My dick https://www.servethehome.com/amd-na ples-competitio

>>59291281

https://www.youtube.com/watch?v=ZtYTPcshyA0

https://www.youtube.com/watch?v=PN93G6Rg2ek

>>59291281
ALL DOSE DIMMS ON A 2P BOARD

File: mpv-shot0012.jpg (121KB, 1280x720px) Image search: [Google]

121KB, 1280x720px

>>59291332
HOLY

>>59291281
More I/O, more bandwidth, more memory, more cores, lower package and platform power.


Intel is gonna have its hands busy phoning HP and Dell

>>59291281
I'd suck several dicks for this machine

>>59291430
I looks like AMD is putting there fist up intel's ass.

post screenshot there intel server gets "out of memory" error.

File: 1488921553589.png (21KB, 653x174px) Image search: [Google]

21KB, 653x174px

>>59291281
Reminder.

http://www.pcworld.com/article/3168319/components-processors/intel-demotes-pcs-giving-datacenter-chips-first-crack-at-new-technologies.html
https://winbuzzer.com/2017/02/13/intel-demotes-pcs-favor-data-center-first-strategy-xcxwbn/

File: Screenshot from 04-03-2017 03:47:54.png (54KB, 1489x889px) Image search: [Google]

54KB, 1489x889px

>>59291513
They likely just gave up when they saw that Zeppelin can pull under 35W at 2.0GHz clocks at 8 cores while Broadwell EP's are 145W TDP parts 2.4-2.6GHz

>>59291513
It's time to innovate and stop relying on shrinks Intel, I'm sure you can upgrade that P6 core.

how does it compare to i5 in 720p gaming?

>>59291609
>I'm sure you can upgrade that P6 core
>Not reviving Tejas and Jayhawk
Pleb.

File: chrome_2017-03-07_21-01-56.png (161KB, 397x156px) Image search: [Google]

161KB, 397x156px

>>59291281

Wow, look these capacitors! I never seen a capacitor like that.

wasn't the zen architecture designed with data centers as the main priority? Certainly looks like it may be able to gain some following, as it stands AMD is pretty dead in the server space.

>32 cores
how will intel even compete?

>>59292700
They're trying to push out a 32 core Skylake Xeon with 6 channel memory.

>>59292740
amd will be 85% performance per core at half the price

File: UWF1V470MCL1GS.jpg (22KB, 362x372px) Image search: [Google]

22KB, 362x372px

>>59292649
You've never seen surface mount electrolytic capacitors?

File: 1488910643715.png (246KB, 882x758px) Image search: [Google]

246KB, 882x758px

>>59291281
DELET THIS SIR

>>59292790

As this size? No.

Intel currently has an approximate 99 percent server monopoly

Any AMD movement into the server market is going to make Intel look bad, cut into Intel profits and make AMD more money

Most have missed the point of Ryzen. The FX line was so shit as a CPU, AMD could only go up in profits or go out of business, and Ryzen isn't shitty enough to kill AMD.

Regardless of the complaints about Ryzen launch performance AMD is going to ship more AM4 cpus and server chips in 2017 than they did AM3+ cpus in 2016

>>59292649
>>59292842
What are you even on about anon?
They're SMD caps. On a motherboard. There's nothing special about this at all.
What is your joke? Is there a joke? Is the joke that you don't have one?

File: 1488926684742.jpg (13KB, 150x200px) Image search: [Google]

13KB, 150x200px

>>59292958

Damn, relax! Oo

I'm talking about the marked capacitors circled in red.. They are much larger than the normal capacitors present on the motherboards, such as the smaller capacitors next to it.

>>59291513
Reminder of what? Intel incrementalism?

Naples is coming Q2, Intel is NOT conceiving, designing, and manufacturing something which can compete with Naples by Q2 2018, let alone Q2 2017.

>>59293218
Comparing them to the thickness of the PCI-e slots, yes they're big, but looking at my machine it has caps as thick at the PCI-e slots, so something maybe 1.5x the thickness, in a high end server board, isn't that bizarre.

>>59291281
is amd gonna release a atom server version?

>>59292910
AMD has probably sold more processors this week alone than during the bulldozer launch up until january 2017.

