AMD is trying to re-enter the server market segment, Intel continues

File: horse.gif (182KB, 352x288px) Image search: [Google]

182KB, 352x288px

>>61003313
>I'm a blatant Intel shill.

>>61003313
Given that Intel's new offering is going to end up being a 100k USD, 800 watt reactor just to compete with AMD's more coars I don't think that they have much of a future in HPC.

>>61003313
Every single sale AMD makes is lost revenue for Intel, they're shitting themselves right now, and lying to their investors about future guidance.

AMD has already succeeded in re-entering the server market segment. The reality of the situation is that intel is desperately trying to stay relevant, now that they've been beaten by both ARM and AMD

>>61003414
>be AMD
>make 12 core CPU
>forced to sell it at the same price as a quad core Intel with no hyperthreading because otherwise no one would buy it

>>61003466
>be Intel
>make 500 watt CPU
>burn all your users and end up with zero marketshare

>>61003466
4 cores? Servers don't even sneeze on 4 cores.

>>61003313
Meanwhile >>60995716

Good or bad you gotta say that amd's shenanigans has caused intel to innovate....if thats what you'd call it.

File: 2sperf.png (65KB, 620x326px) Image search: [Google]

65KB, 620x326px

>>61003466
>be intel
>second holocaust

>>61003313
>what are opterons

kill yourself

File: latency-pingtimes-7900x.png (110KB, 831x537px) Image search: [Google]

110KB, 831x537px

CRAWLING IN MY SKIN
THESE WOUNDS THEY WILL NOT HEAL

>implying AMD and Intel can keep up with IBM POWER

File: is-there-antitrust-inside-intel-25-728.jpg (108KB, 728x546px) Image search: [Google]

108KB, 728x546px

Careful Intel. The industry is watching you sling slime.

>>61003313
*autistic screeching*

>be intel
>make a even hotter processor than powerpc 970
>charge a arm and a leg for it
>doesn't include sub zero cooler

>>61003313

An X299 RAID decryption key ($300 MSRP) has been overnight-shipped to your home address.

Thanks again for shilling for Intel.

>>61003313
EPYC
D O A
O
A

>>61003313
Lulz. Innovation and success?

Let me blow your mind...
>Itanium

Real 64bit computing is courtesy of AMD. AMD invented modern server platforms with their Opterons.

Anyone remember the post I made a while back when I said that intel would start pushing 'reliability' and 'trustworthiness'?

>>61003652
>implying that intel shills get good compensation for their shilling
Intel gives them pin badges and pens.

File: sqSFmXVI9t3V1sshTNPZzRykNWPWBo5ki6NbHkv6QQg.png (177KB, 1024x427px) Image search: [Google]

177KB, 1024x427px

at least i found a better way to spend my money

>>61003668
yes anonymous user I do remember you specifically

>>61003541
>innovate
Presshot 2.0: gigavectors edition is not a fucking innovation.
>>61003668
Yes.

>all these anons who don't realize that this is a copypasta of Intel's response to Epyc

>>61003313
is this another copy pasta to add to my collection?

>I-I swear i-it's stitched together! Like with glue a-and shit! It's not even a real processor!

I would pay good money to see a live reaction video on Intel CEO's face when watching the AMD press conference

https://www.youtube.com/watch?v=5MPlS8RF0zE&feature=youtu.be

>>61003713
Dunno about video, but the sound would be an "Oy vey!" every time a new slide is shown.

>>61003713
well they did release an awesome statement

Intel issued a statement, distributed via email, saying that it would continue to top AMD’s efforts with Intel’s “Xeon” server processors.

http://www.barrons.com/articles/amd-reveals-epyc-details-intel-vows-to-top-it-1497997334?mod=yahoobarrons&ru=yahoo&yptr=yahoo

>>61003790
>http://www.barrons.com/articles/amd-reveals-epyc-details-intel-vows-to-top-it-1497997334?mod=yahoobarrons&ru=yahoo&yptr=yahoo
>other deployment complexities
Ah yes, because that processor will require configuration to use.

