IT'S OVER

INTEL IS FINISHED

>>57986772
How will Intel ever recover?

>>57986772
Dont forget 95w vs 140w tdp.

>>57986772
Have they released the video file for this yet so we can test our own PCs?

amd finally did it. they have slain the jew.

thank you based su.

>>57987147
AMD can claim whatever TDP they want.
In reality the CPU will still consume >300W.

Typical new AMD product scenario:
>AMD fags say it's the good time AMD surpasses intel/nVidia
>First, obviously not impartial, tests, say their products won
>You later find out their TDP is awful
>Later tests show they ain't even competitive with intel/nVidia rival
>We end up buying intel/nVidia and are fine with that
>AMDfags buy AMD and whine on /g/ that it's better to comfort themselves after they see their games run like shit.

File: 98797987987.png (8KB, 363x364px) Image search: [Google]

8KB, 363x364px

>>57987361
>X was bad so Y has to be bad too

This meme needs to die

>>57986772
Source?

>>57987411
It's not a meme, it's the natural coping mechanism of people who don't want to admit they made a shitty purchase.

>>57987416
AMDs zen stream

>>57987411
Man, it's just natural to not trust AMD after knowing how things went in the past. Maybe this is really the time they do something good, but since they failed time after time it's quite retarded to claim "OMG le AMD won!!11" after just a couple of fake benchmarks.

>>57987441
>Not even out
>Shitty purchase

>>57987498
Or you can just not be a retard and look at the facts. Zen had the same designer as AMD's most successful CPUs to this date, (leaked) footage indicated several similarities to Intel's Haswell including SMT, with a declared focus on IPC instead of just "muh cores" from the very beginning.

And no, these benchmarks are not "fake" unless AMD wants to actively fraud investors and risk lawsuits. They won't be necessarily 100% accurate, but it shows the direction they are going. Handbrake is more of an indicator than Blender, at least. Their 8350 wouldn't beat Intel's 6700k on it, let alone a 6900k.

Zen won't be the big savior in terms of performance, you're retarded if you ever thought that - But it will likely be the closest they've been since their K8.

>>57987655
Why are you attacking on me while there are 13 yrs old retards all over here who are claiming "IT'S OVER, AMD WON!!!11" over a little margin which is probably the "little inaccuracy" you're talking of?

I never denied this could be better than previous meme CPUs AMD did, but all these kids orgasming over bemchmarks AMD made (which even if not 100% false, are most likely made in a particular condition in which they know they'll be advantaged) are quite pathetic.

When more objective benchmarks will come out we'll be talking seriously.

>>57987196
yeah
my 8350 at 4.76 got 2min and 29 sec

File: [HorribleSubs] Mobile Suit Gundam - Iron-Blooded Orphans - 23 [720p].mkv_snapshot_22.05_[2016.03.13_20.55.32].jpg (151KB, 1280x720px)

151KB, 1280x720px

>>57987243
do i need to call suicide watch for you?

File: 1439320168559.jpg (13KB, 381x286px) Image search: [Google]

13KB, 381x286px

>expecting anything less from Based Keller

Apologize

File: keller_amd_great_again.png (418KB, 627x627px) Image search: [Google]

418KB, 627x627px

>>57986772
JEWTEL

BTFO
T
F
O

>>57986772
Have they released benchmarks for single-threaded benchmarks yet? AMD's issue - even under Excavator and Bulldozer - was never multi-threaded scaling.

>>57988113
Zen's 8C/16T is matching the competition on a core-to-core and clock-to-clock basis. A Bulldozer with eight "cores" within four Modules, e.g. their 8350, couldn't even beat a 6700k in both of the shown benches. Zen won (barely, that is) against a 6900k.

This indicates competitive IPC.

>tfw we're back to the athlon/phenom glory days when AMD was a viable budget pick


who /955BE/ here

Reminder that AMD haven't released a competitive CPU since 2005, when most posters here were barely out of diapers.

>>57987147
to be fair, my 6850k, at least according to hwinfo64, doesn't draw more than 90 watts at its stock 3.6ghz with turbo off during some abusive stress test like avx2 linx or avx2 prime. normal stuff, like gaming, or even intel burn test stress test, doesn't cause it to draw more than 60 watts tops.

140 tdp is there to ensure it can reach its boost speed of 4ghz with turbo boost 3.0 support. turbo boost 2.0 its 3.8ghz and its tdp is lower.

>>57986772
That was a Zen part locked at 3.4ghz while the 6900k was allowed to use turbo states.

Over indeed.

>>57987748
Is that in blender? Maybe I mishead but I thought they were releasing the video they used in handbrake too.

>>57989462
that was blender yeah

>bulldozer 2.0
>IT'S OVER GOYTEL IS FINISHED :^)
can't wait to see single-core performance

I'm speechless. As a harcore intelfag I honestly though Zen was going to faceplant hard and leave rivers of amdfag tears for me to drink.

Fuck man, it's not easy accepting this. It's like watching your older brother become homeless and you being powerless to do anything about it.

>>57988651
Intel has only had a competitive processor since 2009 and Sandy Bridge though. I'd say 1999-2009 it's been mostly AMD. Intel had a decent celeron you could overclock on the cheap back in the day before the 1Ghz barrier was broken.

>>57988209
>This indicates competitive IPC.

In the specific workloads shown. That is the caveat that you can't forget. All architectures have their nuances. Strong performance in one metric is not necessarily indicative of performance in another.
Though it was a good sign that they said they had beaten their 40% average IPC uplift.

Zen is the biggest core AMD has ever made, and everything in the front end looks adequate to feed it. Even if they wind up 15% behind Broadwell in a good handful of benches it will still be an incredible showing for a fucking brand new arch.

