I hear you... ES chips ? Yeah, why not. ES chips. AMD reminds me of Pentax. While Sony, Nikon and Canon had all released full frames cameras, pentax didn't release their FF until last year, by the time all their loyal fans have left. It's no wonder not many pros use Pentax, although it does offer more value.

Only problem with this is that intel Xeons with 20 cores runs on very low frequency, are hardly overclockable and is too expensive. Thing is, many of us enthusisasts could use more then 4 cores but dont need dozens of them. Advantage with 6/8 cores is that they have almost as good singlethread performance and at same time double of cores compared to quad core. Ant to be honest, quad cores now mainstream in both mobiles and notebook and is there for like 10 years with close to zero performance increase in last few intel CPU generations, its freaking time to get 6-8 core to mainstream desktop and I mean NOW and not in 2019.

@fry178 well you could have said the same by Z87 or Z97 release, manufacturers needed few months to iron bios out, saw the same with Z170.. That said, "complaining" about not being good by release is a moot point imo. Speaking of crappy VRM, well if you buy a cheap mobo of course it will be cheaply build too.. That's why I bought this mobo with 16phase back then, I wanted to be as stable as possible and of curse to have best OC as possible, and now I see I made the right choice back then. Yeah it was 100$ more but hey quality is $$. 😀

From overclock.net............. Zen 8/16 @ 3.2 ghz- $499 Zen 8/16 @ 3.0 ghz- $349 Zen 6/12 @ ?- $249 Zen 4/8 @ ?- $149 Those are some killer prices to be honest on all of them, I just hope they are good cant wait.

wow that 4core is ultra cheap, almost too cheap, if at 4.2 - 4.4ghz it will be serious LGA115x competitor and probably sell like hot cakes too lol its basically a intel i7 4core, 6core looks mighty too for 250€ if ~3.7-3.9ghz

If those prices and specs are real, and the chips can OC to 4.0GHz+ with IPC equivalent to Broadwell-E, then Intel's current lineup with the current prices is obsolete overnight. If.

what are the PCIe lanes count? Any exact numbers? So far all I've read was 2x 16x Gen 3 and additional lanes with PLX chip on mobo..

32 on the CPU itself and a DMI bus to connect to the motherboard PHYs.

Looks like every model will have SMT Then is much posible... 4c 8t 6c 12t 8c 16t Since AMD said that Summit Ridge has 2x Orochi on multi threaded (using all threads) i spected same performance per clock on all threaded operations but now we know Zen's its lower clocks then this says higher performance per clock (on 2 threaded per core scenarios), i spect monothreaded IPC between Haswell and Broadwell, perfect since virtually no have graphics bottleneck Other thought, if every model has SMT can't be comcompared in this way... Sr3 (8t) vs i3 Sr5 (12t) vs i5 Sr7 (16t) vs i7 Since even the "slower model" looks "i7 level" performance but having higher Power compsumition per Core and slower clocks they can't compete at all, every initial model and could be... Sr3 sounds 45-65w Sr5 sounds 65-80w and Sr7 sounds +95w Maybe the Sr3 and Sr5 not get over 3.45GHz on Max Turbo initially and 3.15GHz base to make people buy more Sr7 models

Looks like every model will have SMT Then is much posible... 4c 8t 6c 12t 8c 16t Since AMD said that Summit Ridge has 2x Orochi on multi threaded (using all threads) i spected same performance per clock on all threaded operations but now we know Zen's its lower clocks then this says higher performance per clock (on 2 threaded per core scenarios), i spect monothreaded IPC between Haswell and Broadwell, perfect since virtually no have graphics bottleneck Other thought, if every model has SMT can't be comcompared in this way... Sr3 (8t) vs i3 Sr5 (12t) vs i5 Sr7 (16t) vs i7 Since even the "slower model" looks "i7 level" performance but having higher Power compsumition per Core and slower clocks they can't compete at all, every initial model and could be... Sr3 sounds 45-65w Sr5 sounds 65-80w and Sr7 sounds +95w Maybe the Sr3 and Sr5 not get over 3.45GHz on Max Turbo initially and 3.15GHz base to make people buy more Sr7 models

If the 3.5GHz turbo is correct about the 8/16 part, then it actually turbos higher than the equivalent Intel CPU.

if its 2 p less than the usual max (8+2/2), but better quality on the components, you wont notice that on pwr delivery. at least i dont remember a review having trouble ocing 58xx/69x on asus boards. http://sinhardware.com/index.php/vrm-list seems the latest asrock x99 (gaming/overclocking series) are better than (most) asus...

Thanks for the link, much appreciated. I have to say I'm a bit disappointed by the stats on the ASRock boards. But I suppose it doesn't matter, it doesn't make a difference what board I use, this CPU is a lemon compared to review samples, I doubt any board can get even another 100MHz out of it.