Anyone else here think Zen will be an amazing long term investment?

No, maybe the next iteration in two years will be better, but ryzen didn't have what it takes... I hate that fact, but that's how things are

Unfortunately not. What Ryzen is to AMD is like what Nehalem was to Intel.

Not to say Nehalem was bad, but there where some issues that needed to be ironed out. (Unironically the memory issues)

Then Sandy Bridge came along. I expect Ryzen 2 so be what Sandy Bridge was to Intel.

>>59425264
Lol
>Cpus
>Long term "investment"

2/10

Saged and neck yoself

>>59425444
Somewhat agree, I'm happy with ryzen, but if your looking long term, they're expecting another 12-15% ipc gain on zen2 which comes out next year, you also don't have to be a beta tester that way, and can purchase a more refined product. I'm still buying zen, but I'm ok with the occasional upgrade, and the socket will be the same for zen2/3.

Remember in 2013 when everyone said "Bulldozer will get better guys! Multithreading is the future! Devs will optimize games! Your PDF reader will be multithreaded! You music player will use more cores! It's the future!"

And look how that turned out, faggot.
Multithreading is a pain in the ass to program, and it's only suited for certain problems and applications. This is why we still need higher instructions per clock.

The only AMD products that have appreciated are their GPUs for two reasons:
1) DX12 takes advantage of asynchronous compute, which AMD architectures have in hardware
2) Not having annual GPU releases means they need to keep selling older architectures, meaning those cards need active driver updates to sell. Jewvidia releases new shit frequently and stops allocating resources to develop drivers for architectures older than about 2 generations. Remember how Kepler started faring worse and worse with new titles?

I had a Titan Black and benchmarked it before and after about 20 driver revisions. Some compute performance went up, some went down. Jewvidia had new, expensive chips to sell.

"Optimization" is also a meme that gets thrown around. Do you understand what that means, to optimize for a specific architecture? The only thing I can think of that would be holding Ryzen back at the moment is cache usage, since it's supposedly different, but updating software to make use of that will likely involve firmware updates, compiler optimization (to make use of Ryzen's instruction set), and other software engineering efforts you or I don't understand. That means: Don't count on it happening any time soon.

>>59425590

>>59425631
do you only run one program at a time?

>>59425264
>>59425444
>>59425621

Same here, I have a 8320 (am an AMD fag that can^'t buy intel) that is clocked to 4.6 GHz for a gaymin' PC that I only play cs:go and CK2 on at home while I use 4930k for my workstation which I do 3d/2d rendering for my company at office. Unfortunately for me there wasn't any good workstation CPUs at the time I was renewing my office hardware and I had to go with intel on that one.

So considering that I don't game much anymore and 4930k is still a beast, I am just going to hope that Zen2 will have 5%~ IPC improvement and capable of reaching 4.5 Ghz on 20% of their chips. Then I will get two Zen2 for myself, an 8 core for a workstation and a 6/8 core for gaymin' and shit.

Computers are becoming a mature market. And the vast majority of the market is laptops which are weaker than most desktop CPUs anyways. Of fucking course you'll be able to use Ryzen in three years, you'll be able to use kaby lake in three years too. Browsing the web and typing in Word aren't very intensive tasks anymore. Most applications are designed with laptops in mind too, so your desktop CPU that has a much higher power draw to enable more performance, will be good for a long time.

If you have any other use case scenario, you're either a gaymer or your company is probably buying your PC for you.

>>59425724
the gayming pc market is growing and mainstream users are switching to laptops

>>59425607
my 2600k is long term investment.

>>59425724
I'm using Ivy until it dies on me. Getting some serious mileage out of what was a sandy i3 and hd 6000 series. I figure if you can bench whatever the newest locked i5 will do you're still okay barring hardware failure.

>>59425590

Still running 920 Nehalem that I built in late '09 with 24gigs of RAM. Does everything I need.

Maybe I'm just cheap but the last time I thought about upgrading I didn't see a large enough performance increase for the money(some years ago now). That's probably changed by now.

Would seriously like some USB3 though.

I'm happy with my long term investment. Eight years out of a $250 CPU and $200ish mobo.

>>59425607
4770K will last a long time, still

>>59425264
I'm going get 1600x and upgrade to next gen 8core on same board in a year, the fact that you don't need new socket already makes it future proof and much cheaper than intel

>>59425631
AMD releases a new architecture every two years too kek

They just haven't changed the basic layout of 16 wide SIMD arrays in groups of four, while Nvidia keeps fucking with different ratios of control to compute. Their basic shaders are still unchanged since 2012.

>>59425645
fucking retard, you run one game at a time, even if you're streaming it's better to have a 7700k at 5 GHz, and for general desktop use you're not getting anywhere near 100% cpu usage, you run at most one cpu-intensive program at once, anything that's in the background isn't using much cpu

>>59426174
>wanting a 200 watt processor
ok

I'll stick to a nice 65 watts in games and everything being above 60hz anyways.

>>59426193
so you're spending hundreds of dollars on a ryzen "upgrade" just to save 30 watts and to play at 60 hz

File: tmp_8195-14824432604771429544096.jpg (35KB, 385x375px) Image search: [Google]

35KB, 385x375px

>>59425590
> waiting another 8 years for ryzen2

File: ryzen-streaming.png (129KB, 1440x1080px) Image search: [Google]

129KB, 1440x1080px

>>59426174
This

Ryzen is even awful at streaming, it's bad at multi-tasking, it's just shit

AMDtards are just having delusions if they think it's going to somehow get better with time

>>59425621
>they're expecting another 12-15% ipc gain on zen2
Probably more like a 10% IPC gain as the absolute max, but 10%+ higher clock speeds as well.

>>59427175
>bias methodology "benchmark" brought to you in collaboration with Intel's marketing department and ArsTechnica
Try it on the medium preset and 60fps, shill.

File: 1403837987365.png (48KB, 400x389px) Image search: [Google]

48KB, 400x389px

>>59425264
>investment
How much do you expect to profit?

>>59427175
thats a hard pill to swallow


thanks, i guess?

>>59425264

>Anyone else here think Bulldozer will be an amazing long term investment?

>>59427175
but that's a proprietary benchmark.
That tells you nothing of importance of significance.

when is anandslowtech's ryzen gaming benchmarks going to be released so we can put all of this to rest?

>>59425816
i7 960 here, still fast enough

Sandy Bridge will still be good enough in 2020

>>59427585
kek

File: 1489152808881.gif (2MB, 320x240px) Image search: [Google]

2MB, 320x240px

Do you think Ryzen 2 will bring better clocks? Should I buy a high end AM4 board, a cheap 4 core/8 thread ryzen CPU then upgrade to a more matured process?

>>59428553
Go for it, with the additional IPC gains they are expecting it is gonna be a nice platform. Im waiting for zen+ as well.

The second generation version of the r7 1700 will be the next 2500k if I had to guess. They're already promising a +15% performance uplift and that should be primarily ipc or frequentcy since I doubt there is much room left for improvement with their great smt. So we're looking at roughly skylake single thread performance on a 8c 16t device for $330. 2500k stayed good for what like 5 year until it fell considerably behind, I would expect the same out of the r7 2700. The r7 1700 will fall behind faster than that because of its low frequency unless mobo manufacturers pull something out of their ass that let's you pick up another 300 mhz or so

>>59427598
Post a different one then. Most benchmarks probably won't make use of Ryzen well unless they're updated to take advantage of its architectural improvements.
You're not being helpful just shooting down a benchmark and not posting something to counter it.