>>57987748
4.76ghz got 159s
For 3.4ghz it should be 113s.

It was actually 54s so drawing on this, it's a ~90% improvement in IPC. Something isn't right. Guess maybe the ram or something.

>>57989606
Zen is on par with 6900k whos single core is on par with 4770k

Wishful thinking but maybe you could overclock it into a 8 core skylake without burning your house down

>>57988702
Honestly can't imagine how shitty you must be feeling of your trouble purchase. So much money wasted.

>>57986772
Intel runs all cores at boost as long as temps aren't reaching the limit. So unless they were using stock cooler on both there could still be a bias in favor of AMD since Wraith is exponentially better than anything Intel offers.

the reason bulldozer was bashed was because of its terrible single core IPC which made it perform worse than its predecessor in games, which is the only metric autists care about. Bulldozer was a monster at tasks that used all 8 modules, beating even Sandy Bridge back then.

Now that games use more cores/threads, all AMD has to do is win the gamer autists and Zen will be a bonafide hit.

>>57989901
>Stock cooler
>6900k
Might as well kill yourself. Lisa said the setups were identical so literally identical cooling. Cooling would be a future advantage for RyZen (SenseMi Extended Frequency (which wasn't implemented)). I'm sure RyZen will do much better and chew the 6900k at launch.

>>57989919
It was a "monster" at something that nobody had any use for. Also, it's not a "monster" at 8 core workloads that don't do integer math. It's really weak in floating point.

The only time Bulldozer had good performance was when you gave it the correct workload, and a processor isn't good when it only works with the correct workload.

>>57986772
>cherry picking benchmarks

its almost as if they are selectively releasing benchmarks known to make use of all of AMDs cores.

It will be a different story when it comes to applications that do not use multiple cores.

>>57990092
both parts have the same amount of cores and threads though

>>57990092
Handbrake and Blender are hardly cherrypicking.

>>57987243
>>57988702
TDP is not the same as power consumption. What the TDP does mean is that the AMD processor will be easier to cool, so it can go into more things.

>>57990092
You're stupid

File: more to life than rendering.jpg (164KB, 1313x382px) Image search: [Google]

164KB, 1313x382px

>>57986772
just look at pic
this is the desktop Excavator cpu/apu released earlier this year.

You can win be kind of handbrake than then suck ass at gaming at the same time.

http://www.anandtech.com/show/10436/amd-carrizo-tested-generational-deep-dive-athlon-x4-845

>>57990121
They mean nothing to me. Fire up Arma 2 or Arma 3, or play Planetside 2 in a 96+/96+ clusterfuck and let me know how they run.

That's what I need to know.

>>57986772
Holy shit they actually did it. I have a 5820k and am now officially on suicide watch.

File: smug.gif (2MB, 260x462px) Image search: [Google]

2MB, 260x462px

>6 Cores, 12 threads, 3.2 Ghz unlocked with higher than sandybridge IPC for ~$250

Holy fucking shit

Intel actually BTFO

>>57986772
who cares about a oldass barely loses to a i7 6900 vs a brand new amd 'zen' with double the cores and threads

WHERE THE FUCK IS THE 490?

>>57990374
>higher than sandybridge IPC
congratufuckinglations
stupid /v/ gaymers

>>57986772
ZEN OPTERONS WHEN

>>57990413
>excited about more threads and not single core performance

>/v/ gaymers

Wow you're fucking stupid

>>57990442
Yeah man, that's what I want to know and they didn't tell me. I don't care about productivity shit or streaming shit.

>>57990279
>ASUS

>>57989608

you should be happy, if AMD delivers it, Jewtel will be forced to lower their prices

>>57990683
If they will.

>>57990374
>for ~$250
where are you getting this?

>>57990683
The wont deliver.

>>57990683
Intel won't lower their prices because people will buy it no matter what because it's the "safe choice."

>>57986772
DELET

>>57990957
his ass

>>57990130
Heat is a function of power consumption, unless you found transistors that generate zero heat

>>57988075
saved

>>57987498
both intel and nvidia have released utterly shit products, and not too long ago either

this is just fanboyism

File: 1474778786460.jpg (16KB, 400x400px) Image search: [Google]

16KB, 400x400px

>all these people wanting it to annihilate intel
>all these other people wanting it to be lol ayyymd housefire

It's better for all of us if they show up with something that is i5 or i7 comparable but hopefully costs you say $30-50 less. Then we're all happy because competition is great.

>>57987838
that's Roddy piper

so 3 years of hype

finally gets shown

when can we get it? a fucking year from now?

>>57986772
can't wait for 7000k series benchmarks to come out :^)

>>57992542
q1 2017

in otherwords, at the very end of march

>>57992577
lmao cannon lake will be nearly out by then

>>57992616
Intel will probably push it back to 1H 2018 like they did with Kabyfail

>>57992616
let's hope it's not a fucking disappointment like kaby lake

>>57992344
Heat is the by product, if it only produces 95w of heat that means it's more efficient or can clock itself more efficiently over a 140w processor.

Hence why Intel's started to sell 4w processors as processors that can TDP up to 9w or TDP down to 2w as needed by adjusting clocks.

>>57990322
>muh gaymes
>>>/v/

>>57989834
>In the specific workloads shown. That is the caveat that you can't forget.
Blender it's literally what Intel recommends to bench their own processors you stupid shill nigger

>The final two tweaks—"Neural Net Prediction" and "Smart Prefetch"—help to shuffle data through the CPU more effectively. AMD claims there's a "true artificial network" and "learning algorithms" inside Ryzen that are able to predict future decisions, pre-load instructions, and learn application data access patterns in order to improve performance. That's aided by a beefy 4MB of L2 cache and 16MB of L3 cache.
Holy shit.