>>61003790
oh god intel is getting better and better on making the best worst decisions ever

>With our next-generation Xeon Scalable processors, we expect to continue offering the highest core and system performance versus AMD. AMD’s approach of stitching together 4 desktop die in a processor is expected to lead to inconsistent performance and other deployment complexities in the data center.

i guess mesh turned into a mess xD

Is this what it was like when fx-55 spanked jewtel? Are they about to throw an anticompetitive tantrum?

>>61003833
well yeah they cant stitch cores together
also it will be fun to watch intel stock today its already on a downfall

>>61003335
>yfw this was actually Intels public statement

AMD just announce project EPYC, a 32 core processor aimed for servers. It's the size of a sauce plate.

>>61003335
This was literally Intel's official statement about today's rape, it's already a meme and it will be spammed for weeks
At last I'm doing that

Try to make it less obvious next time.

>>61003313
INTEL IS SO BTFO, THEYVE RESORTED TO SHIT FLINGING.

>>61003313
>That entire statement
Do Intel not realise that data centres are run by people who KNOW what they're doing?
Do they think this pathetic attempt at damage control is going to do anything?

>>61004402
That's the thing with Kikes, They believe the goyim are actually this dumb.

THEY ARE SO FUCKING ASS MAD THAT THEY DON'T HAVE -ANYTHING- TO COMPETE WITH INFINITY FABRIC

File: Herriman-Fire-2.jpg (66KB, 720x480px) Image search: [Google]

66KB, 720x480px

>>61003555
A HEART THAT CANNOT FEEL

HOPING THAT THESE HORRORS WILL RECEDE

HOPING THAT TOMORROW WE'LL ALL BE FREED

SICKNESS TO INSANITY

PRAYER TO PROFANITY

DAYS AND WEEKS AND MONTHS GO BY

CAN'T FEEL THE HUNGER TOO WEAK TO CRY

File: 2B7ED83800000578-3203576-Clinton_insisted_on_Tuesday_that_nobody_asks_her_about_her_email-a-4_1440072659239.jpg (47KB, 634x354px) Image search: [Google]

47KB, 634x354px

>>61003710
>Infinity Fabric? Like a cloth or something?

File: 86479.png (50KB, 250x250px) Image search: [Google]

50KB, 250x250px

>>61003313
>I know! Let's tell them bullshit! Everyone trusts us because we're Intel and we know best.

>>61003790
desu what other statement can they release? "We are sorry, we totally didn't see it coming and shitted ourselves"

>>61003313
Intel are already in denial mode from AMD's 1P Holocaust

i was expecting amd show to start with this
https://www.youtube.com/watch?v=XjRaKId4Rps

cant find a better song to what they are doing to intel

>The guitar riff as HPE rep walks on
nigga

>>61003466
If you should take note of something, it's the fact that AMD CAN do that. Intel can't. Their chips are expensive as fuck to fab.

>>61003313
>Intel is this butthurt and scared they are literally resorting to spreading FUD about a competitor

>>61003313
>AMD’s approach of stitching together 4 desktop die in a processor is expected to lead to inconsistent performance


Meanwhile
>I won't lie, the past two weeks with the Core i7 7900X and the X299 boards I have tested have been an absolute challenge. Memory XMP profiles would not stick, power consumption with one BIOS was OK, the other through the roof. But most of all the processor performance was all over the place. We've seen perf differences of up-to 20% in-between merely different motherboards. MSI however with it's latest BIOS seems to have found equilibrium. Now the biggest fight the two weeks was ironically game performance, it was severely lacking.

>20%

>>61003466
>be amd
>realises that they have 80% yields
>su literally thinks world domination
>7nm is coming with even greater yields
>su is literally changing into SUper SUyan
>amd literally takes the userbase like ostara took the spring in american gods
>amd bankrupts and buys intel
>su turns her head and lays down a sight on nvidia
>su burns the leather jacket guy
>su finally dominates the world

>>61004889
>>amd literally takes the userbase like ostara took the spring in american gods
WOW SPOILERS M8

>>61004402
It's aimed at non tech savvy investors to stop them dumping too much intel stock.

>>61004903
One one hand
>The ENTIRE server industry on a stage with AMD, saying how epic EPYC is

The other
>Intel saying "d-don't worry, i---i-i-it's fine!"