>>59293783
Zen is capable of scaling down to low power envelopes and scaling up to high performance. There's also K12 which is like Zen but with ARM instructions

File: Ps3oiEn.png (790KB, 2015x703px) Image search: [Google]

790KB, 2015x703px

Ouch

>>59292768
Price doesn't matter. It's just that Zen is league ahead in terms of perf/watt at <3.3ghz.

>>59294937
How will Intel ever recover?

File: 1458440007273.gif (2MB, 200x180px) Image search: [Google]

2MB, 200x180px

>>59291281

>64 cores / 128 threads
>512GB RAM with 170 GB/s bandwitch
>4TB of NVM SSD
>compile gentoo, linux kernel, QT and all other shit software at the same time
>all in a few minutes
>multiples VMs
>multiples compilations instances

>>59295222
You could compile gentoo in how many seconds?

>>59294937
>insufficient memory to load
when will RAMlets ever learn?

>>59292549
>reviving Tejas and Jayhawk
They want to compete, not nuke their customers datacenters.

>>59291281
>2017
>visible wires

nope

>>59293951
Idk I want x86 and a C2750 amd version

>>59295419
>Has never seen the inside of a server
K

>>59295385
That's the only way. Scorched earth. They'll leave no datacenters for AMD to sell their Naples.

>>59295440
It's more likely we'll see a competitor to the Xeon d-1500 series.

File: 1488778759330.jpg (25KB, 320x320px) Image search: [Google]

25KB, 320x320px

>>59291281
>Octachannel DDR4
HOLY FUCK

>>59295448
Have you seen the inside of a Sun/Oracle or Fujitsu server from the past 10 years?

No wires.

Everything is on mezzanine cards that slide in and get secured with tensioned screws on springs.

>>59295500
Sure, but those aren't x86 servers. They're carefully designed systems that aren't meant to be compatible with other parts (I know from experience).

>>59295458
I'm pretty sure that those datacenter owners will sue Intel to hell and thus side with AMD forever.
>inb4 AMD having datacenter monopoly because of Intel's stupid Scorched Earth tactics

>>59295235

I don't know. kek

>>59295530
Oracle and Fujitsu both have x86 servers that are designed like that.

OPENCAPI WHEN

holy fuck, A M fucking D

>>59295235
Maybe 15-20 seconds and that's with some major overhead

>>59291999
the trips have spoken.
Can't have a good server without that low res gaming performance

>>59295448
I want this "wires are ugly" trend to end. Black liquorice cables are uglier than the old colour coded ones. Tidy cable management looks great.

File: Jim Keller.png (400KB, 816x1020px) Image search: [Google]

400KB, 816x1020px

>>59291281
You haven't even seen my final form.

>>59291430
>Intel is gonna have its hands busy phoning HP and Dell

And that's exactly why AMD won't manage to sell this.

Any company buying these will get massive penalties from Intel for desktop chips, mobile chips, Atom chips, other server chips, SSDs, Phis, etc.

Maybe they'll end up being fined for 2 billion € this time, but that's standard operational cost when you make 50 billion $ profit a year.

>>59296720
No matter how good will Intel bribe HP/Dell that p/watt is just TOO delicious.

>>59296720
I bet the accountants are all fired up. Maybe the savings from AMD will offset the penalties.

>>59296720

They don't need to dominate the server market.

They just need to crack it open. If in a year AMD can get 5% in the first year. That's when the REAL fun begins

>>59296872
DELET

>>59296720
>Intel for desktop chips
Killed by Ryzen

>Mobile Chips
Killed by summit ridge

>Atom chips
Never even alive

>>59296903
>Killed by summit ridge
Raven Ridge.

when does intel release snippers cove ?

>>59296921
Thanks for the correction

>>59296755
personally, if amd was this far ahead, and I owned a business i'd tell intel exactly what they are missing out on, and if they want to be cunts, tell em to get fucked and go amd.

unless intel has something major, and they don't, amd will be a viable, competitive alternative for the next 10 years, for a period of time in that 10 years they may not be the best solution, but viable nonetheless. I would then make all the business dealings very public with the message 'You want to fuck me, well I got news for you, I got a cock too'

That would immediately cut all ties to intel we have, go full amd, and then send out papers to everyone we have and don't have business with just running benchmarks comparing intel systems to amd ones of equal value. If there was a good chance I would no longer have to put up with intel's extortion ever again, I would do it in a heartbeat.

