Why do AMD products always use so much more power than their

>>59182348
>double the cores of a 7700K
>draws less power

INTEL HOUSEFIRES BTFO

Imagine how little juice the Ryzen quad cores will need...

>>59182348
>95w CPU vs 91w CPU
>95w chip with 8c/16t uses less power under load

lol

>>59182348
son, are you, by any chance, retarded?

File: 1468311873778.png (66KB, 650x387px) Image search: [Google]

66KB, 650x387px

how did they manage to run i7s that hot at idle?

File: 1465220584565.png (66KB, 650x379px) Image search: [Google]

66KB, 650x379px

load is lower than 4c.

>>59182348
Is reading graphcs not part of Indian education?

>>59182535
Probably because they live in some sandy Middle Eastern shithole where it's boiling hot all the time.

>>59182560
This review is from Iran.
wew

>>59182552
>95°C
Jesus Christ Intel

>>59182348
anon, tell me
422 and 437, which number is bigger?

>>59182583
That Intel-brand TIM semen at work.

>>59182348
The bars look like penises

>>59182535
AYYMD HOUSEFIRES

>>59182600
He's making joax because apparently AMD has the most power-efficient X86 design on the planet now.

Intel is straight fucked in the server market. AMD can provide 99% of the performance at a lower power draw and resultant thermal output. Servers are going to devour these chips.

>>59182558
Is recognizing sarcasm not part of common sense?

Maybe we should s_(.*)_\1/s_ to make you newfags feel at home...

I'm not a gaymer, I'm straight.

>>59182710
But 49 is more than 46!

>>59182348
ANOTHER SHOAH

>>59182763
3w at idle means fuckall to a server farm.

>>59182937
> to a server farm.
It's a home CPU.

File: 1476240000930.jpg (10KB, 225x225px) Image search: [Google]

10KB, 225x225px

>>59182552
>iranian heatsink installers

>>59182937
You sure about that?

When you have thousands of cpus in a server farm, any efficiency is identified and implemented if possible.

3w @ idle sounds like fuck all but it's extra power, extra heat and extra cooling needed.

>>59182974
But a server farm isn't going to spend much time at idle, retard. It'll spend the majority of it's time at load, where Ryzen BTFO's Intel in terms of power/performance.

File: tmp_16287-ss+(2017-03-01+at+07.41.00)-22106064.png (23KB, 694x261px) Image search: [Google]

23KB, 694x261px

And apparently, if this is accurate, they were using a crippled mobo that could only clock the memory at 2133mhz

Not bad

>>59182947
it scales to server clusters

>>59182974
>>59182997
It's also twice the core and thread count for that 3 watts.
Intel is literally fucking dead in the server sector.

>>59182552
>>59182535
>>59182348

INTEL BTFO
T
F
O

>>59182937
But it means a whole lot for laptops, embeded, and most mobile shit.
Not all servers run 100k cores at all times, some will have to wait for serial crap to finish, and during this they'll be idle, lower wattage here is important.

>>59183011
at half the price

>>59182348
SHUT IT DOWN
H
U
T

I
T

D
O
W
N

>>59183011
The top Naples SKU apparently is only 180w.
32c/64t with clocks ahead of Broadwell Xeons.
The Zen core really shines at lower power.

>>59183039
laptops will use APUs not 8core16thread chips.

>>59183054
Don't state the obvious, the cores is still the same, lowering the count just makes it use less power.
Matter of fact is AMD has a huge perf/watt advantage now, ever since 2005.

>>59183011

Not to mention it's throughput in multicore. It's not even funny much better Zen is when pushed to it's limit.

AMD will hopefully crack open the server market and year over year improve from that.

"Lools like meats back on the menu boys!"

File: Worried_Gengar.png (24KB, 200x200px) Image search: [Google]

24KB, 200x200px

>>59183011
>>59183037
>>59183040
>>59182997
>>59182937
>>59182710
>>59182600
>>59182583
>>59182558
>>59182461
>>59182423
>>59182405


Fuck off AMD pajeet shills, Ryzen is going to flop and burn just like ever fucking AMD product ever made. Intel does. And always will make the best Processor regardless of what your shitty little pajeet company says is """"""""""better""""""""""

Just because it's better on paper doesn't mean real life performance is better you fucking morons.