Gee, how to inspire investor confidence

>>61004902
oh fuck off who doesnt expect from amd to kill intel?

>>61003816
literally /g/ troll level arguments

>>61004475
trying to use FUD aimed at enterprise folks will not work, they're not dumb /v/ kids

they're literally being childish

>>61003691
retard spotted

>>61006194
Why?

In the far flung future, Lisa Su and Hitler will be remembered as the same person

File: 14979924598501488569071.jpg (3MB, 3968x2976px) Image search: [Google]

3MB, 3968x2976px

>SUSE support

Welcome to my home, AMD

>>61006547
>No gentoo
/g/ is not going to like this

>>61003313
>Xeon CPU architecture is proven and battle tested

At least downconvert the product brochure to something someone would actually say you hopeless fuck.

>>61003313
Bank of America downgrades Intel

https://seekingalpha.com/news/3274716-bofaml-downgrades-intel-amd-epyc-reveal-shares-1_6-percent

Intel is trying to re-enter the server market segment.

>>61006568
intel said exactly this, dumbie

http://www.barrons.com/articles/amd-reveals-epyc-details-intel-vows-to-top-it-1497997334

File: 1492463302260.jpg (205KB, 1600x917px) Image search: [Google]

205KB, 1600x917px

>>61006580
EVERYTHING IS FINE

Meanwhile $AMD is up 11% while $INTC lost 6 billion in market value overnight.

>>61006609
make that 12%

>>61006547
GCC logo is so CUTE

>>61006622
MAKE THAT 13%

THEY'RE GOING TO DO IT
THE MADMEN
THEY'RE GOING TO SMASH 13.50 PREMARKET

>>61006612
>skylake-x
>6c up to 10c

>then they announce all the way up to 18c because amd caught them sleeping
>12c+ will not be available until next year
>threadripper will cost less than $1000 for the 16c/32t and perform better than intel's 18c/36t

best timeline

>>61006609
source, good man?

>>61003659
>Real 64bit computing is courtesy of AMD. AMD invented modern server platforms with their Opterons

Intel shills REEEEEEEEEEEEEEEEEE


>muh hyperthreading

Back before Intel was anything other than a fucking spec on your windshield. And I'm even running an Intel rig. But you know they're blowing smoke with this new release.

>>61006636
Nigga AMD's single socket 32c/64t 7551p only costs $2100

>>61006641
http://www.google.com/finance?q=NASDAQ:INTC&ei=XGZKWamYOdTDsAGA64qIAQ

File: 2017-06-21-142730_21x30_scrot.png (1KB, 256x256px) Image search: [Google]

1KB, 256x256px

>>61006628
C U T E

>>61006641
literally any website with real time stock prices

>>61006641
https://finance.yahoo.com/quote/AMD/
https://finance.yahoo.com/quote/INTC

>$13.52

WE DID IT BOYS
WE GOING TO THE MOON

File: 62f.jpg (36KB, 630x630px) Image search: [Google]

36KB, 630x630px

>>61006657
>>61006674
>>61006646

thank you very much, /g/entleman.

also
>REEEEEEE ANTI SEMITE

File: AMD_Reveals_‘Epyc’_Details;_Intel_Vows_to_Top_It_-_Barron_s_-_2017-06-21_02.16.39.png (324KB, 1191x1587px) Image search: [Google]

324KB, 1191x1587px

kek it's real. Intel in total damage control mode.

7.12% PM

8.00% boys

Someone get me a fucking spacesuit.
It's hitting $14

>>61006707
i bet this was written by the tripfag

>>61006813
maybe he was actually a high ranking paid shill

Seems like it'll stabilize at $13.7 before AH

File: 1482392577750.gif (2MB, 695x392px) Image search: [Google]

2MB, 695x392px

IT'S OVER INTEL IS FINISHED AND BANKRUPT

>>61006707
>Their shit sucks plz buy our $9k xeons goyim

>>61006883
more like $12000 for the good ones

>>61006896
12000 dollar cpu? what 15000 dollar cpu are you talking about? please buy our 18000 dollar cpus for maximum stability!

