ZEN When?
Images are sometimes not shown due to bandwidth/network limitations. Refreshing the page usually helps.
You are currently reading a thread in /g/ - Technology
You are currently reading a thread in /g/ - Technology
When do you think Zen will be out in fully fledged consumer version and will you buy one?
Will AMD get into the high performing segment of the market again?
>>51742723
There's rumors of Zen-compatible motherboards coming in march.
no, i built my lmaolake pc when it was released. thank god i did bought it on release tho, the prices right now are retarded. £370 for the 6700k, i paid like £270 the day it came out.
i do hope zen is a competitive product, and we see something with more ipc than at least a 2500k lol
Q2-Q3 2016.
And no I won't buy one. I have no issue with AMD, and I hope Zen is great to create some competition for Intel. But I have a 4790k and will have no real reason to upgrade for a while.
>>51742783
Amd should maintain the more cores stint too.
Dual core Pentium competitor
Tri core i3 competitor with ht
Quad core i5 competitor with ht
Hex core i7 competitor with ht.
As well as some 6 cores without ht around i5 prices too.
If their ipc will only be a little behind Intel, having more cores will put their processors pretty far ahead.
I'd buy a tri core zen at i3 prices.
>>51742723
>When do you think Zen will be out in fully fledged consumer version and will you buy one?
Mid-2016. Not buying into the hype, I have to see to believe. I have hopes though.
If it's a good deal and proves to be solid (~a month after release) I may get myself one.
>Will AMD get into the high performing segment of the market again?
Not immediately but this is their entry ticket, maybe after a year if all things go well.
>>51742896
>Dual core
>2016-23days
no.
>>51743079
So you don't think a budget dual core cpu should be sold?
So they should start at i5 quad core and sell nothing below that?
You're a fucking moron.
>>51743468
>So they should start at i5 quad core and sell nothing below that?
Not him but that's a bit of a strawman. Both Intel and AMD have their low power, low performance cores. Imagine buying an atom based celeron on a desktop, sounds horrific.
>>51743468
>>51743553
Given as a rule moar coars is a good thing and how far ahead on the integrated graphics front AMD is just selling low power apus is a far better option than some shit-tier dual core chip.
>>51743553
I'm actually thinking of picking up an atom.
Throw in an ssd and you have a decent low powered computer you can leave running overnight for large downloads.
It'll ease my electricity bill somewhat as opposed to my overclocked i7 870 and gtx 670.
Also good for general browsing and shitposting, and much cheaper than a laptop.
Can pick up as arock j1900m micro atx for aud 50. Refurbished. Was originally considering the athlon 5350 but prices went up, fuck the Australian dollar ;_;
>>51743079
You're a fucking moron if you think budget dual core processor are going anywhere any time soon.
A g3240 is enough for most people who just want to browse the web, study, and just regular computer shit.
>>51743597
I'm halfway inclined to agree with you but it doesn't work all the way. You need single core performance, even now. The kitty cat cores may be more than acceptable for their TDP in the extreme low range, they're not something you'd want in a desktop if you can avoid it.
I'd rather a dual core kaveri module running at 3.5GHz than a quad core beema module running at 2.4GHz. Even with the same IGP on both, the beefier core of kaveri would be way better for an end user.
>>51743597
It's not just about what's better, you also have to consider what's enough.
For the average user who doesn't do anything overly demanding, a budget dual core processor is plenty.
>>51743631
>Was originally considering the athlon 5350 but prices went up, fuck the Australian dollar ;_;
Well, at least you've the option of buying it. In my country the shops stock niche things like that until they sell off the first batch and that's it. They won't even order it.
You know you could buy a laptop motherboard with a soldered on CPU. I'm eyeing up some of the old kaveri ones and it should be very possible to just throw it in a case and enjoy a 19 watt thin client with very decent performance. Shouldn't surprise me if you found quite a few core i series available that way, for very little money.
>>51742896
That would be neat. I've seen at least one person claim that Zen cores are produced in groups of four with no hardware capability to disable cores though which could leave AMD without much flexibility.