>>57993173

You didn't get the memo. From the moment AMD first showed Zen all those months ago it automagically (and retroactively) made piledriver good in blender because that is how AMD cherrypicks.

Nevermind that piledriver gets slaughtered in blender.

>>57986772
This seems to be a pretty standard AMD practice these days
Fanboys post a few pictures here and there of their products only operating at 50% and still beating the competition, saying BTFO and INTEL NVIDIA BANKRUPT. Then when the product releases it's fucking nothing and they go quiet for a while until rinse and repeast

>>57986772
Is there a high resolution version of this? I'm interested in which codec and filters they used for this benchmark.
If the price is reasonable I'll buy a ton of those.

File: 05-Handbrake-2c11b0a4d95381b0.jpg (1MB, 2000x1474px) Image search: [Google]

1MB, 2000x1474px

>>57993569
Looks like they just used the appleTV 3 preset

>>57986772

Intel sat on their zionist asses for 4 years of tick/tock faggotry while Jim Keller was in the basement building this beast.

>>57993173
Don't care. Show me benchmarks in specific CPU intensive games and compare then to the 4790k, the 6700k the 7700k.

If it wins or at least comes very close in them, then it's worth buying.

If it doesn't then into the trash it goes. You can keep it for your workstations.

File: carmack-1080p-in-1995.jpg (41KB, 468x332px) Image search: [Google]

41KB, 468x332px

I've never been more happy/proud for a corporate entity.

>>57993671
>muh games
Back to /v/reddit

File: Screen Shot 2016-12-13 at 11.19.30 PM.png (58KB, 651x487px) Image search: [Google]

58KB, 651x487px

Never thought I'd see the day.

>>57988593
My CPU History
>955BE
>Bullshit FX-6100
>Master Intel i7-6700k
Fuck AMD

File: c06_intel[6.jpg (47KB, 575x282px) Image search: [Google]

47KB, 575x282px

>>57993735
who could be behind this post?

>>57993735
HAHA YOU ARE DESPERATLY TRYING TO FIND A REASON TO NOT KILL YOURSELF RIGHT KNOW SO YOU POST ON 4CHAN LMAO INTEL KEK ON SUICIDE WATCH BTFO

File: AMD master race.jpg (118KB, 775x577px) Image search: [Google]

118KB, 775x577px

>>57993735
Gtfo jewtel shill

>>57986772
A 8C 3.4GHz has 10% more perf than the last gen 8C 3.2GHz from Intel ? Absolutely nothing to fear, it's awful. Intel will come with new CPU with like +25% perf.

>>57993735
>>Master Intel i7-6700k

Yeah you better say that - you still 30 more installments on your i7, kek.

>>57993825
There's no new Intel CPU's with actual improvements until 2018, even then it could be delayed like Kaby Lake was
We already saw Kaby Lake, literally zero improvement IPC wise and about 100 mhz faster base clocks

The only thing that can save Intel now is going all out in bribing, or aggressively cutting prices

>>57993825

95w vs 140w ;)

>>57993825
>JUST
>WAIT

>>57993868
Is there any standardization in measuring the TDP? Or companies just pull the values out of their asses?

>>57993860
I like how you imply that intel needs saving.
AMD is going to get stomped, for the same reasons that they got stomped during the athlon 64 days.

>>57993914
I like how you imply that AMD needs saving.
intel is going to get stomped, for the same reasons that they got stomped during the K6 days.

>>57993700
Yes I want the CPU that gives the best performance possible in games. I want to squeeze out every frame I can to make use of this 144hz monitor.

I could not possibly care less about work related shit like streaming or video rendering and all that other shit I don't do.

>>57993888
>Is there any standardization in measuring the TDP?

No. Its not something that could be standardized. Take two chips that are identical in all regards except one of them has a max temp of 50c, and the other 80c.
The 50c chip needs more cooling to keep it operating within its operating range. The 80c chip can run hotter, and therefore get by with a more sparse solution.

Temp range is just one factor in how TDP for a given part is determined. Another is logic density, even layout of the logic itself. Every chip is different, so TDP can never be an ISO metric.

>>57993914
>AMD is going to get stomped, for the same reasons that they got stomped during the athlon 64 days.
Intel can't pull out a bribe all the OEM's with less repercussions than they had during the Athlon 64 days
Also, AMD was pretty close to 50% of marketshare during those days, despite the massive bribing Intel did, they didn't got stomped at all
If Intel pulls the same shit again they could end up broken up by the federal government though, the only thing they can do is price cuts
>>57993981
There were muh gaems benchmarks in the stream, the entire stream was catered to vidya retards like you
And Blender is pretty much an equal workload to most games

>>57994006
plus bribing retailers isn't nearly as effective as it once was since everyone orders stuff online now and Amazon would just tell them to fuck off

>>57993914
>AMD is going to get stomped, for the same reasons that they got stomped during the athlon 64 days.

Illegal market manipulation?

>>57994006
>There were muh gaems benchmarks in the stream, the entire stream was catered to vidya retards like you

Yeah all DX12 games that mean nothing to me. An 8350 could do well in those. They also did them at resolutions I don't run.

A 6700k could spank the shit out of it for all I get to know.

>>57994006
>And Blender is pretty much an equal workload to most games
You fucking what?

>>57993996
So, there is nothing to stop the companies from using optimal test conditions and the bragging of how efficient they are?