>>59426171
Never thought about it like that, good point.

AMD has said they are sticking with AM4 through 2020 at least. I'll boy a ryzen now and upgrade on their last AM4 release. Have fun with your changing socket every year, Intel users.

>>59426174

> 7700k can stream Fallout 4 with OBS without dropping frames every 2 seconds

Kek. Thats how I know you are a shill. You cant game on Ultra/High and software stream with any quality at the same time with a 7700k period. Meanwhile I can OBS stream fallout 4 with perfect smoothness and quality at Ultra, and have multiple internet tabs open shitposting at Intelkeks, FOR LESS MONEY.

>>59425264
meh who needs to upgrade anyways
my cpu will hold 5 more years out and it´s not like there´s anything worth doing with a new cpu

>>59425607
It is in this era of tick-tock-shill-shill harder.

>>59426174
>you run one game at a time
what year is it?
https://www.youtube.com/watch?v=Njm0MBOwFTM

>>59425264
I can't vow for this when Zen+/zenver2 is coming within a year with up to a 15% IPC improvement
However that's on AM4, so buying a R5 seems like a great idea
>>59430538
ServeTheHome and Phoronix already did dumb shill

>>59431575
>just wait

>>59431609
>just wait
>buy the R5
Nice reading comprehension dumb shill

>>59431629
>pay less now and then buy Zen+ cos ebin AM4 ending up paying more
>just wait

File: 1467296306101.png (91KB, 653x726px) Image search: [Google]

91KB, 653x726px

But Zen's extra cores and threads are totally worth it! Most ppl don't browse the Web and play games only a-anyway.. Your i5 will be obsolete!!

>>59431648
>upgrading
>paying more
>not selling old parts
wew you just went full retarded

>>59426591
>>59427175
shills

>>59427597
>>59431708
fuck off faggot shill

>>59431751
>>59431761
Stop giving him (you)s you dumbfucks, you're replying to samefag.

>>59427175
Even that fucked-ass benchmark proves Ryzen 7 takes a 5% performance hit from OBS while a 7700K takes a 16% performance hit.

File: pepewojak.gif (3MB, 500x375px) Image search: [Google]

3MB, 500x375px

>>59431744
>R5 with 2 CCX
>blown away by a Core i3.
>worth it
>muh resell value
>I am not a manlet like you and do more than just browse and play!

>>59431802
(you).

>>59431802
(you)

>It's more to it than just gaming and normal day to day usage!
But that's what most ppl do on their computers. An i5 will suffice for another 3-5 years.

>>59431802
(you),

>>59425607

this

fucking autists on this board holy fuck. AMD dick sucking faggot retards.

>>59431854
A 2.6 GHz Conroe C2D with a low end GPU that's still supported suffices as well

>>59428553
Or go Intel and call it a day.

>>59431974
Yep.

>>59431871
fuck off shlomo

>>59431979
or RMA yourself

>>59431979
>just get or new socket dumb goy

>>59425607
What an idiot

>>59425631
>And look how that turned out, faggot.

that all happened?

maybe not your dumb music player.

File: 1466700992477.jpg (33KB, 500x375px) Image search: [Google]

33KB, 500x375px

1.AMD has released a statement that the W10 scheduler is not the problem for the Ryzen

2.AMD has also stated that whether or not their CPUs will gain performance is going to depend on the software coding and optimization


to points 1: It's really fascinating how shills even still deny the reality after the manufacture has released a statement that should have shut up their mouths

to point 2: The majority of people have a 4 core or less CPU running.
Game producers will pander to the majority of course and not to the minority.
So unless Intel is starting to go for more corez the better route, 4 cores with 8 threads will be more than good enough for the years to come.
Keep your wishful thinking shit for yourselves.

>>59426193
>I'll stick to a nice 65 watts
>believes in the TDP

Are you retarded or what?

The 1700 you are thinking of will eat over 200 Watts once you overclock it to 4Ghz on all cores.

Just saying

>>59432145
6900k eats even more when OC'd to 4ghz, so what?

>>59432109
1. AMD said the scheduler is working properly, not that everything is working properly.
2. computerbase could boost performance by simply turning 2 windows 10 features OFF.

>>59431871
Kill yourself.

>>59425607
>hurr durr i love intel's jizz dripping down my chin while i buy 5% performance upgrades and new motherboards every other year at retarded markups

Eat shit and die.

>>59432033
Or stick with your superior i5/i7 and overclock it sky-high.

>>59432109
>The majority of people have a 4 core or less CPU running.
This is due to Intel artifically keeping core/thread count down with their price gouging bullshit. Which will hopefully be coming to an end.

>>59432058
fb2k is heavily threaded in anything that isn't lightweight enough to be done on a P4

>>59432215
After delidding, goy.

File: nothingtosee.png (17KB, 630x210px) Image search: [Google]

17KB, 630x210px

>>59432109
>everything is workling properly

>>59432219
Not this decade, not before their new uArch comes into play.

>>59432240
AMD acknowledged the difference in performance between the default configuration of win10 and win7 but it seems to be because win10 throttles down cores more aggressively than 7. AMD actually told reviewers to turn that functionality off but apparently a lot of them didn't.

>>59432232
No need to friendly RedTeam+ sandnig-member.

>>59432244
Unless it's Tejas: the housefire rises.
>>59432274
>a lot of them didn't
Or Intel told them not to do so.

>>59432286
Here we go again.

>>59432058
fb2k is heavily threaded in anything that isn't lightweight enough to be done on a P4
>>59432145
Except it won't, at most it does 140W when OC'd to 4 GHz on all cores
>>59432215
Just delid voiding your warranty and risking destroying your chip, buy our premium K parts and our ultra expensive boards that allow overclocking
And hope we don't force motherboard manufacturers to disable the feature a year down the line! Dumb goyim
>>59432283
Lrn2english before shilling dumb pajeet

>>59432244
>m-muh secret sauce!
>intel totally has a brand new arch just waiting to release, they've been hiding it all this time!

Uh huh, tell me another one. Maybe about how Coffee Lake will totally be a 15% performance increase.

>>59432283
The fuck is that English, Ramesh?

File: 1483629357673.png (181KB, 293x397px) Image search: [Google]

181KB, 293x397px

>>59432321

>>59432326
No, he means Krzanich literally ordered a brand new microarch.

>>59432364
That's nice. I ordered a sandwich at subway today. Doesn't mean jack shit.

>>59432364
Remember what happened the last 4 times that was done?

>>59425607
my 3570k still does alright, heck even my old q6600 does ok for some stuff.

>>59432394
Yes it does.

>>59432400
Yes. iAPX432, Itanium, NetBurst, Larrabee?

File: 14769305444123.jpg (5KB, 225x225px) Image search: [Google]

5KB, 225x225px

>>59425264
>CPU
>amazing long term investment

>>59428553
Intel are on the third iteration of their 14nm process. If only Ryzan could be made with Intel's fabs.

>>59428553
That's basically what I'm planning...

but this 2+2 core shit is going to be disappointing as fuck.
I bet the i3-7350K is faster in most games. Granted, it costs more and has no upgrade path.