Still I'm hoping that we see them able to disable cores. Even if the 'groups of four' rumor is true we'll at the very least have the possibility of seeing a 12 core chip.
>>51744025
>with no hardware capability to disable cores though
That could potentially be a first for quite some time. Unless I see a first party source I'm calling bullshit on this, very loudly.
>Daily Zen thread until October 2016
The joy
>>51743751
I was considering the laptop cpu mobo deal, could get them for a little more.
But the atom would still use less power and be plenty fast with an ssd.
$50 for the atom, or $75 for a c2d laptop mobo+cpu.
I saw an i3 for $99 though, but meh, it's only $50 for the atom.
I'd love to snag a VIA but no one sells them here, only ones I've found are horribly outdated ones on ebay.
>>51744126
Yeah, I really hope its bullshit.
A pricing structure like this would be great.
$100-$130 3 core zen
$130-$160 4 core zen
$160-$190 6 core zen
$190-$250 8 core zen
>>51743468
>and sell nothing below that
i5's should be 6cores
i3's should be 4cores
i7's should be 8cores
You seem to have convinced yourself that Dual Core processors are the only option. You probably like to watch your partner and a black man fuck too.
>>51742783
Fucking this, £350 upward for the 6700K. I got a 5820K on sale for £290 instead.
>>51742723
Summit Ridge will have at least a paper launch before Q4 2016. A slide deck showed availability starting mid way into Q3.
>>51742742
Thats not a rumor, and its not relevant.
The desktop version of Carrizo called Bristol Ridge is on socket AM4, and its launching Q2 2016. There'll be AM4 boards on the market for a long time before any Zen based Summit Ridge chips are available.
I will always buy Intel™ because I only play games with Intel Inside™. Intel also pioneers innovative new technologies like Hyper Threading Technology™, Intel Rapid Start Technology™ and the highest quality chipsets to ever grace motherboards.
When I boot up with a brand new Intel™ i7™ with the latest Z chipset, I can enjoy the games the way they where meant to be with Intel Inside™. Intel™ also delivers a far more silkysmooth experience with its Hyper Threading Technology™.
Intel i5™ is also very power efficient. A processor is the most power hungry device in your house. Air conditioners, water heaters, lights, etc all use less power than a processor. Which is why Intel™ puts gamers first by ensuring that their gaming experience is of the highest quality while looking out for gamers by giving them the most value in their electrical bill.
At this point in time, there's really no reasons to consider an AMD processor at all. I tried once, it caused so much heat that it exploded and nearly burnt down my house. It also consumed so much power that it produced an EMP and destroyed not only the rest of my computer but my entire neighborhood.
Intel™ also pioneered how useless MORE CORES is with the i™ series processors. Years ago, everyone thought MORE CORES were the future. Now, Intel™ has debunked that myth entirely and increased efficiency. Now you can save thousands a year in electricity thanks to Intel™ with its powerful IPC. MORE CORES will never be part of Intel's™ line up.
It's quite clear that OPs an AMD shill trying to convince you to settle on something less than the optimal experience with Intel Inside™. Intel™ is the only real way to play games. We have seen recently that they offer incredible libraries for software developers like Intel C++ Compiler. He is probably too poor to afford the Intel Inside™ experience and can not afford to play any games.
Don't be a poor gamer with bad chipsets and a huge power bills. Play games with Intel Inside™
>>51742723
>will you buy one?
Yes, if it offers a decent performance boost over my 4770K.
AMD has been disappointing for years already.
>watch for bulldozer!
>wait for APUs!
>wait for ZEN!
They make chips that are 2/3 as fast as Intels, and they can only compete by being half the price.
I feel that Zen will be no different.
>>51744639
So, no more pentiums?
You're a fucking moron.
>>51744639
Why force a more expensive quad core on people when a dual core is plenty powerful for them?
>>51745316
You probably won't be getting one then.
Odds are it will end up around the per core performance of Sandy Bridge or a touch better.
Will Zen's die be soldered to heatspreader? Does AMD solder or does it use thermal paste? Intel went full jew after Sandy Bridge.
>>51748303
AMD has been doing both on and off. Some chips have had TIM under the heat spreader, their most recent Godavari chips are soldered.