>>61003466
You know AMD is about to have yet another major flop because they pay chinks to shill on 4chan and seekingalpha lol

File: trump_smug_smile_wwe.jpg (24KB, 581x406px) Image search: [Google]

24KB, 581x406px

>>61007044
stay mad, shekelchaser.

File: 1497658837285.gif (365KB, 2000x1153px) Image search: [Google]

365KB, 2000x1153px

>>61007044

>>61007067
>>61007085

I'm not the one who pays money to shlll on an internet forum with my product being pushed too early and about to come crashing down to earth like every other product release by AMD.

Post the benchmarks if you're so proud of it. Go ahead, let's see this next gen CPU barely outperform last gen.

>>61006208
because I have no money and you just made me feel bad. you fucking retard

>>61006914
18000 dollar no, buy the 24000 dollar ones for the MAXIMUM PERFORMANCE goyim

>>61007094
>SPEC benchmarks are worthless
>literally 80% of the server market supporting them is worthless

kek

>>61007094

It's over my fellow Inteldude. Give up. The AMDbros have won.

>>61007121
>>61007129

Let's see the benchmarks if it's so impressive shilliam. I'm waiting.

inb4 you keep posting memes and refuse to accept your piece of shit has the specs of last gen.

File: 1434832613988.jpg (139KB, 300x399px) Image search: [Google]

139KB, 300x399px

>>61007044
>being this assblasted

>>61003544
>ignoring xeon e7 series on purpose
(You)
AyyMD was always competetive against i5s and other pleb tier cpus.

>>61007138
see
>>61007134
>inb4 you keep posting memes and refuse to accept your piece of shit has the specs of last gen.


Notice how AMD shills can't even post benchmarks because it's such an underperforming rushed product.

>>61007121
helps when after 11 years they decide to "updte" their benches because the branch prediction of amd is too good

>>61006598
>Company is coasting
>Company tries to use diversity hires and HR fuckery to "reinvent" itself
>Company crashes and burns
fucking clockwork

>>61007156
Notice how AMD shills can't even post the specs because it's literally a last-gen CPU being pushed as next gen.

>>61007044
amd paid seekingalpha...
AMD PAID A SITE THAT SHITS ON AMD FOR AS LONG AS I CAN REMEMBER
AMD PAID A SITE THAT 90% OF THEM ARE LONG ON NVIDIA
KEK

>>61007148
literally on the presentation they made yesterday

File: 1459857970742.gif (184KB, 500x394px) Image search: [Google]

184KB, 500x394px

>>61007148
>he thinks I'm working for AMD

I'm just here to enjoy you're fanboy tears. Fanboys are the worst.

>>61007169
Notice how AMD shills keep repeating their same paid lines whilst ignoring
>>61007164
>Notice how AMD shills can't even post the specs because it's literally a last-gen CPU being pushed as next gen.

>>61007148
>underperforming
>2h2017 rollout for 2 hyperscalers
0/10 bait.

File: 1446697892826.jpg (32KB, 720x480px) Image search: [Google]

32KB, 720x480px

>>61007176
>calls people out on repeating paid lines while literally repeating his own paid lines

>>61007148
you want benchmarks of epyc?
we have so many reviews about ryzen that we know its scales on 97.1% they dont NEED to do benchmarks we know the perfomance already

File: intel killer!!!!.jpg (27KB, 827x391px)

27KB, 827x391px

>>61007215
>>61007183
>>61007225

Hey since you guys can't do it after I've asked so nicely I'll provide it:

AMD's brand-new, paid-marketing campaign is all for this CPU

4.0 core speed. Laughable.

>>61003313
>innovation
Nope, just moar cores.

>>61007238
MOAR BINBU-

MOAR MESH YOU ANTI-SEMITE

>>61007228
Anon that's some low-effort bait.

File: Time.jpg (165KB, 685x480px) Image search: [Google]

165KB, 685x480px

>>61007121
>>61007134
>>61007144
>>61007148
>>61007164
>>61007228
>8am
>Intel Shills misteriously appear
Really made me think

>>61007228
dear god odin throw a thunder to this infidel DEUS VULT his ass

he is stupid he doesnt even realise his own stupidity

File: Threadripper Specs Cut Down.jpg (67KB, 590x332px) Image search: [Google]

67KB, 590x332px

>>61007225
>>61007215
>>61007183

https://www.pcgamesn.com/amd/amd-ryzen-threadripper-specs

Here's another spec.