So maybe when Zen2 comes, I can upgrade to a 4 or 6 core that actually works properly on top of having faster clocks. It does seem likely Zen2 will have at least 10% higher clocks as supposedly it's going to be a different process that's more desktop focused.

How can poo in loos justify Ryzen's poor performance in single-thread, low clock speed and lower IPC gaymen applications without bringing out their Jewtel card?
>muh optimized workloads
>it will get better!

>>59432458
Said 2+2 chip has fucking 16mb of L3 so anything can happen.

>>59432466
>low clock speed
>>>>>>>>muh megahurtz
You know these can't be directly compared anyway, right shill?
>gaymen
Oh no, 10 less FPS at >60 FPS, we're doomed.
>single-thread
If you're that desperate for single threaded performance on the cheap, get an i3-7350K.

File: wojakself.png (293KB, 633x758px) Image search: [Google]

293KB, 633x758px

>>59432551
So, you do not even deny Ryzen is simply shit where it counts for most people anymore.

>>59432449
Intel's model of keeping their fabs to themselves is close to being unsustainable, just the fabs eat more than 80% of their R&D, and their majority of employees are fab workers
They will have to open them to third parties, specifically ARM vendors, unless they develop a magic uArch able to so what Atom never could and able to compete with Zen

>>59432510
Doesn't matter when there's milliseconds each frame wasted looping through cache misses because the scheduler keeps moving the thread between CCX
Or because a game is programmed to use a shared L3 cache each frame and has to hit RAM so often.

This shit is sad.

People did tests of the 2+2 and 4+0 configuration for an idea of what performance looked like.
The 4+0 configuration was better than every Intel 4 core on average except for the 7700k.
The 2+2 was pretty much on par with the i5-2500k and i3-7350k.
So duuurrr of course AMD gives us a gimped ass CPU that doesn't properly compete.

No one buys a 4 core for workstation tasks. There's no fucking point of this stupid 2+2 configuration, especially the 1400 with half the L3 on each disabled.

>>59432587
Where the fuck did you read that? Most people care about getting a smooth experience without having to pay an arm and a leg for it. Multiple cores does that. Having 10 more FPS in non-GPU bound VIDYA GAEM does not.

>>59432662
And now watch as plebit/amd goes into full damage control yet again.

>>59432662
>1400 with half the L3 on each disabled
Source?

>>59432718
I think he's just assuming it's like that. It makes more sense for AMD to disable an entire CCX than it does for them to disable part of two of them.

>>59432699
The whole win7/10, CCX infinite fabric, memory issues etc. make it less and less appealing for me to buy an R5 chip while it's inferior the price doesn't make it up for me sadly.

>>59425631
And then this become literally true when the FX8350 surpassed the 2500k, programs will become multithread unless we replace silicon limitations with something else.

>>59432749
Yeah, I'm hoping the 1400 is the saving grace here. The CCX issue really sucks and I'm saying this as someone who plans on upgrading to Ryzen next year. Will probably get a 1700 because fuck it.

>>59432662
That statement isn't an actual statement, the (((journalist))) pulled shit out of his ass

>>59432798
There will be 1xCCX chips because 8 cores doesn't make sense on the lowest end products where they will always be disabled. Too much waste for it to be 2xCCXes for those. You can at least expect the Raven Ridge(APUs) to have 1xCCX and there will probably be some with the iGPU disabled.

>>59432662
You sure do get a lot of (You)'s here

>>59427175
https://www.youtube.com/watch?v=J5ONXgQrdqc

>>59432778
>just wait

File: deus-ex-gaming.jpg (55KB, 848x672px) Image search: [Google]

55KB, 848x672px

>>59432798
You're all getting trolled by known shill Ryan Shrout from PCPer.
Inter CCX communication isn't really a problem. Using a faster memory kit increases the data fabric clock, and that increases performance. Mobo vendors will be pushing out yet another BIOS to improve memory compatibility, and they're all working to increase their list of supported memory kits. AMD partnered with a few memory vendors to sell Ryzen branded kits with high clocks and tight timings to eliminate the headache.

Legit reviews did a sparse review on memory scaling themselves. Going from 2400mhz to a mediocre 3200mhz kit increased performance in a game by nearly 8%. The data fabric clock has a fixed ratio of 1:2 to the memory clock, so faster memory increases inter CCX bandwidth.

>>59427175
>>59432870

So who's lying here?

>>59425444
Next year, not two years from now.

File: good goy.png (604KB, 800x523px) Image search: [Google]

604KB, 800x523px

>>59432941
PCPer is a known shillhub, but why would they show up here?
>>59432949
Do you even need to ask?

>>59432941
>it's alright
>just wait

File: amd-ryzen-ddr4-scaling-latency.jpg (77KB, 849x531px) Image search: [Google]

77KB, 849x531px

>>59432941
DRAM latency from AIDA64 as well.
Ryzen prefers fast memory with tight timings.

>>59432949
Arstechnica used default settings with a low bitrate, settings that would result in an ugly artifact riddled mess that no streamer would make the mistake of using.
Higher bitrate = more CPU load = moar coars perform better.

>>59432981
We've known since launch that the BIOSes are kinda bugged out right now.

>>59432996
And some people have already tested kits and shown 60ns~ DRAM latency.
Pretty much all the FUD about Zen has been completely false.

>>59432949
nobody. launch was disastrous and plagued with issues. I've never such wild fluctuations in hardware reviews before.

it sucks that many people weren't able to show Ryzen at its best but in the end this is AMD's fault. they gave motherboard manufacturers too little time and now it's messing with their big exciting release extremely hard. plus, if even tech journalists can't get optimal results out of those chips, then most consumers won't either.

>>59433015
Or Ryzen is simply flawed by design and nothing can salvage it to such an extent to make it worth buying.

>>59432981
>>59433060
what do you have to gain by being such a partisan cunt? this isn't a video game console we're talking about. if it ends up being good, you benefit. if it doesn't, just buy Intel.

as it stands, based on what we know about the industry and the uarch itself, the "wait and see" theory is plausible. if nothing improves in the next two months you can start shitposting. until then your speculation is not worth more than anybody else's.

depends

from march 2 suddenly its not worth it at all apparently according to the intel shills

you should keep a dual core because its the best vaule according to them

in reality its an 8 core server chip that on march 29-april 5 will get the updated load balancer patch and it will fix one major problem

its certenly a cpu that will be able to keep up for at least 4-5 years

>>59433133
they cant stand that fact that keller for once more shit on their company i even saw today someone saying that a g450 is a better choice for gaming

>>59433146
it's not a fucking server chip. stop trying to invent a scenario in which Ryzen's flaws are acceptable. just watch the fucking story unfold.

>>59425444
Actual ryzen version is more of a nehalem with bugs of a sandy bridge.
It's good but not enough to be a truly long standing platform.

>>59433133
But AMD came out with a statement already and there won't be much more than future games optimized for it.

>>59425631
Its simple, devs have to stop being lazy and develop optimized software, it will have to happen eventually.

>>59433146
>incoming magical patch
Sauce?

>>59432941
I just got this off their site. https://www.pcper.com/reviews/Processors/AMD-Announces-Ryzen-5-Series-CPUs

Update: AMD got back to me this morning about a question I asked them about the makeup of cores, CCX units, and L3 cache. Here is their response.