>>59291281
Dose lanes

>>59291281
Can it get 60FPS on crysis?

>>59293796
likely sold more to consumers in the short launch weak then they did all consumer bought cpus sense bulldozer, not 100% sure on oem sold ones though, they still may have done better there.

>>59296872
>look guys we're faster than a system that's starved of memory (which is obviously Jewtel's fault)
????

>>59295137
first, intel needs octa channel memory
second they need a low power architecture...
then they need to realize they failed so hard trying to make one for the last 10 years they can barely give their lower power architectures away for free...
then they need to bribe oems and hope oems are ok with the extortion...

prey to whatever god they believe in that they can make a new arch that isn't shit.

>>59297965
Lets be frank though: Do you (hell does anyone) see them moving away from their P6 derived architecture anytime soon?

IIRC they've tried at least twice (Itanium, Netburst) to pull away from P6 for a mainstream core, and both were failures. There's also the fact intel cannot for the fucking life of them make decent GPUs, which in turn locks them out of the GPGPU department (and no Xeon Phi doesnt count.). The only really low power architecture they've done was Atom and we all know that Atom was a slow as shit shitpile only good for netbooks and other extreme low power systems.

Their current arch with its fuckhuge FPU is not helping their position either.

>>59298078
which is why literally 8 weeks ago the ceo got his thumb out of his ass and ordered a new uarch. 2020-2022 is when it will be out, possibly sooner if intel gets deep dicked in every single market they once had a strangle hold on.

>>59298098
>Brian we found a solution
>MAKE LONGER PIPELINE

>>59298098
Huh. Still, 2020-2022 for a actually brand new arch. Between now and then they still have to use P6-derived cores, and all the while AMD is going to be turbo-shitpunching them with Naples and its successors.

What's not helping Intel's situation either is that they insist that their top-end server silicon be monolithic. These dies are >500mm^2 (their largest was almost 700) and are very hard to get any decent yields out of due to their size. Naples on the otherhand uses MCMs, which while they do have their downsides, are far cheaper to manufacture as AMD can pull any 4 Zeppelin dies, slap them on a package, and sell it. They're even SOCs so external silicon for PCI-E, disk control, and other I/O is no longer required, further saving costs.

These coming years will be VERY interesting.

>>59298098
2022
>6 years too late

>>59298236
And I'd like to add, this is all assuming intel doesn't fuck it up AGAIN.

>>59298170
They were gonna try that with Tejas and Jayhawk. 50 pipeline stages and 7ghz does not a powerful CPU make.

>>59296720
>desktop chips
Replaced by Ryzen (Summit Ridge and Raven Ridge)
>mobile chips
Replaced by Ryzen (Raven Ridge) and Qualcomm+Mediatek
Friendly reminder that Intel abandoned their efforts at penetrating the smartphone market
>Atom chips
Replaced by Naples/Snowy Owl
>other server chips
Replaced by Naples/Snowy Owl
>SSDs
There's dozens of SSD manufacturers pushing shit equally as good as Intel, sometimes better and always cheaper
Optane was Intel's try at dominating the SSD market but it's a massive failure that's in development hell with 2+ years of delay
>Phis
Radeon Instinct rapes this, along Naples
>>59291513
>>59293259
Reminder that Intel will start a new era, a shitfest of bugs that will rival the FDIV bug, they literally have no validation teams
>>59292910
>Any AMD movement into the server market is going to make Intel look bad, cut into Intel profits and make AMD more money
AMD singlehandedly destroyed Itanium/EPIC, despite Intel convincing every other vendor to abandon their server lineups
This is just as disruptive as Opteron was in 2003
>>59298098
>2020-2022
That's 3-5 years, with no real competition against Zen
And then Zen+ will already be out, with 15% improvement IPC wise, possible Zen++ would be out, even 7nm would be the norm by then
Intel is literally becoming 2011 AMD
And that's not counting all the extravagant IP that AMD has been developing the last 5 years

>>59291281
Dick status: diamonds

>>59291352
That's a pretty normal amount of DIMM slots for a 2 socket board.