>Benchmark scores for AMD’s Ryzen Threadripper 1950X reportedly appeared on Primatelabs’ database yesterday before being abruptly taken down. The 16-core variant managed a 4,167 single-core score and a 24,539 multi-core score, but it’s not certain whether these scores are based on a performance benchmark or just a run to see if everything’s stable. It also doesn’t mention whether SMT was turned on.

>Provided the screenshot of the database listing is indeed real, and the database was genuinely listing a Threadripper, we’ll be looking at a 3.39GHz clock speed for the chip - whether that’s a base or boost clock isn’t made clear. It also shows the 1950X as having 32MB of L3 cache and 8MB of L2 cache.

This kills the AMD shills.

>>61003313
I see they're using the deflection type damage control

>>61007259
That's even worse than the last one. I think you should go back to /v/.

>>61007259
amd sandbagging the industry cause they want to ass rape blue smurf asses?

you know kinda what they did with ryzen and today with epyc?

CALL ME SHOCKED

File: 1443776271440.gif (453KB, 255x144px)

453KB, 255x144px

>>61007259
Here you go faggot.

https://www.heise.de/newsticker/meldung/AMD-Epyc-legt-los-3748615.html?artikelseite=2

>>61007259
>>61007228
>>61007148
>>61007094
>>61007044
Hi JIDF, how's Chaim doing?

>>61007293
>>61007274
>>61007261
https://www.pcgamesn.com/amd/amd-ryzen-threadripper-specs

>Benchmark scores for AMD’s Ryzen Threadripper 1950X reportedly appeared on Primatelabs’ database yesterday before being abruptly taken down. The 16-core variant managed a 4,167 single-core score and a 24,539 multi-core score, but it’s not certain whether these scores are based on a performance benchmark or just a run to see if everything’s stable. It also doesn’t mention whether SMT was turned on.

>Provided the screenshot of the database listing is indeed real, and the database was genuinely listing a Threadripper, we’ll be looking at a 3.39GHz clock speed for the chip - whether that’s a base or boost clock isn’t made clear. It also shows the 1950X as having 32MB of L3 cache and 8MB of L2 cache.


3.39 core clock speed.

File: MV5BMTkwODk2ODM4MV5BMl5BanBnXkFtZTYwNjk1NjMz._V1_UX214_CR0,0,214,317_AL_.jpg (16KB, 214x317px) Image search: [Google]

16KB, 214x317px

>>61007300

About to patent a very common human gesture, and you?

>>61007300
>>61007298
https://www.pcgamesn.com/amd/amd-ryzen-threadripper-specs

>Benchmark scores for AMD’s Ryzen Threadripper 1950X reportedly appeared on Primatelabs’ database yesterday before being abruptly taken down. The 16-core variant managed a 4,167 single-core score and a 24,539 multi-core score, but it’s not certain whether these scores are based on a performance benchmark or just a run to see if everything’s stable. It also doesn’t mention whether SMT was turned on.

>Provided the screenshot of the database listing is indeed real, and the database was genuinely listing a Threadripper, we’ll be looking at a 3.39GHz clock speed for the chip - whether that’s a base or boost clock isn’t made clear. It also shows the 1950X as having 32MB of L3 cache and 8MB of L2 cache.

3.39 core clock speed.

>>61007306
this was already proved fake

those multithreaded scores are similar to one 1700

>>61007325
Let's see some competing benchmarks then


Oh right AMD wont put them out because it's pathetic.

>>61007306
>tripfag asks for epyc benchs
>proceeds to btfo us with threadripper passes
>next in line is going to find an fx95xx to showcase how good intel is doing with their tdp

File: 1381579125993.jpg (448KB, 3840x2160px) Image search: [Google]

448KB, 3840x2160px

>>61007306
>>61007318
Kek I think we just triggered Pajeet into rage mode.