>>59182348
JESUS FUCK BABY AMD IS BACK

AMD
IS
BACK

This can't be fucking real, look at that perf/watt, that's fucking insane, Intel will need two full node shrinks to get to that and only if AMD doesn't upgrade its core next year.

>>59183099
topkek

>>59182997
You'd be surprised at how often servers are actually at idle or low usage. They're not getting gunned at 90+% all day every day.

Indeed back when I was working in them, systems in the operations centre were designed to warn if systems were using excessive resources over extended periods.

>>59182348
Hey guys remember how shit GloFo is?

>>59182937
>3w at idle means fuckall to a server farm
3w x 50000=you save money

>>59183099

>That img

It really put a geniune smile on my face. Thank you for your hard work as a dedicated poster on this Kazakh throat singing board. Take care of yourself and be happy.

>>59183137
server farms don't idle.

>>59183170
You're pretty clueless at this.

HPC farms might not idle, much, they still do since everything has serial loads, even HPC shit, but server ones sure as fuck do idle a lot.

>>59183181
>>59183137

Anon you are grasping at straws now.

The img you are looking at is the idle of 1700x and no server farm is going to use that cpu.

They are going to use the 65W 1700s, which probably idles even lower than most KabyLake i5s.

>>59182348
you do know this is the ES sample they had eh iran cant get retail stuff

File: IMG_04700.jpg (186KB, 600x598px) Image search: [Google]

186KB, 600x598px

>>59183249
samefag here

the picture they provided along with the bench

it had a nice run being a shilled thread

>>59183265
So that means this isn't real performance and performance will be even lower?

AMD BTFO

>>59183282
yeah an ES being on 3.15ghz is going to have more perf from one being on 3.2ghz

#logic

>>59183265
oh, that's where all the samples go when product gets released

>>59183305
AMD is hiding the truth again, putting a cherrypicked ES instead of retail units is misleading performance to make them look better than they are

>>59182348
What are you saying, Geoffrey? The AMD processor has the lowest power use values on that chart.

>>59183249

>Iran can't get retail stuff?

What really? Wtf do they do, go and buy their PC related stuff from Turkey/Azerbaijan/Armenia or something?

And how the fuck do they have servers or hell, gaymin?

>>59183364
they were buying ps3s and ps4s untill someone tipped them off that they were building an HPC with them (iraq bought them and then sold them to iran for oil)

File: certified shit wrecker.png (400KB, 816x1020px) Image search: [Google]

400KB, 816x1020px

How does he do it?

File: Screenshot from 01-03-2017 17:47:43.png (1MB, 1920x957px) Image search: [Google]

1MB, 1920x957px

Looks real fake, I don't trust turks with anything.
Here's a real benchmark from America and more in line with standard testing and results.

>>59183419
>7700k at 5Ghz has lower average than 7700k stock
>real benchmark
Fuck off, guy.

>>59183419
bashing a fake review with an even more fake review

/g/ in a nutshell

>>59183265
>Made in China
WTF, You fags said their fabs were in NY
Fuck you guys I'm cancelling my pre-order

>>59183476
>WTF, You fags said their fabs were in NY

Fab 8, located in Luther Forest Technology Campus, Saratoga County, New York, United States is a new 300 mm Fab. This fabrication plant was constructed by GF as a green field fab for advanced technologies. It is capable of manufacturing 14 nm node technology. The plant's construction began in July 2009 and the company started mass production in 2012.[5][10] It has a maximum manufacturing capacity of 60,000 of 300 mm wafers/month, or the equivalent of over 135,000 of 200 mm wafers/month. in September 2016, GlobalFoundries announced it would make a multibillion-dollar investment to refit Fab 8 to produce 7 nm FinFET parts starting in the second half of 2018.[11] The process is planned to initially use Deep ultraviolet lithography, and eventually transition to Extreme ultraviolet lithography.[12]


>diffused in USA
Meaning cut from the wafer in USA
>made in China
Assembled in China

>>59183499
>Assembled in China
I don't like this.
Fuck this shit

>>59183511
All they do is set the chip on the package and put the IHS on top of it.