>>59291566
>Broadwell EP's are 145W TDP parts 2.4-2.6GHz
8 cores?

No.

>>59295222
>512GB RAM with 170 GB/s bandwitch
It's going to need to be able to address at least triple that amount to be competitive.

>>59298320
This is why you fear AMDs engineers. It's even worse now that it's led by one.

File: 1442967126425.jpg (409KB, 1157x772px) Image search: [Google]

409KB, 1157x772px

>>59296856
>AMD releases second impact of Ryzen server chips
>Has enough TDP to spare and can go simultaneously: Moar coars, higher turbo, more base clocks.
>The devestation is immense, Intel stock drops overnight and not a rebound for 8 months.
>Gen 1 chips rain down at affordable prices like when facebook dropped their v1 xeons.

>>59298320
on the phis, no, they are effectively different markets but similar.

a phi is a fuck ton of x86 cpus that hit a parity point where a bunch of small retards are able to outperform faster but less bug cores, suited to tasks that gpus just are not fit to accelerate.

and lets just be fair to intel, in the consumer market, they still own the desktop till more applications adequately handle multithreading, mobile though... intel has severe throttling issues in macbooks, and if amd is lower power...

Intel has its ass on the fire and they have literally made nothing worth a damn sense the p6, meanwhile even though amd stumbled with bulldozer, bulldozer still never fully conceded itel to clear they win at everything like the pentium 4 did to athlons. even at amds worst, they were better then intel at their worst.

>>59295500
To be fair, there's no way whatever this is costs as much as a Sun/Fujitsu unit.

>59298320

Don't forget this is all accomplished by AMD with ~1/10th of Intel's R&D budget. Just imagine AMD with a healthy reserve of cash.

>>59298451
get ready nvidia, she comin for you
https://www.google.com/search?q=giant+strapon&biw=1035&bih=602&tbm=shop&source=lnms&sa=X&ved=0ahUKEwiHlcXFocbSAhUrj1QKHfRoBfEQ_AUIBygC&dpr=1.75#tbm=shop&q=The+Annihilator+Xxxl&*

>>59298438
>they still own the desktop till more applications adequately handle multithreading
They don't, they do own the enthusiast and gaymur market, but for everything else Raven Ridge will be running circles around i5's and i3's
They're more power efficient, almost as fast ST-wise, faster at multi threading and absolutely rape Intel GPU-wise, while being cheaper
>>59298451
Intel's R&D on x86 isn't all that bigger than AMD, Intel invests in a lot of random shit
Like Optane and their plethora of FPGA acquisitions, and their fabs specially take a shitton of resources

>>59298475

>>59298511
power efficient on a desktop doesn't matter, multithread... look at photoshop, that is the most common thing a normal person will do on their computer that isn't browse the web, and that shit barely threads.

as for apu, yea, amd is going to kick intel till they piss blood, but normal mc bumbfuck knows intel is faster for gaming, so never even considered amd an option, then the drm, gives people 4k netflix.

the only way amd is going to win desktops is if oems abandon intel, or more things use multithreading. server wise, high end pc, workstation, especially if you are on a budget, amd kicks intel's ass, but you see it in the benches, programs still are barely threaded so intel's better single thread makes them look that much better.

>>59297964
But it is. They can't support enough memory for that task.

WE INFINITY FABRIC™ NOW BOIS

>>59298575
Assuming they were using a dual E5-2699 v4 system, that's up to 3.08 TB of memory, so I really doubt that. Even if it is the case, it's still bullshit because big problems like that are generally solved on clusters anyway.

>>59298741
>needs moar hardware
>additional costs
>still not an issue
Bitch pls.