>>51748336
I could really consider purchasing Zen if they'd solder it.
>>51742723
If it comes close to Haswell in performance and supports ECC in consumer versions, I'll just immediately.
>>51742723
yes im buying zen , no i don't care as long as its better than bulldozer and vishera.
seriously i just don't care I just want AMD with m.2 support and some decent black and green mobo for the MATX class of cases.
as for getting in to the high proformace segment its hard to say but if its flagship is better than the 9590 then yes, there is hope.
for me as long as it does what I want i don't care.
>>51745365
If their performance improvement chart is anything to go by it means zen will be at least as fast as haswell per clock
>>51744216
you can get whole 2nd hand thinkpad x200/x300's with duo2cuos for like $50 AUD, less if they have an easily fixed issue like fan error boot failures or a BIOS password
>>51745801
It doesn't have to be expensive. AMD has been selling cheap quad cores for 5 years now.
If Zen is even 85% of haswell performance then yes, I'll be buying their too model and good stout overclocking motherboard without hesitation. I had an fx-9370 for the longest time until the sabertooth motherboard it was in started dying on me. There were no decent am3+ boards readily available so I went out and bought z97 and an i5 chip. Felt like a traitor doing it. At the same time I'm not going to buy into another bulldozer disaster. They had better bring everything they got to the table or the only thing I'll be buying from them is video cards. Even that it's up on the site looking at his hard furyX failed. Overclockers dream my ass.
>>51749759
And dual cores are selling even cheaper.
>>51749829
Fury x may not be overclockers dream, but I wouldn't call it a fail.
With new drivers it outperforms 980 ti at 4k, and trades blows at 1440p.
It's also better in crossfire than the 980 ti is in sli
>>51749829
Sabertooths have a 5 year warranty, rma that shit
I'm excited to see what they'll do with the Raven Ridge APUs due out Q1 2017.
I hope they have HBM on package.
I'm waiting for the Opterons. I switched from a pair of Opteron 6128's to a Xeon E5-1620 and currently using a Xeon E5-2660.
>>51744261
What about 1 core+HT Zen for $50?
>>51751347
That would be a hilarious SKU, but I doubt they'll ever offer less than a dual core.
>>51749998
>proprietary benchmarks
For what purpose?
>>51751597
Do your own open source benches then
>>51744261
Budget 60 dollar dual core zen?
bought a lmaowell 4690k and going to keep it until zen comes out, if zen's performance is 30% faster, i will replace, if not, will wait for another intel release probably (3-4years)
>>51752356
wanted to add, my next cpu will be 8 physical cores, unless new 4core version is ~50% faster single core speed
>>51742723
>Will AMD get into the high performing segment of the market again?
they better hope so.
My FX-8350 build will last me another two or three years, at which point Zen will definitely be out. Hyperthreaded eight core Zen would be a hell of an upgrade.
>>51752371
gl m8, i'll wait with my e6600 till amd is 139% faster than xenon 48238 @ 0.4GHz
>>51752420
4690k at stock is faster than e6600, 104% to 300% depends what computing task is used
>>51745365
But APUs are great. You easily save $200 by getting the A10-7870K over an equivalent Intel CPU + NVidia GPU and this is not counting the lower energy bills.
>>51752945
For the money involed AMD's apus have absolutely zero competition from Intel. If you go for a system without a dedicated gpu you go AMD every time.
Inb4 iris pro
>>51751347
>>51752192
I expect that the bottom end will continue to be filled by cat cores and bulldozer derivatives for a while. After that it will probably be Zen APUs with the graphics hardware disabled.
I also wouldn't be surprised at all if what we get with Zen this year is only the higher end chips with all of the more mainstream chips coming as APUs in 2017.
>>51753078
>inb4 the one thing APUs can't compete against
Fuck off.
>>51753193
The iri pro is only on chips that cost many, many times those APU. If you are buying an iris pro chip you have specialised needs because honestly anything the iris pro can do a dedicated gpu can do better.
>>51753094
>Zen this year is only the higher end chips
Really, i don't mind, i would sell my intel i5 4440 + 280x rig and update to AMD.