1600X: 3+3 with 16MB L3 cache. 1600: 3+3 with 16MB L3 cache. 1500X: 2+2 with 16MB L3 cache. 1400: 2+2 with 8MB L3 cache. As with Ryzen 7, each core still has 512KB local L2 cache.

File: Untitled 1.png (95KB, 1922x497px) Image search: [Google]

95KB, 1922x497px

>>59433165
i dont even know what to respond..

the fact that the new bios already is showing its results

the fact that on march 29 ms will release the load balancer patch and it will shit on your parade?

or the fact that you are so tech illitarate that you dont even know basic stuff

File: frustratedwojak.jpg (204KB, 900x973px) Image search: [Google]

204KB, 900x973px

>>59433208

>>59433205
>witchcraft happens

windows 10 ver 17xx currently on the fast ring

>>59425264
>Zen will be an amazing long term investment?
what kind of a fucking idiot thinks computer parts are an investment?
maybe if you're playing the stock market or something..

>>59433180
>won't be much more than future games optimized for it.
They're talking about SMT you nig

>>59433241
ofc its an investment you chinese ape

File: 1488408453536.jpg (138KB, 653x726px) Image search: [Google]

138KB, 653x726px

>>59433208
we need faster memory

File: certified shit wrecker.jpg (13KB, 381x286px) Image search: [Google]

13KB, 381x286px

>AMD will go bankrupt
>AMD will never release Zen
>14nm LPP will never clock over 2.8ghz
>it'll only have Sandy Bridge IPC
>it'll never clock over 3ghz
>I'll draw 200w
>it'll never support high speed DDR4
>it'll only have single channel memory per CCX
>there won't be 32c/64t server chips
>they won't have 8 channel memory
>it'll never cost under $700
>its SMT will never compare to intel's


Over the past 2 years all the antiAMD fanboys have been utterly wrong on pretty much every single point they have shitposted about.
Safe to say Jim Keller delivered.

File: 1486810667607.jpg (66KB, 568x612px) Image search: [Google]

66KB, 568x612px

>>59433271

>>59433161
but he didn't. if anything this is a return to the Core vs Phenom days. slight tech lead in favor of Intel, with AMD having the more feature-complete products. another K8 would have been cool but Zen isn't it. the market isn't what it was back then.

ignore the shitposters and don't succumb to wishful thinking or confirmation bias.

>>59433180
they said nothing was wrong with the Windows 10 CPU scheduler. this leaves game engine optimizations and motherboard firmware updates.

I don't expect Ryzen to catch up with Kaby Lake in terms of vidya FPS but the gap will most likely be bridged to some extent.

>>59433219
Ryzen is not a fucking server chip. it was never marketed as such. it's a consumer fucking product. Naples is the Zen product line for servers.

>>59433271
on that list since march 2
>tdp doesnt matter anymore
>5% difference is HUGE
>intel paid shills pretending that their HEDT doesnt exist
>g4500 is the best cpu
>the only thing that matters is games
>the only resolution /g/ plays is the casual 480p and if the cpu can handle it the 720p
>never talks about netburst and first gen skylakes or x99
>suddenly there is a price/fps on cpus
>>59433298
ofc he did the fact that for over 3 weeks now the shills havent gone to level so insane as to invent new stuff to talk about is pretty much a dead giveaway...

>>59432859
>>59432798
>>59432859
>>59432811
EVENTUALLY there will be 4+0, yes.

With the APUs, they will surely have ones with defective iGPUs that they disable and there is only 1CCX.

But it does not look like any of the R5s launching April 11th are 4+0.

Many, including Anandtech, have said they spoke with AMD and confirmed that all the upcoming 4core models being launched in April are 2+2.

>>59433208

>>59433298
the fuck are you saying

the ccx complex is being built with servers in mind amd doesnt have 2 fucking productions lines

its a fucking server chip regardless of you like it or not

File: 1485527902639.gif (2MB, 350x300px) Image search: [Google]

2MB, 350x300px

>>59433133
Holy shit! Just wait you guys!

>>59433346
The CCX interlink is based on the memory speed and the APUs will need fast memory for the GPU too. Ryzen is designed to take full advantage of fast memory.

>>59433344
fanboys gonna fanboy. Keller brought AMD back into relevance and that's a major fucking feat.

>>59433360
they designed the uarch to fit all purposes

that's why they're using it in CONSUMER chips (Ryzen) as well as SERVER chips (Naples)

>CCX Complex
>CPU Complex Complex
jesus fuck you're retarded

>>59433449
no shit sherlock you know what makes naples more different than ryzen?

the fact that it has all the good cores and nothing more

>>59433477
actually it's got 8-channel memory, registered ECC memory support, extra PCI lanes, etc

stop embarrassing yourself

How can it be the new 'Sandy Bridge', when infact it - arch released to public in 2017 - loses to it in most single threaded operations? (3-10% slower).
Do you actually believe that there will be some magical 'optimizations' for zenver1? Did such thing come for bulldozer archs?

Sandy Bridge on other hand - just like Nehalem before it - had great performance from the launch. And there still isn't anything that beats the living shit out of it in single threaded operations. Have six core SB-e chip and you have chip that is very competitive to zenver1.

>>59433529
a 32 core monster has extra pcie lanes? WHO WOULD HAVE KNEW THAT

in case you dont know its an mcm chip

also https://www.overclock3d.net/news/cpu_mainboard/amd_confirms_that_ryzen_supports_ecc_memory/1

next time you answer please have your facts straight

>>59433608
>Inba cpuboss and some other shitty site.

>Ryzen
>Buyer's remorse
Both go hand in hand like Zen's shitty CCX.

>>59425816
>>59428177

I'm not saying Nehalem was bad, it definitely has its merits. Hell it was the start of an awesome new arch from Intel, and really put intel where it is today.

What I was saying is that Sandy Bridge was just more popular, stable and had/has awesome $/perf ratios.

I'm not a bleeding edge guy either, my last upgrade was to a 4770k with 16gb of ddr3 from an aging dual core Athlon 64 x2 with 8 gb of ddr (I actually still have it, and it does pretty well as a home server).

I'll probably wait until the second iteration on Ryzen 7 to upgrade because I'm shifting more from gaming to workstation use (Machine learning, data science and development), and the issues its been having being a new platform and all are putting me off for the time being. And I'm honestly hoping Vega turns out to be good, as my r9 290 has been treating me wonderfully in all my tasks so far.

>>59433638
Yeah, I have no idea what you're implying, but having 1700x with prime x370-pro, dozen setups with 2500k and Asus x79 deluxe with 3930k, I know how zenver1 compares to sb and in all gaming tests, which only use one or two thread, 1700x@stock loses to 2500k@stock.

I've already posted results too many times. If in doubt, why not do your own with those same games? (Armed assault, Arma2, SHOC, CoP and Lost alpha).

>>59433621
literally from AMD themselves
>ECC is not disabled. It works, but not validated for our consumer client platform.
>consumer client platform

absolutely fucking shattered yourself mate

also I mentioned REGISTERED ECC you dumb shit head

>in case you dont know its an mcm chip
this has nothing to do with anything, you are grasping at straws
>m-maybe if I repeat this acronym I heard he'll be intimidated...