>>59298758
The fuck are you talking about? It looks pretty damn obvious they're comparing two multi-socket systems. Get the cock out of your mouth,

>>59298564
>power efficient on a desktop doesn't matter
Of course it fucking matters dumb nigger, why do you think AIO's based on laptop parts are so popular? It;s small as fuck and the normies love when their shit is as small as possible
OEM's care a lot about power consumption
>multithread... look at photoshop
Ryzen is fast enough at single threading to work well enough on anything that's single threaded, your average faggot isn't runing a delidded 7700K with a custom water cooler loop
>as for apu, yea, amd is going to kick intel till they piss blood, but normal mc bumbfuck knows intel is faster for gaming, so never even considered amd an option, then the drm, gives people 4k netflix.
Any APU wrecks anything Intel at Overwatch and even Minecraft, and no one wants 4K Netflix, specially since it requires quite more bandwidth than 1080p, is more expensive and eats burger datacaps faster
>the only way amd is going to win desktops is if oems abandon intel, or more things use multithreading. server wise, high end pc, workstation, especially if you are on a budget, amd kicks intel's ass, but you see it in the benches, programs still are barely threaded so intel's better single thread makes them look that much better.
OEM's were abandoning Intel during the P4 days, pulling another anti trust would destroy them with Trump in power

>>59299191
Please note, normal mc bumblefuck, see mr. nmb doesn't look at benchmarks, what he looks at is does it have intel inside and knows amd is not the way its meant to be played.

anyone who needs a computer for work/hoby will look at benchmarks and oh will you look at that, that is either in their price range or the best for the job is not multithreaded properly, so the 7700k is better.

as for aio, had to look up what that was, it's those computers with a monitor attached, hate to break this to you but I have never seen one of those that were out in the wild that wasn't a mac.

as for non de lidded 7700, doesn't matter, the benchmarks clearly show the 7700k is at the top, ms nmb is to fucking stupid to know that her version will never preform the same.

For laptops, amd may be king because if they offer a cheap, possibly cheaper solution than intel, they can put the cpu in and have it perform overall better than intel due to no throttling or far less, but there also aren't many people who buy laptops based on specs, or the ultrabook never would have taken off.

"no one wants 4K Netflix"
Hate to break it to you but mr nmb jr sure as fuck does,
"is more expensive"
ms nmb jr doesn't care, she just wants it to work and its what, an extra 3 dollars in total, where as nmb's parents are use to paying close to 300$ for their cable packages, its a drop in the bucket that most nmb's don't even think about.

and with trump... really? you think someone using a shitty business tactic would be checked by trump? the person who appointed some fuck who wants to not regulate the internet again? Yea, i'll believe that when I see it.

File: 1399601060282.jpg (64KB, 420x288px) Image search: [Google]

64KB, 420x288px

>>59296644

>>59295488

per processor

>>59299408
>Please note, normal mc bumblefuck, see mr. nmb doesn't look at benchmarks, what he looks at is does it have intel inside and knows amd is not the way its meant to be played.
OEM's do, I see you have no idea about what you're talking about since you didn't even know what an AIO was
>muh marketing
Yeah, you sure aren't dumb redditlord
>anyone who needs a computer for work/hoby will look at benchmarks and oh will you look at that, that is either in their price range or the best for the job is not multithreaded properly, so the 7700k is better.
Yeah, I already told you the enthusiast/gaymur market will stick with Intel, guess what, they're a minority at best
>as for aio, had to look up what that was, hate to break this to you but I have never seen one of those that were out in the wild that wasn't a mac.
Sorry to break it to you but this are the most sold desktops out there, maybe you should stop staring at Apple shops
>For laptops, amd may be king because if they offer a cheap
APU's will offer better battery life, which is everything the normies care about
>Hate to break it to you but mr nmb jr sure as fuck does,
Of course, you have zero source on this, it this was true Win10 and Kaby Lake would be selling out
They might do for their meme TV's, but guess what faggot, they buy Roku's for that
>ms nmb jr doesn't care, she just wants it to work and its what, an extra 3 dollars in total, where as nmb's parents are use to paying close to 300$ for their cable packages, its a drop in the bucket that most nmb's don't even think about.
Price is fucking everything massive retard, the few people who won't go with the best value will stick to MBP's
>and with trump... really?
Do you seriously believe he wouldn't pick on a company who is going full anti competitive to other major American companies? Yeah sure nigga, GM got scared for far less

>>59297964
>384 GB of memory
>not enough

I understand they need fuckloads of it but how did intel fail to see this coming

Is 512GiB per processor the max on Naples? I know that they have double the channels/bandwidth of the very top offering from Intel, but the top end Intel chip can address 1.5TiB per processor afaik. I understand if that's not the price segment that they're going for, but still I'm curious.