Literally go kill yourself.

>>59183511
Everything besides packaging is done in fatland, anon.
Hence the >diffused in USA
part

>>59183531
>All they do is set the chip on the package and put the IHS on top of it.
So why not do it in the US?

>>59183395

>they were buying ps3s and ps4s untill someone tipped them off that they were building an HPC with them (iraq bought them and then sold them to iran for oil)

LOL! Those brown fags were right though, PS3 CPU was really good for the price for multi-threaded workloads.

>>59183511
>Like every gpu vendor too
Yeah , go fuck yourself

>>59183545
Because assembly is cheaper in China

>>59183419
It hilariously how bad this fake review is fooling everybody

>>59183545

shipping the high tech made products to China, having the chinkbots assemble it there and ship it back to US/EU or whatever is cheaper than assembling it by actual humans and paying them the lowest possible wage + benefits anon.

Welcome to the new world.

File: amd gods zen 2.jpg (90KB, 2560x1440px) Image search: [Google]

90KB, 2560x1440px

1700 4.0ghz + 1080

http://tieba.baidu.com/p/5003192198?see_lz=1

http://tieba.baidu.com/p/5003192198?see_lz=1&pn=2

>>59183639

>no specs
>don't know the game

>chinese benchmark

discarded

>>59183680
it's ROTR

>>59183419
>looks real fake
>post fake

cant make this shit up

>>59183639

Would help knowing what the fuck that is

>>59183680
watch link above your head cunt i won't post 3 images since i'm too lazy just as you are

>>59183695
>ROTR
at what res?

>>59183695
its from the recently leaked one? the ones that showed the min on the ryzen being -34% than the intel?

thats some nvidia dx12 magic right there

>>59183754
CHECK LINK YOU PIECE OF LAZY SHIT FFS

http://tieba.baidu.com/p/5003192198?see_lz=1

>>59183419
>looks fake guis so i'ma show the fake bench that caters to my needs ;DDDDDDddd

kys

>>59183660
>they claim to have OC'd a 1700 to 4.0 on the stock cooler
yeaaaaaah, I'm calling bullshit on that one
no way in hell you do a 33% OC with the shitty Wraith Spire

Reminder that this is a fucking ES running stock clocks at 2133MHz memory

see
>>59183265


AHAHAHAHAHA

>>59183563
>>59183626
I've seen Ryzens that were assembled in Malaysia as well

>>59182348
>Twice as many cores
>Lower power
Top kek, are Jewtel even trying?

>>59184693
Easier to just pay off OEMs.

File: 1487838437461.jpg (18KB, 267x297px) Image search: [Google]

18KB, 267x297px

>>59182348

Actually they messed up a lot of things:

>A: ES SAMPLE
>B: Broken non updated bios
>C: Ddr4 stuck @ 2133
>D: Mem latency of a 2004 pentium
>E: No boost
>F: No amp profile
>G: No xfr

INTEL BTFO!

>>59183099
>3 (you)s
thanks :^)

>>59183137
>server farm with 4 core CPUs
lol

>>59184835
>>G: No xfr
he oced it to 4.0ghz you moron in bioshock test

>>59182348
This is being ironic r-right?

It clearly is more power effective than intel

Keller is a master of low power design

>>59185038
>found persian scum

Why does the 7700K has a lower power consumption than the 6700K? That doesn't make sense.

>>59185084
keller didnt design the zen ppl

he helped reduced the tpd on the mobile zen platform and he finished up the k12

zen came purely from amd

(he probably helped on desktop zen a bit but it wasnt his design the ldo for example it was his idea basicly the voltage going directly into the cores is 0.2 0.3 lower than its actually showing they will update later the cpuz to show this )

>>59185482
has anyone really been far even as decided to use even go want to do look more like

>>59183396
He cums on ur mum m8

>>59183545
1. Cheap labor
2. They can avoid heavy import taxes when selling in Chinese markets since it was technically produced there.

>>59183099
Sorry you can't read you lazy brainlet