>a 32 core monster has extra pcie lanes? WHO WOULD HAVE KNEW THAT
there is no inherent relation between core count and PCI lane count

they are there to accommodate professional requirements

>>59433702
>Sandy Bridge was just more popular, stable and had/has awesome $/perf ratios.
Nope, it just had the biggest 'IPC' improvement between archs ever and still has, since Zen doesn't deliver that promised 50%+ 'IPC' improvement across the board (single threaded games).

>>59433408
The fuck does that have to do with anything I said?

>>59433790
Zen actually did deliver that 50%+ IPC improvement. Problem is, clock speeds went down.

>>59425673
What is your cinebench singlethread rating?
How fast is your memory? cpu-nb?

>>59425607
Anyone who bought a sandy bridge in 2011 is laughing their balls of at this comment.

>>59425264
>j-just wait, it will get b-better in the future!

Yup, it's an AMD shill with stockholm syndrome, alright.

>>59433823
With that promised '50%+ IPC improvement' (even if counting from Piledriver) it would have more 'IPC' than Sandy Bridge, but clock-to-clock, it loses to Sandy Bridge on all single threaded games.
So it doesn't deliver that across the board. Sandy Bridge on the other hand provided 40% 'IPC' improvement across the board - in each and every single threaded game and software - compared to Nehalem.

>>59433805
As memory speeds improve the fact that a chip is 1 or 2 CCX becomes irrelevant. It's already almost irrelevant, there's hardly any notable difference in most applications to begin with.

>>59433883
wtf are you smoking?
>hurr durr its slower on sandy bridge
>CLOCK TO CLOCK

i dont even..

File: best_cpu_choice.jpg (196KB, 724x1084px) Image search: [Google]

196KB, 724x1084px

>>59425264
ofcourse, you practically future proofed yourself till ddr5 comes around, which will be in around ~5 years

>>59433885
Upping the bclk above 105-107mhz downgrades PCIe3.0 to PCIe2.0

We're getting to the point that GPUs are going to more than fully saturate PCIe2.0 16x, so that doesn't really work AFAIK.

I mean it works, it makes the CPU run faster, but it'll choke higher end graphics cards to come.

>>59433925
[email protected] loses to [email protected] in all single (or dual) threaded games. Example of such games: Armed Assault, Arma2, SHOC, CoP and Lost alpha.

If the statement that zenver1 delivered 50%+ 'IPC' improvement (even if counting from Piledriver) was true on all software, it would be faster than SB at same clocks. But it isn't.

What is it that you don't understand?

File: 45786237892.jpg (75KB, 499x474px) Image search: [Google]

75KB, 499x474px

>>59433976
>can't even beat i5s, let alone i7s, while costing many times as much
>"futureproof"

>>59433999
you're retarded

every single bench says otherwise but keep spewing lies

>haha got ya!

>>59433852
So, what's so great about sandy bridge ?

>>59433999
Because those games see it as a 16 core generally. Bad programming.
Not a fault of the CPU.
But if all you want to play is old games optimized for Sandy Bridge, go ahead.

>>59433999
and how many of those games are actually being developed for ryzen?

its much much more complicated than that...

>>59434060
it's also complete bullshit

File: 5.jpg (2MB, 2304x3504px) Image search: [Google]

2MB, 2304x3504px

>>59432870
>4k streaming
>4k gaming
>cpu of the future

>>59434091
yeap a complete bullshit

OS cant handle it properly
bios are wonky
literally almost no programs out there are developed for it

ITS A COMPLETE FAILURE

#logic

File: total-war-NEW.png (569KB, 600x600px) Image search: [Google]

569KB, 600x600px

>>59434051
Games released before (and not receiving updates after, except for Arma2) SB was released optimized for SB? Goddamn you're delusional.

>>59434159
Huh? Total Warhammer came out long after Sandy Bridge. Years.

>>59434184
So? Were those the games I listed? The Kitguru review obviously doesn't iclude those, but that game gives good idea and is completely in line with my experience with Armed Assault, Arma2, SHOC etc. when comparing 1700x to 2500k.

Enjoy your (almost) 2011 performance in 2017.

>>59425607
wtf, is this bait?
i still use the same CPU that I got 6 years ago and it still runs perfectly fine

File: ryzen3200.jpg (79KB, 618x613px) Image search: [Google]

79KB, 618x613px

>>59433988
You can set memory clocks to 3200mhz-3500mhz without touching the base clock.
AMD also has a notice posted on their site that another memory compatibility BIOS update is coming.

>>59434239
Also see the Tomb Raider performance@Kitguru - a game that uses as many threads as is available - and was even 'AMD (GPU) optimized', but yet zenver1 doesn't outrun SB at all.

>>59434239
so when everything gets fixed and ryzen gets the boost they need you will still be shilling for games that people dont play anymore ?

>>59434298
the first rotr was an amd game the second was a bloated "i run on the first thread because im too bored to develop my own third party api" gameworks

File: untitled-21.png (30KB, 683x709px) Image search: [Google]

30KB, 683x709px

>1700 loses for only 2 frames
>meeeeeehhhhh I'm going to intel because it's better :^)

>>59434300
'Everything get fixed', lol, you have no idea what you're writing about.
There won't be any major performance boost without a new arch (zenver2/Zen+).Solving little uefi problems and memory incombatibility won't bring massive performance boost. MS won't change cpu scheduler in any currently released OS either and there is changes anticipated to CFS or BFS on linux side either.

>>59434367
>there is no changes anticipated for CFS or BFS

>>59434367
you dont need MAJOR improvement since its not even 10% behind...

and second yeah ms already has the patch on the fast ring on 17xx ver

Thread blocked by my brand new Shillter (shill filter).

>>59431802
Given that SMT is buggy right now, it means that the gaming performance could be similar to a dual core given the 2+2 CCX design.
They had one fucking chance to destroy the entire intel i5 lineup, but they blew it.

>>59434367
>one week after launch many reviews posted articles stating new BIOS updates increased performance up to 25%
>using higher speed memory can increase game performance nearly 8%, likely higher with an even better kit
>turning off HEPT and setting power plan to high performance can increase frame rates in games by 18% on average
>hurr nothing will get fixed all the initial launch reviews are still totally accurate, goy

Don't you ever get tired of lying on the internet?

File: brainproblems.png (22KB, 623x371px) Image search: [Google]

22KB, 623x371px

>>59425264
>consumer electronics
>investment

>>59434361
>Conveniently not testing SB chips - which can be had used for pennies - which doesn't lose to latest Intel arch much either and especially not to Zen

>>59432662
Also, SMT is not working properly in games right now, so the 2+2 design might be closer to Pentium.

>Consumer products
>Investment
The only investment you could have made was riding the Ryzen hype wave and abuse the stock market with it, you're already way past that chance

>>59434367

>He doesn't know infinity fabric is memory speed bound
>He doesn't know that ryzen is memory speed sensitive
>He doesn't realize certain games are optimized for GDDR5 system memory
>He doesn't know that making draw calls across core complexes slows down multithreading performance that can be solved with sodtware patches and AMD compilers
>He still thinks Zen won't improve
>He doesn't know that

>>59433298
>Naples is the Zen product line for servers
and is literally the same chip slapped on a MCM

>>59434454
>SMT is not working properly in games right now
SMT leads to less effective single thread performance by design

https://forums.dolphin-emu.org/Thread-quick-tip-20-performance-improvement-for-hyperthreaded-dual-core-cpus-i3-etc

http://www.eecs.harvard.edu/~fedorova/papers/wiosca06.pdf

>>59434564
They mean that most games are seeing Ryzen 7 as 16 cores instead of properly seeing the SMT threads.