And then we have fanboys using this non-comparable values to wage shill wars? Noice

>>57994020
You can't even do a basic Google search, can you?
The main Intel bribing was OEM's
>>57994024
>it wasn't exactly my setup so it's invalid
The blender run was there for most ballpark comparisons, it's pretty much at least slightly better than a 6900K, probably faster once boosting is enabled
>>57994030
Yeah, DX11 games don't scale nearly as well as Blender does on multiple threads, but it was against a 6900K with an equal amount of threads

>>57994057
Correct.
Though in most all cases every manufacturer out there is being very honest, aside from intel who briefly tried to push the "SDP" metric with their Atom line. SDP was a figure measuring average draw in simple workloads like browsing the web, or playing back a shitty low resolution movie. This let intel put 10w chips into sub 5w devices by saying something along the lines of "It only draws 2.5w some of the time!"
The only problem is enthusiast wannabes who fail to understand what TDP is, or why its meaningful. and then shitpost all over the internet throwing around numbers that are totally over their heads.

Anyone else hates the Ryzen name and logo?

>2011
>trusting amd

>>57994170
Yes, very much. Athlon, Phenom, Sempron, Duron all sounded much better.

>>57994183
those were all fucking retarded names too, just like pentium and core i-whatever.

>>57986772
well jim ryzen keller delivered once more...

BUT im really sceptical about the blender demo...
either amd did some mumbo jumbo to complete it on 34 sec or their new extension for it is fucking awesome...(the one that is on development and i assume they use it it..)

File: 900x900px-LL-ca927bfa_download.png (435KB, 796x498px) Image search: [Google]

435KB, 796x498px

>>57988075
for the love of god someone needs to shop this hat on keller head

>>57990322
You are a legitimate fucking retard. Both Blender and Handbrake are heavily used for media creation, you know, stuff that actually makes you money?

>>57992443
But AMD has released nothing but shit for years. It's grounded skepticism. I will not be on the hype train until the processors are in the hands of independent reviewers. If it still compares this favorably once it's subjected to tests that aren't cherry-picked, then I will get excited. Until then, AMD has let me down too many times to believe they won't do it again.

>>57994261
they just dont get it lol they think a badly programmed game can somehow be better than a program that really uses all of the cpu features..
>>57994274
this is literally the first cpu amd has shown with lisa on the helm...
in reality everything amd has won so far is because of her and raja and keller..

>>57994285
>this is literally the first cpu amd has shown with lisa on the helm...
Having a new boss don't mean shit if it's still the same retards drooling on the keyboards. Again, I'm withholding judgement until we see independent testing.

>>57994305
indeed i agree with you.. but on overclock net people already have replicated most of the results so till now amd has actually delivered
http://www.overclock.net/t/1617227/amd-new-horizon-zen-preview-on-12-13-at-3-pm-cst/600
its a bit of a wall of texts 20+ pages but you can get an idea at least it works and its not another bulldozer shit

File: its time to stop posting.jpg (14KB, 191x226px) Image search: [Google]

14KB, 191x226px

>>57992344
>TDP doesn't mean that it will only use 95W its gonna use 300W of power OMG AMD SHITS STAHP D:

>Heat is a function of power consumption OMG AMDFAGS STOOOWP :C

>>57986772

>Benchmark provided by the manufacturer of the product

Yeah k, lets post some nvidia benchmarks while we're at it.

>>57994368
amd released all the files and the specs of the system they did their tests so that everyone can do it...
so no their benches have already been replicated and its accurate

>>57994379
They've already been replicated with Intel hardware. Wait for Zen to land in the hands of reviewers before jumping on the hype train. Again, it still needs to be seen if it compares as well when the tests are not cherry picked to favor it.

>>57993303
Skynet

>>57994403
well yeah that is why they release it lol
also they used a version of blender that is about to get released with avx265 not the current one
in the current one only a dual cpu setup can reach up to 34 seconds..

>>57988075
From now on. Jim Keller is a Saint of /g/

Say it with me

Saint SHITWRECKER Keller

>>57990130
Calculations are not work, CPUs are essentially highly advanced heaters with nearly perfect efficiency.

whoever has amd shared be prepared to get a good spike today

>>57988075
Thank mr Keller

>>57993713
fuck why didn't I buy this shit when it was like 2 dollars a share

>>57986772
AMD still has hardware botnet and locked down BIOS, while ME partition on Intel systems can be erased.

File: 1477173328909.png (211KB, 500x361px) Image search: [Google]

211KB, 500x361px

>>57994916
For the past 10 years this was no problem at all to anyone but Stallman, and suddenly in the wake of Zen, being able to disable ME is the second coming of Christ?

>>57995027
le epik post ekkkssdee

>>57986772
Problem is that we still don't know single core performance.
Even faildozer and vishera weren't bad in heavy multithread benchmarks.

I-is this finally happening?

>>57995056
Cool argument

>>57995027
well yeah its a typical intel nvidia argument
when intel is about to get shit suddenly something becomes bad

File: 1475730162042.png (728KB, 1165x1075px) Image search: [Google]

728KB, 1165x1075px

>>57986772
>2016
>the year silicon Jew was defeated

>>57995276
sauce?

File: Screenshot_20161214-045131.png (323KB, 1080x1920px) Image search: [Google]

323KB, 1080x1920px

>>57995291
funny junk dot com

>>57995076
We actually do know that it's comparable to Intel's since both chips were 8c/16t, unless AMD's SMT implementation is vastly superior to Intel's.
At this specific workload, that is.
The major deciding factor will be Zen's single core turbo frequency.
The 6900k goes up to 3.7GHz in single threaded workloads and 3.5 I believe in multi threaded ones.

A core parking mechanism like you see in mobile ARM CPUs would be pretty dope.
That way AMD could switch off unused cores and use the additional TDP overhead for higher single core turbo frequencies.