Fuck, i forgot about emulators, let's hope that it won't suck much in them.
>>51753289
With the IPC improvements it should be right up there with Intel chips in emulators. Hopefully we also see emulators as early adopters of Vulkan and the new version of OpenCL. That could really help their CPU bottleneck.
>>51744135
Where else am I supposed to wait?
>>51753378
I stoll dont get this emulator maymay. Using an FX 8320 and both when its at stock and at its current OC (4.2ghz, hardly an OC), i get perfect emulation w dolphin (ssbbrawl 50 fps constant).
My buddy has an i5 3570k and ssbb runs the same on his rig.
Zen isn't coming out til fucking Q4 2016:
http://wccftech.com/amd-zen-launch-q4-2016/
>>51753434
It is amusing how Intel users are declaring Zen dead on arrival because it isn't arriving RIGHT NOW GODDAMNIT despite the fact said intel users were never going to upgrade anyway.
>i5 6400
>2.7 ghz
what
>>51753739
I guess they thought 4460 offered too much for the money. It's still a better buy than 6400.
That's what happens when you don't have competition, you are free to do whatever.
>>51753432
Apparently Haswell has some architectural bits that allow it to be far better than its predecessors at emulation.
>>51753739
That is only the idle power. You have to look at the speed with turbo active.
>>51742723
I already bought a x99 system that should last me another 4-5 years like my previous FX system did. I see no reason to upgrade from x99.
>>51742723
They said q4 2016
Zen is too little to late
>>51754671
Skylake revival in q3 2016 will crush zen
>>51743604
>>51743608
>Zen hive mind
>>51748520
To be fair you're not asking for much.
even a 30% per core average IPC improvement would mean that Zen only needs to clock to 3.4Ghz to be just as quick as a 4.8Ghz 8350/9370
AMD is saying Zen is 40% faster than Excavator, Excavator is about 8% faster than Piledriver.
That means, purely mathematically, Zen at 2.5Ghz is as powerful as Piledriver at 4.8 (4.8 * .52 = 2.495)
>>51754671
this has been my feelings as well.
Zen at most will match haswell. which is great and all. Haswell provides great performance.
But intel already has skylake out. And intel plans on releasing skylake refresh, "kaby lake", at the same time zen is post to be released.
Kaby lake may indeed match skylake's performance since its just a refresh, but it will bring a newer chipset along with some minor changes to benefit overclockers. Much like devils canyon did with haswell.
Ignoring the fact that skylake brought a boost over haswells performance clock for clock. Might not be much, but ultimately, why pay $300+ for a zen thats slower than its intel counterpart for the same price?
by the time zen hits amd once again is two and a half (due to kaby being a refresh) generations behind. amd is already two generations behind if they released zen at this point.
yes i could see zen being somewhat of a success if they released a true 8 core zen for $350 to complete with the 6700K but Intel already has a answer to that, its called x99, haswell-e and soon to be, broadwell-e. Zen will match haswell-e single core performance and core count, but will be behind broadwell-e in single core performance. Broadwell-e which too will becoming out around that time frame.
I can't see zen causing a mass, or even a minor exodus of intel haswell and above users to switch to amd. All it might do is give something to amd fanboys to stay happy and cause a decent portion of sandy bridge and ivy users to come over simply because they want a change in face. But nothing to revive the athlon 64 glory days. which is something amd needs to stay relevant.
>>51754788
skylake right now already crushes zen in pure single threaded performance.
all zen will have to counter it is core count if they decided to go with core counts. a 8 core zen would be a better buy than a 6700k. but if amd goes 2, 4, and tops off with 6 cores, then skylake will stay extremely competitive and be the better choice for most users.
ignore the fact of haswell-e which already provides you with that zen performance and core count. let alone broadwell-e which is coming out next year which will have the core count and faster performance.
>>51754927
Hell I'd do it. An unlocked 8C/16t CPU that gets the same performance as a 5960x at half of the price? I'd be all over it. It'd make a fine upgrade from my current E5-2690 (8c/16t SB-E Xeon).