>>59291281
Pricing?

>>59295222
>30 minutes POST time.

>>59295222
Don't forget about AMD’s Firepro S7100X either. Hardware Virtualised, that shit can handle up to 16 VMs on the hardware.

>>59292681
they have 90% in thin clients though, not that it matters at all

>>59293796
they sold all of them here in rus lands, all is left overpriced resellers
people buy them even without boards

>>59298236
>These coming years
q4 financial report going to be hilarious either way
two variants
amd fails: we will laugh at that per usual
second variant amd succeeds: intel loses several billions, nvidia loses half it's computing market and 10 more percent in gaming share

>>59300517
Don't forget, Intel is leasing iGPU related IP from AMD, $300M per annum/5 years was the deal Intel held with nVidia for it, so let's assume the same deal applies to AMD. That $300M changing hands will be big.

>>59291566
Dude. I have a 2.3GHz Xeon with 12 cores and it's a Haswell. It does 72W at full load. There is no fucking way a 14nm Broadwell at 2.4GHz would 145W with 8 cores. That'd mean they lost massive amounts of energy efficiency by moving to 14nm with basically the same microarchitecture. How the fuck would that be possible, huh?

>>59300517
I'm getting ready to short Intel stock on Q4

time to buy up on amd stock

intel is finished

>>59300555
they dropped 2 per share, but mkt cap is 1B higher

what happened there?

>>59298407
2.4 ghz on 22 cores, not 32.

Intel is really getting wrecked here, just admit it. Zeppelin has 1.5x perf per watt under 3ghz.

None of you guys see the big picture.

The problem for AMD is that things are not hardware agnostic in the datacenter. AMD might dominate the datacenter in 5 years, but not until they can prove proper support.

>>59300662
isn't it not how it works for x86?

GPU you have a point, but 90% of x86 is standard?

no idea, just asking

>>59298429
It addresses 2TB max.
I don't know where anon got those numbers. Each one can support 2TB.
2TB is the max it supports, and it has 16 memory channels. 16x128=2TB.

http://www.techspot.com/news/68431-amd-squeezes-32-zen-cores-naples-server-cpus.html

"With 16 GB DIMMs, it will be possible to kit out a Naples server with 512 GB of DDR4 memory, although AMD states that each Naples CPU supports up to 2 TB of memory (4 TB in a dual-socket system)."

https://www.extremetech.com/computing/245462-amds-new-32-core-naples-server-cpu-will-attack-intels-datacenter-monopoly

"AMD has glued together multiple four-core CCXs into at least two multi-chip-modules (MCMs). Each MCM has four memory controllers and there are two groups of four in a 32-core chip, which is how AMD hits its 8-channel memory target. Total memory bandwidth per CPU is 170 GB/s."

>>59300769

So why is a "Naples server" limited to 512GB, whilst each actual Naples CPU can handle 2TB? Would that need a retardo mainboard or PCIe memory banks or some shit?

>>59300769
>>59300788

I'm a retardo. Could just use 128GB DIMMs for 4096GB.

>>59299784
2TB each is max.
So 4TB on a Naples system.

>>59300688
I'm too tired to respond in full, but no, not really.

Lots of Intel proprietary stuff is very useful and supported. Think about hypervisors like Xen and how well they support anything besides Intel.

If you make money using computers in a datacenter, you cant afford to risk it. It's cheaper to buy a less performent Intel processor to not have to deal with the problems.

TPM, VT-d, VT-x, GVT, none of this is the same between Intel and AMD. EPT is not the same as RVI.

Intel wouldn't have that much issues if they had something in the next 2 years, they don't, that's the issue, while Zen's upcoming iterations have plenty of room for growth-

PER SOCKET AMD has more cores, more I/O, more bandwidth, more memory amount..