>>61007331
>16 core cpu would perform exactly the same as 8 core on the same clockspeeds
right

>>61007331
hahahahhahahahahahaahahahahahahahhaahhhahahahahahahaha oh wow

File: 1423432148132.jpg (12KB, 251x242px) Image search: [Google]

12KB, 251x242px

>goddamnit boss i demand more rupees for this bullshit. they're all making fun of me! it's getting on my nerves..

>>61007339
>>61007442
OH LOOK. A SAMEFAG

File: 1432817901948.jpg (59KB, 395x401px) Image search: [Google]

59KB, 395x401px

>>61007456
Nice argument Rajesh

>>61007339
>>61007442

woah this one >>61007442 hits close to home, poojeet?

>>61007487
meant >>61007456

>>61007259
The day when all of india gets constipated draws nigh

>>61007456
>OH NO, THE GOYIM KNOW BETTER MAKE A NEW BREAD

>>61007278
>>61007278
>>61007278

>>61007540
Just report the shitposter, he got banned more than a few times.
If enough people report, he gets banned automatigically.

>>61007547
lmao why would you ban such a miserable litte shit? it's just too funny reading pathetic shitpost attempts

>>61003313
Xeon used to be nice. Today they're just intel core without GPU

>>61007576
Because he spams the board and it's annoying.

>>61003651
>>charge a arm and a leg for it
Had to read it 3 times before understanding that you meant the body part and not the architecture

>>61007618
this>>61007618

>>61007610
>amdtards shitting the board with wojak threads for the last 12 hours is ok

>>61007641
at least they have justification with AMD's new EPYC server lineup.

Intel is just cooking in their self-created oven.

>>61003313
> data center innovations
> broad ecosystem investments
> battle tested

MEANINGLESS BUZZWORDS

YOU ARE SAYING NOTHING YOU STUPID FAGGOT

KILL YOURSELF

>>61008075
This is an actual statement from Intel. They're digging their grave deeper and deeper.

File: 14624568450319.jpg (941KB, 1906x1536px) Image search: [Google]

941KB, 1906x1536px

>>61007044
>>61007293

>>61008094

That letter is not meant for weebs on a Tibetan cross-stiching board. It is intended to reassure investors in the company that can easily have millions at stake. As good as Epyc is, Intel has been there and done that against many other competitors, including AMD itself, and has come out on top. That is not a bad thing to say to people that are trusting you with millions.

>>61008405

That being said, how is the statement wrong even technically? Going off die within the MCM is going to be way slower than staying on die. If AMD could fab a chip as large as intel, that is what they would have done. Intel is going to retain the per-thread cpu performance lead, without even considering the 2x boost they will get from avx512.

The only real advantage I see AMD having is that they will have 8 channels of memory to intel's 6. Looking at Epyc's low clock speeds, I don't know how much that is going to matter. How many people are actually going to provision a server with the full 2TB of ram? And all those ram channels and huge package must mean expensive motherboards.

Random observation:
I went looking at AMD supermicro servers last night, and was amused that none of the new ones seemed to come with 10gige. Why? I have no fucking clue. I consider that a mistake.

>>61008552
It remain to be seen if amd infinite fabric does what it's supposed to do. If so it's a great contender. Amd isn't supposed to flip the market but compete on it.

About avx, you really think it make that much of a difference? I mean software that fully utilize avx is scarce, and even if a load could sustain avx usage it's very power hungry.

>>61008618
>About avx, you really think it make that much of a difference? I mean software that fully utilize avx is scarce,

People that want avx are compiling their own codes, so I don't see software availability being a problem for avx.

>>and even if a load could sustain avx usage it's very power hungry.

Yes avx may have high power usage, but if that comes hand in hand with high throughput, then who cares. It is joules/opp. Intel created it to have their cpus compete with GPUs and the huge efficiency advantage they have over cpus. AMD talks about how their lack of avx is not a problem because you will be going GPU compute. They neglect to mention how expensive it is to push and pull data in and out of GPU ram is.

I have yet to see a lot of benchmarks about avx, but iirc the next huge american supercomputer is going to be using xeon phi's with their atom cores and avx-512. That decision would have been made on technical merits, including throughput and efficiency.

I expect within the next year we will start to see some good enough benchmark results.