There's also the fact AMD has some really really good power management IP that they developed for Excavator and Steamroller that ended up coming with spitting distance of intel's far more energy efficient design on a superior node. They throw that into Zen on 14nm and it will end up being an energy efficient beast of a CPU.
>>51755327
> They throw that into Zen on 14nm and it will end up being an energy efficient beast of a CPU.
I suspect that is going to be a major element for Zen and everything based off of it. It is especially important for cracking the laptop market (which intel's jewry still has AMD locked out of) and laptops are one of the big money spinners for Intel.
>>51755408
AMD can wait on the laptops. Its the servers and data centers they need to get into. Lots of well performing, energy efficient cores in one package (lets say, a theoretical socket G35) plus the rumor that they may stick a beefy on-die GPU for HSA stuff, and that would end up being a VERY powerful piece of silicon.
There could ultimately be 2 classes of high end Zen chips: 1 is a pure CPU setup, 32C/64T in the same package, and the other is a server APU, with half the cores and a big fat GPU for workloads that do well on them.
Either way, I'm looking forwards to what AMD will pump out, provided they dont go under first.
>>51755592
> 1 is a pure CPU setup, 32C/64T
On a single die? Imagine a dual cpu board with two of those bolted on.
>>51747543
That wouldn't necessarily be bad at all. My 4770K is dogshit, it barely hits 4.3 at ridiculous voltages, if Zen is a good clocker (hits 4.8GHz+ on a custom water loop) and comes with 6 or 8 cores, that could still be quite a nice upgrade for me.
>>51755720
I never said same die, i said same package. Big difference. Amd's pushing their luck as it is with the nearly 600mm^2 Fury die. It'd be cheaper for AMD to do what they've been doing with their higher end G34 CPUs, and put 2 dies in the same package.
For the pure CPU package, that'd be a pair of 16C dies connected either through the package or an interposer die on the package.
For the APU, it'd be a 16C CPU die w/ a GPU die (maybe some HBM for a local high speed cache) either connected through the package or through an interposer (likely an interposer).
I'm not expecting AMD to pull a >600mm^2 monster chip for their high end SKUs. Not even intel is ballsy enough to do that these days, and during the Nehalem era their largest chips were knocking on the 700mm^2 door.
>>51755592
>the rumor that they may stick a beefy on-die GPU for HSA stuff
a full-blown GPU with all the TMU/ROP/tesselation/etc. fixed function units, or just the shader FP ALU array?
>>51756259
Well, considering its from the Fudzilla slide, it could be either. What i see conceivably happening though is a full shader array, but the other hardware is either cut down to save on die space and power consumption, or omitted entirely.
The slide in question: http://www.extremetech.com/wp-content/uploads/2015/04/AMD-Zen.jpg
>>51756415
that slide shows HBM controllers, which would be largely wasted without a bunch of ROP/TMU units, which are by far the biggest bandwidth hogs on a GPU.
furthermore, it says "multimedia engine", which implies all the codec crap as well.
in conclusion, somebody pulled this out of this ass, and nobody really has any idea yet.
>>51756527
Yep, its a rumor, nothing more. Its a given though that AMD is going to make an attempt to get back into the server market, and that a dual chip pure CPU design is quite possible considering they've done it in the socket G34 based chips.
If they do get a chunk of the server market back though with Zen, it'd be a huge boon for them as the servers and HPC markets are where all the cash is that isnt in the mobile markets.
What I honestly want (although likely wont get) is for AMD to allow their server Zen chips to be overclockable in the right board not too dissimilar to what was going on with LGA 1366 and the EVGA SR2. That would be awesome as hell.
I'll probably build like a lower mid range mATX zen system for fun.
>>51754441
Still lower.
>>51753078
Who cares r7 250 is trash. I'm not impressed by a shit cpu and shit graphics stuck together for the price of their equivalents.
>>51757634
You sound salty.
>>51744261
So a 5960x equivalent for $200. Stay delusional.
>>51757781
Lets say - for the hypothetical discussion - such a chip does exist, would it really change anything?
>>51757677
I am mad a pentium and an r9 270 cost $15 more than that bullshit and you get triple the gpu power.
>>51757852
> pentium