SMT is more efficient in many cases, but only when used with two threads that both have a lot of waits going on.
And AMD's SMT seems to work much better, tons of benchmarks show. But not when Games are thinking it's Bulldozer.

>>59434546
any argument you might make to substantiate the claim that Ryzen is a server chip would also support the notion that Naples is a consumer chip

>>59434639
>games are seeing Ryzen 7 as 16 cores instead of properly seeing the SMT threads.
No game or operating system does this not even for intel lol

>>59434444

This!

>>59425607
amd 1100t bought in 2011

still kickin it for 1080p gaming @4ghz

>>59425264
Fucking tryhards and your stinky chip.

Shoulda waited for Zen 2 like we all told you.

>>59434642

Fuck yes, give me 16core 32thread, octo-channel ram, 120pcie lanes in a seamicro eatx motherboard then two Vega x2s and convert every tablet, laptop, desktop in my home in to a thin client workstation with GPU pass through, and home automation.

>>59434674
Has already been confirmed to happen in some games like F1.

Any game that does its own thread scheduling can make this mistake.

File: circle-6.jpg (43KB, 236x244px) Image search: [Google]

43KB, 236x244px

>>59425264
I have a feeling that the Zen 2 will be the new Sandy Brigde.

They fumbled the launch, i expect them to refine their new architecture for the Gen 2 and it will be fucking golden, especially if they also crank those clocks up.
Sandy Bridge was a second generation too,

>>59434674
They will.

>>59434674
windows is seeing ryzen as 1x 8 core with 136mb of l3

this is the problem of the load balancer this is what we are waiting ms to fix

i dont even know why you answer back with an attitude while you clearly dont know shit

>>59425607
I expect my 5820k to last me at least a decade considering how things are shaping up.

>>59435041
It won't last a decade thanks to transistor decayl. Anything smaller than 32nm will have a short life thanks to that fact.
Only with 28nm and smaller it becomes a real problem affecting effective lifetime of a CPU and such.

File: amd fags and streaming.jpg (283KB, 2493x1005px) Image search: [Google]

283KB, 2493x1005px

>>59431784
hi shill, coffe break over

>>59425621
Getting a 1700 now and if zen2 is great I'll do a drop in upgrade

File: conan turns on his new monitor.gif (2MB, 623x340px) Image search: [Google]

2MB, 623x340px

>>59425264
Yes! Just like Bulldozer with it's large core-count for future-proofing!

>>59435402
How is he wrong?

The 7700 takes a huge hit to both average and minimum frames.

Dota2 just runs like shit on Ryzen for no reason what-so-ever except that it's not updated for it yet.

>>59435917
the fact that the 1700 barely dips at all suggests it's highly underused in Source 2

>>59434642
Naples is a package not a chip.
The 8 core 2x CCX Zeppelin die used in desktop Ryzen was first and foremost designed for servers. AMD does not have the resources to be taping out different dies for every market.

>>59427175
fuck. i almost bought a 1800x, dodged a bullet there.

>>59434361
>bench at 4k
>AMD athlong X4 800 loses by 1 frame
>>meeeeeehhhhh I'm going to intel because it's better :^)
I mean who cares that when you bottleneck CPU by benchmarking at high rez the framerate effectively gets capped.

Look goy at 8K rez, this pentium duo is as good as all the other cpus on the market that cost 500 and 1000 dollars! Why would you get the expensive ones when you can just get the cheap pentium duo

>>59433608
Zen also has great performance on launch, are you one of those who won't see at anything more than Skyrim at 720p?
Ryzen is offering Intel kilobuck performance for 300 dollars
>>59434016
Is Intel outsourcing their shills to reddit nowadays?
>>59434444
2500k's are around 150 used
You can get a 1400 for 180 new
>>59434488
Except he isn't, the hype won't die until Naples profits are known in Q4
>>59434795
I really doubt the clocks will go much higher, the only market that would benefit from a performance oriented process is the gaming market, which is a ridiculous minority
They might squeeze a bit more from their current process but that's kind of far fetched

File: 1488404508616.png (316KB, 882x758px) Image search: [Google]

316KB, 882x758px

It will get better..

>>59431155
>thinking this has any relevance to people who aren't gold farmers
>thinking running a mobile game and some 4 fps shit in the background is supposed to be impressive
kill yourself

>>59436544
what's important is how well the CPU itself performs. removing the GPU bottleneck allows us to examine the CPU in a vacuum (or almost). substitute the 1080 with a future GPU that doesn't bottleneck at 1440p or 4K and what do you know, the i7 will win there too. assuming no significant improvements, of course.

it's good to know that Ryzen doesn't bottleneck high end gaming setups however. you can buy a 1700 (or better) right now for 1440p+ gaming and suffer no consequences from that decision on the short term. if you plan to upgrade in 2-3 years, the difference will never matter. ask yourself whether Ryzen's non-gaming advantages over Kaby Lake are worth its potentially shorter lifespan as a high end gaming chip.

>>59433298
>Ryzen is not a fucking server chip. it was never marketed as such. it's a consumer fucking product. Naples is the Zen product line for servers.
zen was designed from the ground up as a server chip. just because ryzen was marketed to stupid ass gaming doesn't mean it isn't the same server chip that just happens to do gaming (but not as well as intel's mainstream processors)

>>59436845
>I'm only ever going to run Skyrim and only Skyrim
kys yourself

>>59436912
fuck off shill no one cares about your artificial workload

>>59436937
>i-it's not real
>p-please believe me
>m-my pc is a glorified console... it's not fair...

>>59436961
the fuck is your problem idiot

OP
Jesus Christ I didn't think this thread would get so big.
I just shitposted before sleeping, I don't even like AMD.

>>59436989
>I just shitposted
you're in good company. 4chan is mostly just a bunch of kids that are here purely to waste time

>>59425264
I'll get an R5 or R3, a good am4 mobo, and wait for Ryzen 2. There are problems (like the CCX interconnection issues) that simply cannot be ironed out with software updates.

The big showdown is going to be the $235 4c/4t i5 7600k vs. the $250 6c/12t 1600x. If the 1600 cannot outperform the 7600k despite being more expensive and significantly more multithreaded there is NO HOPE for future performance. You will be waiting another decade easily.

File: cine-single-1.png (503KB, 600x600px) Image search: [Google]

503KB, 600x600px

Looks like Sandy Bridge is a great place to upgrade to if you have one of the previous AMD processors.

HAHAHAHAHAHHAHAHAHA

>>59437437
>CPU points (single core)
okay

>>59437455
Hey if you want a affordable cpu to compile Gentoo kernel on, i suggest a 6800K. It's pretty cheap compared to the pricey AMD workstation CPU's.

File: don't open.gif (465KB, 900x1200px) Image search: [Google]

465KB, 900x1200px

I keep getting mixed messages from /g/. The 1700 seems to do fairly well in games and it excels in multitasking scenarios. Why does Ryzen suck again?

>>59437483
Personally I'd suggest waiting for a more affordable R5 or R3 to arrive first.