>>61003816
are they srs? they are offering bunch of cpu's on single mobo and claim it as advantage, yet somehow multiple dies on single cpu is bad?
>damage control

>>61006645
enterprise boards are quite expensive, also you need a ton of RAM

>>61007144
They're comparing chips based on price bracket; Xeon E7 8xxx CPUs cost $8000+ a piece and comprise a tiny fraction of the market where you NEED maximum scale-up performance from a single server and the price of such chips is the least of your expenditures.

>>61007044
what the fuck is seekingalpha? stop projecting you fucking shill, intel lost and it's over, it was about time they fuck off

File: image.jpg (206KB, 630x389px) Image search: [Google]

206KB, 630x389px

>>61003313
>stitching together 4 dies in a processor
You mean like some sort of mesh?

>>61009251
yeah a mesh that actually has the uncore tied into the ram speed and its called infinity fabric unlike something else..

>>61007044
OYYY VEYYYYYYYYYYYY

>>61007044
Shai pls stop

>>61007094
Go ahead, let's see this next gen CPU barely outperform last gen.

But anon, I thought you were defending Intel.

>>61009524

Imagine zen+ supporting 4000mhz DIMMS across each segement - it would make the zen+ Epyc chips blisteringly quick at such huge core counts - not that I expect ECC DIMMS to ever be that fast.

>>61010319
no such thing as zen+

zen 2 is the next step

>>61004367
sauce?

>>61010908
>http://www.barrons.com/articles/amd-reveals-epyc-details-intel-vows-to-top-it-1497997334
>We take all competitors seriously, and while AMD is trying to re-enter the server market segment, Intel continues to deliver 20+ years of uninterrupted data center innovations while maintaining broad ecosystem investments. Our Xeon CPU architecture is proven and battle tested, delivering outstanding performance on a wide range of workloads and specifically designed to maximize data center performance, capabilities, reliability, and manageability. With our next-generation Xeon Scalable processors, we expect to continue offering the highest core and system performance versus AMD. AMD’s approach of stitching together 4 desktop die in a processor is expected to lead to inconsistent performance and other deployment complexities in the data center.

>>61010947
>In response, Intel said through a spokesperson
stopped reading there

>>61010985
>intel is a person

>>61011006
>he missed the point

>>61011013
explain

>>61006609
I was poor back when $1.6. I was poor back when $10. I'm still poor now and I hate it.

>>61007110
actually for 24000 dollar its maximum performance
MAXIMUM PERFORMANCE is 25000 dollar

>>61003313
It's gonna be a lot of clever marketing, kicking screaming, and convincing departments but if they can undercut Intel for the same performance, price and efficiency then as long as they're reliable, its only time before AMD snatches the datacenter market.

>>61012719
>same price
more like half to 1/3 anon, that's how much intel is jewish on the server market

>>61012719
>its only time before AMD snatches the datacenter market
They already did
Literally everyone was there, the only guys missing were the HPC folk from Cray and SGI, probably because they haven't made something that allows 8P EBYN boards

File: ebycspurdo.png (555KB, 801x1500px) Image search: [Google]

555KB, 801x1500px

wooooooooo intel losing wooooooo yayyyyy

>>61009251
Wonder how long it took them to bang that concept out, looks like a fucking mess.

>>61015452
It's just a bunch of smaller ringbuses instead of only a single big one

too bad their fabs are shit and it performs worse han one ringbus

>>61015497
>it's actually worse
JUST

>>61008552
>Looking at Epyc's low clock speeds,
Uh, exactly how high do you think a Xeon clocks?

>>61015661
he's just talking out of his ass while pretending to be reasonable

>>61015452
>Wonder how long it took them
It's a variant of KNL. It was just adapted.
>looks like a fucking mess
In SKL-X, it manages cross-core comms without the horrific latency penalty found on Ryzen cross-CCX communication.

>>61003713
I'm rather interested when they're going to fire Krzanich's sorry ass since he's responsible for Intel sitting on their asses for years in the first place.

>>61015707
the problem it's that the latency between all cores on skl-x is equally shitty, while on ryzen the cores inside the same CCX have small latency

File: react.png (492KB, 614x578px) Image search: [Google]

492KB, 614x578px

>>61003313
Somebody is desperate