>>57989919
fuck off freetard

File: A-10_5700.jpg (69KB, 500x319px) Image search: [Google]

69KB, 500x319px

will it be soldered?
amd did use paste in some of their previous cpu

>>57995304
Based

>>57995343
Biased

File: image_2.jpg (20KB, 255x265px) Image search: [Google]

20KB, 255x265px

>>57995343
>>57995470

I might be able to use my AMD stickers on my cases in an unironic way now.

>>57987748
Where would I get that file etc? Can't seem to find it anywhere

But what about the price?

Performance is solid but useless if price isn't lower than Intel.

>>57995524
download.amd.com/demo/RyzenGraphic_27.blend

File: UUd29da.png (27KB, 634x480px) Image search: [Google]

27KB, 634x480px

Who /loadsamone/ here?

>>57995544
Nobody really knows until it's released and then it's going to depend on region anyways. THere are some good educated guesses at around 500 900ish dollar range.

>>57986772
You're not fooling anyone AMDunces

>>57994305
>Having a new boss don't mean shit if it's still the same retards drooling on the keyboards.

That's incorrect.

Success or failure of a company or any other organization goes with the management, not the workers.

If your company fails, it's not because the workers did something wrong, it's because management did something wrong. Anything else is simply trying to deflect blame from where the responsibilities belong.

>>57995620
This. Management sets the work environment and how well employees perform. Bad management results in bad hiring practices, bad habits, bad project planning, and everything else. Morale is also important like in a Paradox game with your units.

out with jewtel

in with chinkmd

I take it nobody bothered to test blender thing with intel 8 core?

>>57995679
>chinkmd
Need to slide a poo in there somewhere

Intel i7 6700k Shoah Edition
AMD Poo in the loo x4 w/ boost?

>>57995751
Maybe AMD would be best with going with Designated Shitting Pipelines or something

>>57994245
give me a high res pic of keller and lisa su and i'll do my best

>>57987361
Actually a realistic scenario

>AMD releases a huge new product
>NVIDIA and Intel counter the next year doubling and tripling with their Ace card waiting for it
>AMD cries in a corner along with their fanbase

>>57993303
It can learn your moves Anon
>Pornhub
>YouTube
>Steam

Lol kek.

>>57993665
Then Intel counters and AMD gets btfo like NVIDIA did. They always counter AMD.

File: 1481703908447.png (203KB, 500x420px)

203KB, 500x420px

>>57994245

File: trump_a_fucking_straw_1477017028824.jpg (536KB, 1000x1000px) Image search: [Google]

536KB, 1000x1000px

>>57994006
>If Intel pulls the same shit again they could end up broken up by the federal government

You might have had a point back before November 8.

By the end of January, there won't be anyone left in the government with any fucks to give about elitist, pot-smoking hippie faggots whining about regulations and shit.

It's cute how you seem to think there's going to be "rules" or something.

>>57996127
I do think a Trump Administration will be similar to Teddy the Roose is Loose. He doesn't seem shy about breaking up huge monopolies or at least threatening them. Amazon and Google come to mind. God I hope he burns Google.

File: 1474725922286.jpg (89KB, 880x505px) Image search: [Google]

89KB, 880x505px

>>57995899
actual scenario:
>AMD release a huge new product
>Intel tries to counter by funding the HeForShe diversity in tech program
>Nvidia tries to counter by have a special new partnership with Facebook and Google with fun new Farmville themed driver updates
>AMD is used by intelligent people, Intel and Nvidia get picked up by the Macfag Facebook social media crowd of college liberals
>they both have substantial shares of the market

Oy vey. It should be great for my daily x265 encoding.

>>57987147
>muh tdp

kys

>>57996288
Is that why he picked the CEO of Exxon to be secretary of state, or that many of the people he has advising him have expressed support for AT&T acquiring T-Mobile?

File: 1366637041772.jpg (117KB, 642x753px) Image search: [Google]

117KB, 642x753px

>>57986772
>8C/16T 3.4 GHz ZEN VS 3.2GHz 6900k
more like 3.5GHz 6900k
turbo was enabled
all core turbo is 3.5GHz
intel is finished

>>57996827
It depends if trump is going to bow to them or they're going to bow to him
Given trumps personality, do you think he is someone who would bow to people that work for him?

>>57996858
>do you think he is someone who would bow to people that work for him?
I don't think he will "bow down" to them but at the same time Trump is not an ideologue so he won't fight big business if it's what his advisors support.

File: 1473594079159.gif (628KB, 500x500px) Image search: [Google]

628KB, 500x500px

>>57994687
>tfw I did

>>57993303
how will CPU learn? it can't store anything, do we need CPU drivers now?

>>57996901
>it can't store anything
it has a cache
most of that 'learning' is just marketing buzz though, since cpus already do that. that's the whole point of brach prediction

>>57996934
so, basically it doesn't learn then
it's like saying fish learns from it's mistakes

File: thumbs up.jpg (105KB, 640x640px) Image search: [Google]

105KB, 640x640px

>>57996954
>so, basically it doesn't learn then
>it's like saying fish learns from it's mistakes
yup

Like a phoenix they have ryzen from the ashes
(Of burnt mobos with fx9590's)

I hope this brings new APU's

>>57986772
Why does the intel side not have a taskbar? What are they trying to hide?

>>57997177
Most likely plain incompetence.
If they were to fake the benchmark, they would've done so without it being noticeable in the taskbar.
There were also spelling errors in the slides such as
>Platfrom

>>57995560
I wonder if the anon whose wife invested a huge portion of their savings on AMD stock is here.

>>57997177
Anybody with a 6900k should be able to emulate their results. We just have to wait.