File: image.jpg (146KB, 640x819px) Image search: [Google]

146KB, 640x819px

>>59437437
is this bait?

>>59437504
It holds up pretty well against just about every task, even ones that it's not ideally configured for. It's actually intended to compete with $1000 8 core chips but AMD priced it at just under half of those to make them more attractive

>>59437504
it doesn't. intel shills are shilling in full force.

>>59437504
It's not bad, it's just that Intel is better everywhere else unless you're doing video encoding etc and want to shave an hour off your nightly work batch schedule.

>>59437520
Single core performance is even closer on yours. So you tell me, is it bait?

>>59437564
you do realize that the selling point of an 8c/16t processor isn't its single core performance, which btw. exceeds that of the 8c/16t intel in that benchmark?

>>59425264

>Any piece of computer hardware.
>Long term investment.

You're dumb as shit.

>>59437564
>it's just that Intel is better everywhere else unless you're doing video encoding
What if I have a ton of programs open at the same time? Does Intel still beat AMD?

>>59437504
I keep getting mixed messages from /g/. The 7700k seems to do fairly well in work related applications and it excels in gaming scenarios. Why buy a Ryzen again?

>>59437701
I keep getting mixed messages from /d/. The 2500k seems to do fairly well in work related applications and it excels in gaming scenarios. Why buy a Ryzen again?

>>59437675
most programs don't use much cpu resources just by having them open. the benefit would be if you run something like video encoding in the background, but if you're a youtuber or something you probably have a dedicated machine for video encoding.

>>59437726
I keep getting mixed messages from /o/. The 2011 mustang seems to do fairly well in work related driving and it excels in straight line top speed scenarios. Why buy a Ryzen again?

>>59437749
What if you have 20 tabs open in chrome, each one buffering videos and flash games?

File: 1488654951644.jpg (439KB, 1920x1080px) Image search: [Google]

439KB, 1920x1080px

>>59433271

>>59432794
Show me where this happened.

>>59431575
>call me a shill because of my argument
Those are good examples. Why not just post them in the first place? And where are the rest of the updates? In development.

>>59437786
Just your Internet and ram, really.

>>59437771
I keep getting mixed messages from /tg/. Elves seems to do fairly well in magic related applications and they excel in rape scenarios. Why roll a dwarf again?

>>59437786

If I'll ever start doing that, I'll consider seeking medical help.

>>59437786
a 7700k would do just fine, especially if you overclock it. a 7700k at 5 GHz actually comes very close to 8 core ryzen even in the most extreme multithreaded workloads like compiling the linux kernel for example.

File: 20170316_121542.png (198KB, 1109x1029px) Image search: [Google]

198KB, 1109x1029px

Amd retards on reddit

>>59437813
Right now my i5 is bottlenecking the shit out of me in exactly this scenario (and also Overwatch).
>>59437833
So the 1700 would actually be better in that scenario?

>>59437821
I keep getting mixed messages from /fit/. Overhead press seems to do fairly well in aesthetic gains department and it excel in practical strength scenarios. Why buy a bench-press again?

File: 1487638432167.png (347KB, 454x600px) Image search: [Google]

347KB, 454x600px

>>59437834
>4 core ryzen
>3.2 GHz base clock
>3.4 GHz boost clock

1600x looks more interesting but it's only 2 more cores and 4 GHz boost clock. 7700k at 5 GHz would still be better in a lot of applications

>>59437897
>I keep getting mixed messages from /fit/.
Good, you're almost ready for muscle confusion.

>>59437834
An r7 with 4 cores disabled is pretty much on par with a 7700k underclocked to the same frequency. The 1500x will essentially be a slightly cheaper slightly worse performing 7500. I don't know why anyone would consider a 1400.

Hasn't AMD already closed the gap with 7700K in gaymen by fucking with Windows settings?

Isn't AMD already competitive with 6900K and 5960X overall on Windows, and actually beating them on Linux simply because software can be compiled per platform? Have Ryzen optimizations even really surfaced for gcc or Clang?

>>59438035
I am wondering about the same thing desu. Ryzen seems very viable right now.

File: 88a.png (9KB, 215x215px) Image search: [Google]

9KB, 215x215px

>For now, AMD executives said they’ll disclose Ryzen 5 performance numbers at a later date, after orders for the chip have begun.

>>59438035
AMD has said they are starting to work with developers so we'll probably start seeing better multi-threading in games. If that happens r5s will render the i5s obsolete.

>>59427175
Thanks I just cancelled my order and got an 7700k instead

>>59438077

>thinking amd is even close to the jewery that is intel

>>59438156

Poor sap.

>>59438156
>tfw too poor to buy a new cpu and mobo every 2 years because intel changes the socket

>>59438156
good choice anon

>>59438214

Unless you buy budget (aka you're stupid) you usually want to upgrade everything.

That's why I spend a good amount on the cpu and mobo, because they are the most expensive and difficult to upgrade/replace.

File: 1488248377514.png (10KB, 497x556px) Image search: [Google]

10KB, 497x556px

>>59425264

No I think it will be an OK investment in the long run. No updates will make this shine and years later when games FINALLY take advantage of 4 + cores, will this cheap make a 4 core feel slow. By then Intel will release their 4 + chips and AMD will be a meme again.

Seems to me like AMD made a CPU that is at least competitive to Intel. So, of course now everyone on both sides are stirred up into a shitposting frenzy.

Look, if you're an Intel-fag you should be happy AMD put out something that's even good enough to have a debate about. It'll put pressure on Intel.

If you're an AMD-fag you should be happy they managed to put out something that can even pretend to compete with Intel.

>>59437947
>comparing mhz values across architecture
>comparing mhz values at all

Fuck off, retard.

>>59438367
I'm an AMDfag and I'm mad there is no 4+0 model. The 2+2 is going to be worse in most situations.

>>59438607
if you're looking for the best of the best 3.4 GHz is no bueno whether you're looking at kaby lake or ryzen

>>59438035
Somewhat.

On paper, the 8 core Ryzens are purely better than the 7700k except in single threaded performance (which games aren't even now days. Tons of games run better on a lower clocked 6900k over the 7700k now days)
But software isn't close to there yet to utilize it. It still has issues with Windows moving threads around when it shouldn't. It still has games seeing it as a 16 core instead of 8 core with SMT, or just plain treating it like it's Bulldozer when it's more like K10 and Haswell.

Since it's such an alluring CPU that lots of developers are going to buy, I'm waaaay more confident it is going to be optimized for. People said the same shit about Bulldozer, which was baseless since no developers wanted that and the potential really wasn't there much except for specific applications.

>>59438634

Buy an Intel CPU then.

That's how choice works.

That's why this is a good thing, even though Ryzen is "just ok".

>>59438779
The Intel CPUs suck dick for the price as well. $350 is the cheapest for 8 threads.
I wouldn't buy that either.

The 1500X will still be good value for the money, but it won't be great. If it was 4+0 even with the half cache, that'd have been on average over 10% better in gaymes and even for most other applications.

$190 to have 4c/8t would have been fantastic, even if it's clocked less, if only it wasn't this stupid "core 2 quad" shit.

>>59438855

What are you using 8 threads for? Out of curiosity.

>>59438883
e-peen

>>59438883
Marketing tells him he needs 8 cores, so he needs to get moar coars.