How can Intel even fucking compete? It would be okay if the core was significantly slower like the Bulldozer one, but it's practically neck-to-neck with Broadwell AND can clock higher at a lower TDP.

This is looking pretty grim.

>>59299784
It's a 2 socket system, Intel supports less DIMMs per socket and in turn would need exotically priced 64/128GB DIMMs to compete, which AMD can use as well so in the end AMD has more memory by default.

>mfw zen+
>256bit fpu pipes
>improved memory controller
>3 channels per die
>10 cores per die
>more lanes
>higher clocks at the same power


Intel is literally fucked.

https://translate.google.com/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http://www.pcgameshardware.de/AMD-Zen-Codename-261795/News/Zen-2-Pinnacle-Ridge-Release-Termin-1222222/&edit-text=&act=url

Pinnacle ridge early 2018

>>59301403
You forgot that Zen is cheaper to produce core per core, and much cheaper in high core count configurations

this thread makes me feel warm inside

>>59301569
>Naples is 32c/64t for ONE CHIP
>You can build a server with >100 CPU threads when Naples comes out
what a time to be alive

>>59301647
>zen+ coming out in under a year


Just in time for most software, firmware BIOS issues to be fixed, and maybe some more quality motherboards to come out!

My 3570k can wait another year.

>>59302000
Meh.
I'm just going to upgrade to a 1600X once it and some decent affordable motherboards come out. Then I'll do the Zen 7nm upgrade in 2019 as it looks like that'll be the real hot shit.

>>59297965
>>59298078
>Intel
>Got blueprints from IBM
>didn't want to deviate at all from them
>AMD found more optimal way of doing things right away
>Cyrix backwards engineered a better x86 CPU than what Intel ever could
>Only way Intel could survive was shady business tactics

Intel was always 10 years behind competition.

This is a large reason people are excited for Amd's new CPU even if they aren't buying one, it's a new architecture, something Intel lately hasn't done for... since the Pentium 3... so Intel has been floating along on the back of Pentium 3 nearly 20 years, and will be over 20 by the time they make something new, while AMD has had 3 different cores in that time frame with 2 of them being good, one of them being good for specific applications for a short time, would likely still be good enough today if they released full desktop variants but fuck it, when gamers are concerned its hard to argue in favor of construction and cat cores.

Throughout the entirety of the bulldozer base, look at what AMD did gen over gen... and now look at Ryzen, and imagine it to be Bulldozer levels of optimization left to happen. now realize Ryzen currently on engineering samples that throw errors like a mother fucker, is within/over 5% IPC of Intel's current CPU.

1/2

>>59302190
I cant wait to see if either there is a performance stand still, you can't get much faster than Intel is, or if we have been rammed up the ass by Intel for so long we forgot they were even there and its only now that we finally see real vagina we remember a cock is still lodged up our asses.

Intel has way more fuckups than Bulldozer, Bulldozer is AMD's only spectacular fuckup in the CPU market in 40 years, the only other minor fuckup was Phenom which was not even a architecture fuckup since it was fixed in B3 stepping, the other fuckup was K5.
But these are fucking nothing compared to Presshot, Itanium, Larabee, iAPX 432, and their own fuckup with P5 FDIV that wasn't nicely fixed with a stepping, but a fucking recall.
What about their complete failure with the Atom? What an abortion, mobile market? 14nm and lower lithography woes where they allowed their competitors that were lagging 4 years to close the gap to less than a year?

These magnificent failures would destroy 5 companies over, it's a good thing Intel has more money than sense to live through it all.

2/2

>>59300538
Intel still hasn't paid AMD that over 1 billion fine yet either.

>>59302193
Oh hey my pasta is being used around.
I feel so special.

>>59302193
>>59302190
You should add the Real3D acquisition and the massive disappointment that the i740 was, followed by a whole decade of Extreme Graphics and GMA being a decade behind the competition

>>59291332
Who cares, the only relevant way to measure CPU performance is if it can get 500 FPS in CS 1.6 at 640x480

>>59302224
It tells you everything when you find out Intel only started writing 3D drivers with Sandy Bridge.