>>57997031
i would like to insert my penile organ inside her vagina vulgaris

>>57996773
>Intel and Nvidia get picked up by the Macfag

>Nvidia
>Mac

How long until Intel gets Keller suicided?

>>57998537
Keller now works for Samsung. If anyone an heroes him, it'll be Apple.

>>57998596
was he fired from tesla?

>>57986772
>no quad channel memory support on AM4
AMDfags on suicide watch it's over

>>57998537
Most say 6 months after Zen releases

File: 02d.jpg (8KB, 200x254px) Image search: [Google]

8KB, 200x254px

When where you when intel was kill

File: you_asked_for_this.jpg (51KB, 600x515px) Image search: [Google]

51KB, 600x515px

I won't even be buying this or an intel product in the next few years, but god fucking bless.

>>57999350

You should see the intel defence force out in full effect on various tech sites. Word is blender is now AMD biased because FX chips have historically done well in it (they haven't) and this also overlooks that Intel likes to use blender to demo their higher end products.

File: Laughing Whores.gif (117KB, 400x400px) Image search: [Google]

117KB, 400x400px

>>57999350
Goddammit, gets me every time. At least Kaby Lake will have great clock speeds out of the box.

>>57994261
I don't give a shit about media creation, I care about not dropping frames in games because of a shit CPU bottleneck.

>>57996934
It literally says it uses neural networks.

JIM "LEAVE INTEL DEAD IN THE CELLAR" KELLER

>>57986772

OY VEY GOY DELET THIS

>>57995550
Any specific settings or just render image?

amdrones suffer delusions of grandeur while never admitting to themselves that their third rate hardware is designed for stupid people

>>58000648

Set samples to 100.

Note: the general shit you have running in windows can affect scores.

>Bbbbut Intel has Babbylake!

And Zen has Zen+ with another 15% IPC improvement. Leapfrogging competition means healthy competition, which means better pricing, and faster improvement. And Intel is under MUCH deeper scrutiny since they were caught bribing OEMs.

>>57994261
>Blender and Handbrake are heavily used for media creation
Lol, no.

File: 14334708_316045805424811_7504734499628908544_n.jpg (104KB, 750x750px) Image search: [Google]

104KB, 750x750px

>>57997031
Oh hi there Katya. :3

File: 1475797255730.jpg (38KB, 500x307px) Image search: [Google]

38KB, 500x307px

>>57986772

Ryzen had no boost.

Jewtel had boost on.

>>57996854
This, if someone ever used an Intel cpu should have noticed under full load it never gets down to base frequency, even when all cores are maxed. So we can assume that the 6900k was actually running at least a few hundred mhz faster than 3.2ghz

Like some others have said already, however well Zen will perform in the end the best possible outcome will be competition between the companies leading to innovation.

>>58000476
>It literally says it uses neural networks.
branch predictors already are neural networks... that's the point
just nobody bothered calling it that before

>>57996934
branch prediction is not learning

>>58001961
Well I never knew that. Huh.

>>58000729
Ahh I'm doing it in Linux so my results will be skewed af

>>57987486
kek

Trusting intel, ever. :^D

https://www.youtube.com/v/Otcge1cn8Os

File: a.jpg (352KB, 917x946px) Image search: [Google]

352KB, 917x946px

>>57987748
here's what i get on my [email protected]

>>58003821
where are the files so we can all test this?

>>57997057
kek

>>58003821
I got 02:36:06 on 8350 at 4.4Ghz but I also had a lot of tabs open in Chrome.

>>58004030
i have 156 tabs open in icecat, but that doesn't matter, since i paused the icecat process

>>57996896
i bought btc instead and made like 50 dollars on that, i could have made 500% returns instead...

>>58003821
got 2:19 on an i5 4670k stock

>>58004051
I actually just restarted and did it with nothing else open, no chrome and all my applications closed and it only save 2 seconds so made basically no difference.

>>58004178
Doesn't really fucking matter unless every tab has adverts or videos running in the background.

>>58004760
It ties up resources, and Chrome is a resource hog to begin with.

>>58003916

https://hardforum.com/threads/amd-ryzen-blender-benchmark-scores-thread.1919896/

You can get them here.

>>57986772
>Barely faster
>200MHz higher clock speed
So they just barely caught up with current Intel CPUs (if that)?
Took them long enough.
Now they can finally regain their budget-king position while Intel still sits on top.

File: Render.png (1MB, 1903x1005px) Image search: [Google]

1MB, 1903x1005px

5820k stock, with tons of stuff running on the background.

>>58005438
>>58004090
gonna need a webm on that boss

>>58005347
>200MHz higher clock speed
actually the 6900K ran with boost freq so that means a steady 3.5ghz at the very least

boost was turned off for the ryzen chip which makes it faster core for core, clock for clock

good luck finding some other shit to nitpick though

>>58005599
>>58005347
not this guy, but
you know, i originally thought this, but im questioning it too
they didnt specify that it had great cooling, and without good cooling it could "throttle" to base clock, or just not turbo up
with that in mind, it would still be pretty much neck and neck though

>>58005712
I have no clue what they used to cool the 6900K but I'd wager they used something good enough to run it at stock level

I think they can shell out 30 bucks for a 212 Evo or something. that's a bit better than their wraith cooler afaik.

in any case, others have tried to reproduce the same benchmarks on the same CPU and have gotten worse results, so there's no reason to suspect foul play by AMD

>>58005712
Both had their stock coolers straight out of the box.

>>58004090
>>58005438
Turn CUDA acceleration off first.

File: ryzen.png (1MB, 871x858px) Image search: [Google]

1MB, 871x858px

meh ok i guess

If the AMD is using some ridiculous cooling system and their new automatic overclocking, it will be really easy to fudge the numbers.

