When will AM4 motherboards come on the market to prep for AMD Zen?
Right after Computex. Bristol Ridge and Stoney Ridge parts are releasing. Budget AM4 boards will be available shortly.
Will the chipsets even be compatible with Zen?
>>54808268
yes?
>>54808265
It's also possible that we only get mobile Bristol Ridge now and the desktop Bristol Ridge at the end of the year together with Zen.
Word has it (iirc from thermalright) that AM4 will use the same mounting points for cpu coolers that have been in use since AM3. Rather handy if you ask me as it means a mountain of coolers (especially budget and top tier ones) will work on zen based chips.
>>54808534
I think AMD's cooler mount has remain unchanged since AM2
>>54808532
Thats not happening. Bristol Ridge is nothing special, and inventory has been building for a very long time. They're launching that gap filler to hold over to 2017 when Raven Ridge finally releases.
>>54808534
They made a facebook post saying that the mounts had changed, then it was quickly deleted.
>>54808534
AM2, AM2+, AM3, and AM3+ all have the same mounting for CPU coolers.
>>54808557
>>54808627
Even better then as it reduces the entry cost for zen if neither consumers or cooler manufacturers need to replace their coolers for their shiny new zen cpu.
zen will probably get cancelled since amd is finished
>>54808673
k bud
>>54808673
true story
>>54808557
Older. The clamp method to the bracket has not changed since Socket 754. I could take a clamp-on 754 heatsink, and use it on every single socket to date, and vice versa.
The 2x2 bolt through method has remained unchanged in function since AM2, although there have been changes to the brackets as of FM1
Reminder that 2016 is almost halfway over, and the release of Summit Ridge is soon upon us.
>>54810054
>non-square die
is AMD retarded?
>>54810150
You clearly are.
>>54810054
>south bridge on the die
When the fuck did this happen? Also, why even have it at that point?
>>54808811
2x2 was the same on most 939 boards too.
It wasn't 'standard' though, the clamps were and iirc some shitty mother implemented the standard clamps without the 2x2 holes but by their own proprietary hole pattern and plastic piece.
>>54810283
with carrizo.
>>54810283
Boards with tertiary chipsets can support more IO, and a set base level is always supported by whats on die. External chipsets are no longer necessary for board vendors to provide. The on die south bridge also draws less power than any external counter part would.
>>54810150
Thuban was pretty rectangular.
I wonder how spectacularly AMD has to fuck up this time
like, Intel doesnt even care about making fast processors anymore, just the whole efficiency meme.
>>54811755
https://en.wikipedia.org/wiki/Jim_Keller_(engineer) I don't think the designer of Zen has ever had a flop that's launched to the public.
>>54811837
He didn't design Zen, he lead the team. He managed the entire oversight of Zen and K12.
Zen's lead designer is a qt blonde lady named Suzanne.
>>54811837
I just want a proper cpu upgrade
fuck intel and their yearly 10% performance boost with slightly less wattage
>>54811995
Thats literally the best they can manage. Diminishing returns isn't a joke.
>>54812139
people will see the writing on the wall if AMD hits the same issue in the coming years. We might even see silicon alternatives from other parties, too!
>>54812207
They will hit that same wall likely with Zen+ or its immediate successor. AMD will start churning out 5% nominal IPC uplifts, the majority of transistors added will be spent on power management and tertiary features to pad marketing material. Substrate material has nothing to do with the fundamental issue at hand.
The last 30 years has been nothing but an accelerating focus on one paradigm of IC design. Our approach to logic is the same regardless of ISA. Instructions are fetched, executed, and retired in the same manner. The throughput of a core is always going to be limited by what you can actually feed to the ALUs, and the separation of memory and execution provides a finite wall to theoretical performance.
The closest thing to a real answer is IBM's Synapse, but who knows if that will ever see the light of day outside of a lab.
>>54808534
>>54808557
oh my God if this is true
i was worried there for a second that I had to buy an upgrade kit for 10 bucks for my existing CPU cooler
>>54808268
uhh.. are you retarded?
>>54810150
>implying you're 0.1% as smart as jim motherfucking keller
what do you think is jim kellers IQ?
do you think he shitposts in /g/ from time to time?
>>54813210
Probably in the range of 140, maybe higher.
He spends all of his free time on his yacht. Doubt he ever shitposts on someplace as worthless as /g/, he probably ghost writes for seeking alpha and manipulates stocks for luls.
>>54813180
certified shit-wrecker
>>54811837
Ditched AMD when he realized there was no hope.
>>54814119
he actually only par-takes in team leading of design and then leaves every company.
everything he's done.. everything, has succeeded extremely well.
>>54813111
its a ligitimate question, i know bios updates needed to happen for amd cpus on some of the older boards, and then you have current boards that support the 6 "core" amd cpus, but can also fit a 8 core 9000 in it but cant function and will burn up.
>>54810054
will it be good
will it be cheap
will it try to match intel price for price or will it be the value we all hope it to be
those are the only questions i have
>>54815653
Thats different. As far as AMD has indicated. AM4 will be, for the first time in a long while, not backwards compatible with previous sockets. AM4 will not be an upgraded platform for Vishera or Bulldozer, but a new platform for Zen only.
As far as we know. But it would be beyond retarded if AMD let their mobo partners ship boards for Zen with bios that isn't compatible. Either way, my bet is that the answer to >>54808268 is a solid yes. Its not an upgrade for AM3, its a new platform.
One of the things I'm keenly interested in is if they've done the same as intel and moved most of the Northbridge functions into the CPU die. I've never seen an FM2+ board in my life, is there a Northbridge on those?
>>54815749
yea, am4 will be new for new shit only, but that doesn't mean zen will work on the first run boards when zen still isn't out.
>>54815686
These are the only important questions.
Don't accept any answers from anyone until Zen is actually out.
I've only ever owned Intel CPUs but I was thinking of making an AMD build when the new stuff comes out.
>>54815886
well, performance wise its at least sandybridge so that is checked off for me.
however, the more amd talks about it, the more they say without saying anything. "the first time amd has been better then its competitors sense i joined the company" or something along those lines is from an investor meeting, something they cant lie to without getting sued.
its honestly starting to sound like they perform at or better then skylake, and this worries me. i was expecting lesser single core performance but making up for it with cores and threading, but if they are close to if not on par, perhaps surpassing, will they be cheap?
ill buy one for 500$ or less, but if it goes close to what intel charges, i can't pay it, i just cant justify spending 1000$ on a cpu.
>>54810150
This is skylake cpu die
are you retarded?
>yes
>>54810054
>2016
>still using the fucking south bridge
>>54815818
You know as much as me, friend.
>>54815989
they equivocate because they don't want to explicitly suggest their processors are superior. so rest easy I guess, they'll be suitably shit for your price bracket.