>>59438855
what are you on about? why would 4+0 with half the cache be better than 1500x?

>$190 to have 4c/8t would have been fantastic
so 1500x?

>>59438996
https://www.youtube.com/watch?v=BORHnYLLgyY

>>59438156
Sure you did friendo.

>>59439037
Stupid intel shills, don't you realize you can just buy a 1800x or 1700x or 1700 and disable 4 cores on the secondary ccx and get that CPU you want with 4 cores on the same ccx.

>>59438883
Almost every AAA game that's come out the past 2 years uses HT for an average of around a 25% improvement clock-for-clock.

Only a few games reusing super old engines don't.

>>59439152
>buy ryzen because muh cores
>disable half the cores because muh performance
literally just get a 7700k

>>59439265
Ryzen will age like fine wine and shit all over 7700k as soon as they fix the motherboards, memory, windows 10. The performance will improve by at least 50% and don't forget you actually have 8 cores and 16 threads to future proof your computer. Lastly unlike intel who changes sockets yearly the AM4 motherboards will be here for at least 5 years so you can easily buy zen 2 and 3 in a year or two respectively.

>>59439303
kek

>>59439303
>Lastly unlike intel who changes sockets yearly the AM4 motherboards will be here for at least 5 years so you can easily buy zen 2 and 3 in a year or two respectively.
Getting a little too close to the truth there goyim.

>>59439561
the motherboard is only like 75-150 bucks, 250 bucks for the AM4 board that can overclock to 4 GHz, but whatever

>>59439647
Uhhh what? You can OC to 4ghz on a $100 AM4 motherboard.

>>59439647
Fuck off intel shill, don't you understand it's much better to future proof yourself by buying ryzen right now.
And in a year you can buy ryzen 2 because the AM4 motherboard is compatible and in 2 years you can buy ryzen 3 because the same motherboard supports those CPUS.

>>59437504
Its 7 fps behind a 7700k, so thats why it sucks.

>>59439675
can't find the quote but 1700 could only overclock to 4 GHz on the high end motherboard

>>59439679
can't tell if you're serious or trolling. with a 7700k you don't even need to upgrade for at least several years

>>59439717
>can't tell if you're serious or trolling. with a 7700k you don't even need to upgrade for at least several years
7700k doesn't have 8 cores, literally every single game coming out will be made with 8 cores and 16 threads in mind.
Remember FX series, remember how we told you that games will be using more cores? WELL GUESS WHO IS LAUGHING NOW!

File: 1488896527219.png (765KB, 928x681px) Image search: [Google]

765KB, 928x681px

>>59439717
>the 7700k is the most powerful CPU ever. Nothing will ever be bottlenecked by it! Especially not a year old highly optimized game! ps I never have anything running in the background
dumbass

>>59439752
Agreed you need a bare minimum of 8 cores right now and it will only increase in the next years. You would be a fool to use obsolete shit like 4 cores.

>>59439793
I'd say 4c/8t minimum, but 6c/12t recommended.

The upcoming Coffelake 6 cores should be great.

4c/4t is only enough for 60fps. And that might not be true in 2 years.

>>59439813
>I'd say 4c/8t minimum, but 6c/12t recommended.
Fuck off intel shill. 8 cores 16 threads is the minimum today! Would not be surprised with 10/12 cores becoming standard in a couple of years.
Luckily with ryzens awesome architecture they can easily make those for cheap so ryzen 2 or 3 will have us covered with that.

>>59437812
Because you can easily find them, and have been posted thousands of times here
You won't get paid if I post them

>>59437833
To do 5Ghz you need to delid it and also a golden chip, along with a high end cooler
Most people barely get 4.7 of they know what they're doing and have decent cooling

It was expected that Ryzen will deliver single thread performance esimilar to Haswell for months now. I actually think that AMD succeeded in what they were attempting to do.

That said does anyone know if 7600K i5 processor will work on a Z170 board? I have the Asrock pro4s

>>59438077
We already know how they will perform, you can easily emulate their configurations with any R7

>>59439752
if someone has the money to go through 3 cpus in 3 years they don't give a shit if they can reuse their motherboard FUCKING RETARDED SHILL

>>59439752
>Remember FX series, remember how we told you that games will be using more cores? WELL GUESS WHO IS LAUGHING NOW!
lol ok you're having a laugh, good one i took the bait

>>59425796
Same here.

>>59439875
that's bullshit, something like 60% of all 7700Ks can do 5 GHz on +100mV, almost all of them can do 4.8-4.9, all you need is a good cooler, you don't even need to delid

>>59439900
7700k costs 530$
Ryzen costs 300$
Why would you pay more, for less cores?

>>59439967
(You)

>>59439939
You do unless you are fine with it throttling whenever you load it more than a few minutes

>>59440017
+100mV is close to stock voltage, most 7700Ks can do 4.8-4.9 at stock voltage or below, it's not even close to overheating if you have a good cooler

>>59440017
This guy knows whats up, all of intels overclocks are done with Liquid Nitrogen and even then it can only sustain it for 2 min just to get those benchmark results.

M$ and co. will totally fix all of Ryzen's problems because said chip isn't at fault at all! :^)
>just wait

File: 1484442295292.png (204KB, 600x528px) Image search: [Google]

204KB, 600x528px

>>59440040
>>59440017
>I-I swear, Intel can't overclock for shit either!

>>59434795
>circle anco
Good taste anon

File: 1482617810793.png (338KB, 640x480px) Image search: [Google]

338KB, 640x480px

>>59439967
>muh cores
>muh value
At it once more!

>>59440950
yes intel shill are you too dumb to understand 8 cores is more than 4 cores, why would you spend money on less cores

>>59437504
Ryzen is great, if you're a beggar that is.
Otherwise go with Intel 'cus Zen's a clusterfuck of problems atm.

>>59440992
Because it performs better all around for tasks that 99% of the world's population care about. And: >>59441001

>>59440992
quality > quantity

>>59441056
>>59441079
Ryzen is only 5% behind and with windows 10 fixes, memory fixes, optimizations for programs and games it will topple the 7700k by a huge margin.
You would have to be an intel shill to even suggest to get something that will be outdated in a month.

File: 1488636678357.jpg (62KB, 584x560px) Image search: [Google]

62KB, 584x560px

>>59441114
>Just wait
Alright, how much they pay you and your wife's son for posting this? You sure are one happy RedTeam+ member!

>>59441114
>delusional
>AMD shareholder
>shill
pick all three

>>59441143
Didn't you get banned from /g/ intel shill? We are really getting tired of you spamming your benchmarks and nonsense here.

File: 1484863085168.jpg (111KB, 588x647px) Image search: [Google]

111KB, 588x647px

>>59441114
>only 5%
Jewish in heart.
Middle Eastern in mind.
And a total poo-in-loo in body!

>>59436544

Of course, retard. Put a Athlong x4 800 with a GTX 1080 and run a 4k test.

>>59441354
>Ryzen 1400 gets the same FPS as 7700k and 6900k at 6k
>a cheap ryzen cpu is as good as those intel cpus that cost hundreds if not over thousand dollars
And yet you try to argue that people should be intel. Seriously getting tired of this board being raided by intel shills.

>>59441391
>>>/v/

>>59441391
Fuck, pre-ordering one and a 6K monitor right hell now!