>>58005757
Well thats not very fair. Everyone knows intel cpu coolers are just decorative items you throw away after opening the box.

>>58005788
They had the automatic overclocking stuff off for the demo, no turbo boost equivalent, and it was just the stock cpu cooler it will ship with.

>>58005801
if intel's CPUs aren't rated to run at stock settings with the stock cooler then ought to get their shit together

>>58005801
It's perfectly fair. They weren't overclocking it, it was stock settings, which is exactly what the stock cooler is supposed to handle perfectly fine.

The reason you throw away stock coolers is because they don't give you overclocking headroom, which would have just been extra variables added to the test.

>>58005757
>6900k
>doesnt have a stock cooler
you what
>>58005755
thats what i would expect, like i said we dont know so im questioning it
can you post those results? i really want to see, ive been looking myself

>>58005438
What is tons of stuff?

>>58005864
hardcore ghey pr0n

File: 1439364889364.jpg (425KB, 1680x1050px) Image search: [Google]

425KB, 1680x1050px

>>57989837
>Something isn't right.
The Intel kikes sure get anxious when Jim Keller is around

>>58005909
I believe you.

>>57989837
You're forgetting the extra threading it has.

>>57997057
kak

>>58005771

Just checked, it is off.
Considering how they had the 6900k in that test and I'm running 5820k, there shouldn't be anything strange about that 34.5 second result.

File: snip1.png (1MB, 854x842px) Image search: [Google]

1MB, 854x842px

I am the reigning champion.

>>58006549

What do you have there, a dual Xeon workstation?

>>58006569
Yes

>>57986772
Someone tested their stock 6900K and beat Zen's time by two seconds. Something's fucky.

>>58006840
In the live demo they used dual channel on the 6900k

>>58007319
kek, trying to cover up that gimp.

>>58007319
You're incorrect but that's a good guess coming from a total idiot.

>>58007319
source?

>>58007319
They used dual channel for the Ryzen, the 6900k got quad channel.

Ryzen probably would have gotten quad channel as well, but the mobo doesn't support it. That'll be a downside if they don't get any quad channel mobos made.

>>57996106

saved.jpg

>>57998596

>>57998605

Samsung was a rumor....Keller works for Tesla's elite group made up of ex-Apple bigshot engineers.

>>57996854

I love that picture. Original Athlon through Athlon64 were great days.

>>58007758
It's because they're still using PGA. 1300 pins doesn't cut it for DDR4 quad channel.

>>58008009
So you're saying it's a CPU problem not a motherboard problem?

>>58007758
quad channel is pointless. xmp is more important.

>>57986772

not believing any of this shit until some actual reviews come out. that being said Intel better lower the price of their CPUs since AMD is going to make alot of money underselling their cpus at better performance and wattage

Anyone spot VLC on the BF1 or star wars footage?

Wolves hunt in packs, aryzen

>>58007805
stupid cars ruin everything, I wonder where he will wander off next that would be actually interesting

apparently he incorporated arm techniques into zen

>>58008766
don't be a retard, nobody sane would sue unstable corruption out of ass VLC for that

>>58008961

dohohohoho

File: core2duo.jpg (182KB, 1280x700px) Image search: [Google]

182KB, 1280x700px

Surprisingly good result, if you ask me. Would have guessed 10-20 mins.

4:43 on FX8120. 5x faster is pretty decent.

File: 290x.png (869KB, 1377x957px) Image search: [Google]

869KB, 1377x957px

>>58009257
290X using openCL. Still, ryzen offering almost the same speed on a CPU renderer and 90W TDP vs 300W water cooled R290X is pretty good.

File: 2016-12-15 02_46_11-CPU-Z.png (36KB, 418x415px) Image search: [Google]

36KB, 418x415px

>>58009310
>>58009257
>>58009101
How do you do the benchmark?

File: samples.jpg (30KB, 272x477px) Image search: [Google]

30KB, 272x477px

>>58009498
Download and install Blender from https://www.blender.org/download/
Download the file from download.amd.com/demo/RyzenGraphic_27.blend
Open the file (with Blender).
Set the sample count to 100, by default it should say 200; pic related.
Close running programs that might negatively impact performance -- the rendering will occupy all cores.
From the Blender menu, select 'Render' > 'Render Image'

File: 2016-12-15 03_10_40-Blender_ [D__Users_Downloads_RyzenGraphic_27.blend].jpg (81KB, 970x843px) Image search: [Google]

81KB, 970x843px

>>58009565
>>58009498
Im not sure if I did something wrong it seems it was too fast compared to some other people ITT

File: 1477866234982.jpg (125KB, 750x841px) Image search: [Google]

125KB, 750x841px

>>57996127
Spot the hillary shill

File: Screenshot_20161215_031837.png (313KB, 1600x900px) Image search: [Google]

313KB, 1600x900px

>>58009645
here is my laptop

File: Untitled.jpg (263KB, 1920x1080px) Image search: [Google]

263KB, 1920x1080px

4790k 4.6ghz

>>58009645
To be honest, I'm not sure if I did something wrong either.
If you take into account clock speed and number of cores and normalize by these values, any CPU seems to get similar speeds.
E.g. my Core2Duo (which is from like 2008) takes ~6 minutes. The new AMD Zen (8 cores) takes ~1 minute. Considering that the benchmark is nicely parallelized, I could just divide my time by 4 to compare per core speed with the Zen.
So if my CPU had 8 (of the same) cores instead of just 2 it would take about 1:30 (maybe a little more, since parallelization is never 100%). If you also take into account clock speed (2.2 GHz vs 3.4+ Ghz), though, the performance per core and clock seems almost the same.
My guess: the benchmark is probably memory bottlenecked and can't really show whether it has better performance per clock at all.
But maybe we just fucked up some Blender setting.

File: testblend.jpg (323KB, 1915x1087px) Image search: [Google]

323KB, 1915x1087px

>>58009565
Wow that was quite a bit faster than my default 200 run.

File: mad.png (888KB, 1044x972px) Image search: [Google]

888KB, 1044x972px

>>58009565
RYZEN BTFO!

>all this damage control
>AMD w-was using the OC t-thing
>I b-bet the i-intel cpu wasn't ex-copper properly
>b-blender is b-biased

>>57986772
Isn't the Zen CPU that's competing with it meant to cost about 1/3rd the price too?

Intel is fucking finished.

>>58010381
Yea yea, don't get your hopes up. AMD is notorious for hyping their shit to heaven and then delivering something lukewarm.

If they do deliver, great. But will they?

>>57998641
kek because surely every single other system out there has seen so much increase in perf when having a quad channel..

>>58010381
can you actually provide any evidence that amd was using the oc -t thing?

>>58010425
All I care about is competition. Even if they only scratch the surface with this and hot even harder with Zen+ that would be something.

>>58010460
They supposedly weren't, otherwise there's barely a point in clock for clock comparisons.

>>58010514
No, I get that, but buying into hype is first step on the road to disappointment.
I find the numbers floating around to be unrealistic. Specially the idea that AMD is going to sell this shit for peanuts when it so thoroughly trounces intels numbers.

>>57996934
>>58000476
>>58001961
>>58001983
>>58002000
found the origin for the neural network thing

>It is interesting that Samsung has broken ranks by publicly declaring that its prediction engine is a neural network.
>AMD's Zen architect Mike Clark confirmed to us his microarchitecture uses a hashed perceptron system in its branch prediction. "Maybe I should have called it a neural net," he added.

http://www.theregister.co.uk/2016/08/22/samsung_m1_core/

>all those benchmarks almost at the 8cores Ryzen CPU with one or two years old 4cores

Welp, AMD lied once again, are they allowed to do that ? I'm pretty sure we can sue them and make them fall once and for all.

>>58010564
amd isnt going to sell this shit for below 500 dont even dream about it

>>58010564
I won't be disappointed, I have an i5 that still does more than I need it to. I'm not desperately waiting for some product to fit my needs. I'm just excited that SOMETHING is happening in the cpu world.

It won't be peanuts but it will probably be competitively priced.

>>58010574
what the fuck are you talking about, nigger?

>>58010574
The fuck are you on about.

File: 197809911_1280x720.jpg (32KB, 1280x720px) Image search: [Google]

32KB, 1280x720px

>>58010574
>I'm pretty sure we can sue them and make them fall once and for all.
Well said Anonymous 4 Chan user! Well said!
We should sue AMD so Intel® could produce gaming and workstation CPUs without any evil competition forcing them to release weak new CPUs to keep afloat.
I agree with anonymous from 4 Chan here, Intel® and Intel® Xeon® products are the one you can truly trust, for 4 Chan, by 4 Chan.

File: smug.jpg (62KB, 600x327px) Image search: [Google]

62KB, 600x327px

>>58010668
Anon, you deserved a (You) for this post.

>>58010615
>I only want something so I can meme and feel good about myself and because someone else has it and I saw them have it
>I dont actually even need a pc
okay

why would amd sell their product for less than intel does

you're all retarded, especially because you think the brand of number crunching cpu counts for anything

>>58010730
>happy with four decent cores
>HAHAHA LOOK HE DOESNT EVEN NEED A PC

What?

>>58010736
>why would amd sell their product for less than intel does
because that's what they've been doing for since
forever?

File: 100 samples.png (43KB, 498x212px) Image search: [Google]

43KB, 498x212px

>>58003821
>>58010040
>>58009565
since we're doing 100 sample runs

File: 1456650219571.png (150KB, 359x414px) Image search: [Google]

150KB, 359x414px

>>58007319
It was stated as quad channel and it wouldnt have made any difference if they did. Quad channel is utterly pointless outside of multi CPU systems. The only reason the 6900k has it is because it's just a xeon with different microcode. A single 6900k simply doesn't have the throughput to saturate it's memory bandwidth on dual channel.

The Zen CPU shown is a desktop model. Zen server CPUs will be multiple dies on a MCM and will have up to octo channel memory.

>>58010730
>everyone needs 8c/16t

kekd

File: 1457015558156.png (214KB, 600x452px) Image search: [Google]

214KB, 600x452px

>>58010850
There will be cheaper lower binned models with as few as 4 cores 8 threads and MCM monsters for servers with 32 cores 64 threads

>>58010867
I know, that wasn't my point though.

>>57986772
AMD is based God of cpu game. Intel on suicide watch.

INTEL BLOW THE FUCK OUT

Little bit confused here, did Zen get renamed to Ryzen?

>>58011090
>ryZEN
Ryzen CPUs are based on Zen architecture.

>>58011090
Yeah I was confused because the name Zen was spammed for years and suddenly it's RYZEN IS HERE LADS. Zen is like Intel's Core architecture I think and Ryzen is the actual product name.

I'm guessing that the server variant will have its own name like Intel's Xeon.

>>58011101
>>58011122
Fair enough, makes total sense. Thanks!

>>58011090
AMD skipped Zen and released Zen+ called Ryzen instead

>>58009565
You can accelerate the render by a lot by changing the tiles' size in the Performance section.

Whats the best cpu for windows 7?
I want to upgrade but don't want to use windows 10.