[Boards: 3 / a / aco / adv / an / asp / b / bant / biz / c / can / cgl / ck / cm / co / cock / d / diy / e / fa / fap / fit / fitlit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mlpol / mo / mtv / mu / n / news / o / out / outsoc / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / spa / t / tg / toy / trash / trv / tv / u / v / vg / vint / vip / vp / vr / w / wg / wsg / wsr / x / y ] [Search | Free Show | Home]

>z8300 consumes 2w of power Why doesn't Intel make something

This is a blue board which means that it's for everybody (Safe For Work content only). If you see any adult content, please report it.

Thread replies: 65
Thread images: 11

File: 1444292921241.jpg (3MB, 5250x7230px) Image search: [Google]
1444292921241.jpg
3MB, 5250x7230px
>z8300 consumes 2w of power
Why doesn't Intel make something like a 10w 20 core atom board for servers? Why don't they make atx boards for the newer atoms? Would love to upgrade my d2550.
>>
>>56347696
>teletubbies
was britney pandering to lolicons?
>>
>>56347696

>That pic

Man talk about a nostalgia trip. One my first crushes and faps.
>>
File: 1450385857166.jpg (20KB, 479x398px) Image search: [Google]
1450385857166.jpg
20KB, 479x398px
>>56347696
LEAVE BRITNEY ALONE
>>
>>56347696
because unlike fucking children like you, people who buy servers dont want a 10w piece of shit
>>
>>56347894
Yes they do. Modern atoms perform pretty well, 5 of them together would perform great for a small server. Not everyone needs a xeon.

Plus power consumption is important in servers because they're pretty much on 24/7 and the heat they put out needs to be cooled.
>>
>>56347968
you arent looking smart. if a fucking atom can replace your 'server' you are just trying to show off about your epic seedbox or usb 2.0 laptop hard drive.

NO ONE is going to think about replacing a real machine with a piece of shit like that, period. this isnt your parents house where they will notice the extra 300w pull on their power bill
>>
>>56347696
Xeon-D

Also she has a hairy stomach.
>>
>>56348008
>Xeon-D
This. They're way more power efficient than cheap airmont atom cores (really just tiny silvermont cores pretending to be important).
>>
>>56347696
What kind of retarded bullshit are you doing that can multithread over 20 cores but is still super sensitive about power?
>>
>>56348045
Netflix and chill
>>
>>56348045
Video encoding but atom cores suck balls for cpu intensive applications. 2 atom cores = 1 haswell core for example.
>>
>>56348077
>video encoding
if you're OP then yeah, that's some retarded bullshit. It's not even a "server" retard. And no, power is not a concern for video encoding
>>
File: Clipboard01.png (27KB, 673x294px) Image search: [Google]
Clipboard01.png
27KB, 673x294px
>>56347696
They make the Xeon-D line. Those actually have appreciable performance unlike the throttling low performance Atoms with their "SDP" rating.

Though if you wanted a ton of Atom cores you could buy a Xeon Phi, goyim.
>>
>>56348136
>And no, power is not a concern for video encoding
Who the fuck are you tell him to not be concerned about power consumption?
>>
>>56348136
Power consumption is a real issue for servers.

>>56348138
Isn't that a little higher than the 2w of an atom core?
>>
>>56348171
Duh, it's not a shitty Atom core and there's 8 of them on the die.
>>
>>56348136
>if you're OP then yeah, that's some retarded bullshit. It's not even a "server" retard. And no, power is not a concern for video encoding
10-bit HEVC encoding is very time consuming process across all processors especially above 480p video resolutions.
>>
>>56348171
The Xeon D 1557 is 12c
The Xeon D 1581 is 16c

They're pulling 3-4w per core, not exceeding TDP then throttling like crazy which is what Atoms do.

>>56348184
8 core Xeon D is the smaller part. Its not tested here.
>>
>>56348184
Atoms have 4 cores per die too, don't they?

At 120w, you could have 240 z8300 cores and I'm pretty sure that just might perform better...
>>
>>56347696
>let's build a server that requires 60W for RAM, chipset, and LAN, and populate it with a 2W processor.
>>
Not op, but I want to get into video recording and editing.

How slow will 5 orange pi PCs be at this? In a cluster.
>>
>>56348214
>8 core Xeon D is the smaller part.
Oh, I thought the 1557 was the 8 core.
>>56348228
Go buy 240 fucking Atoms, then.
>>
>>56348153
If atoms did the same work for less power it would be all Intel sold retard. If you're doing compute intense work you buy a real CPU or wait 10 times longer for your stuff to finish

>>56348171
>Power consumption is a real issue for servers.
But you don't have a server. You've got another computer in your bedroom that you do video encoding on.

>>56348206
how does that change anything I said?
>>
>>56348228
>you could have 240 z8300 cores
no, it doesn't work like that retard. If it did AMD could just glue two processors together and actually compete with intel
>>
>>56348265
Jesus Christ you're a fucking retard. Power consumption is a concern everywhere. It's a bigger concern in a datacenter, but it's still a big concern at home for the same reasons.
>>
>>56348265
>how does that change anything I said?
Servers do video encoding you fuckstick. The more power efficient the processor the less money processors waste in the long and short term which is what businesses care about.
>>
File: VIA-EPIA-M900.jpg (81KB, 427x449px)
VIA-EPIA-M900.jpg
81KB, 427x449px
>>56348280
Isn't that what VIA was doing until recently?
>>
>>56348228
Atom chips do not operate at a fixed clock speed. They have strict power and thermal target that they're not allowed to exceed, and they will hard throttle their clocks when they exceed that limit for more than a couple seconds at a time.
Atom core arch(latest family is Goldmont) do not have the same performance per clock as their larger Core i arch family.
A 1ghz Goldmont core only offers a fraction of the performance of a 1ghz Broadwell core.

This is why the Xeon-D line exists and why it doesn't use Atom core arch. Their performance per watt is too low to be a compelling enterprise offering.
>>
>>56348303
I already answered your post. Repeating it doesn't change that "If atoms did the same work for less power it would be all Intel sold retard."

>>56348310
1. OP has a computer for video encoding in his bedroom. I don't know what all this "server" and "business" bullshit is
2. See above. There's a reason actual servers aren't filled with 6000 atoms. They're not effective for that kind of work.
>>
>>56347696
It's called xeon phi
>>
>>56348303
People are not dropping their blades for a box full of atoms because of power consumption. A modern xeon is plenty efficient. If you even consider replacing a real box with a fucking '20 core atom' you don't have a server, you have a repurposed piece of shit family computer.

shit my home lab pulls about 275-300w constant and that isnt even much.

people are making fun of you because anyone who thinks a fucking atom is a server replacement is a fucking moron, full fucking stop.
>>
>>56348316
no
>>
Atoms are server processors too, next gen atoms will be initially launched as server boards, then other possibilities will be considered like laptops.
>>
>>56348363
Storage servers, the super low intensity work that ARM cores handle in a NAS.
>>
>>56348363
Not for rendering farms and similar dummy
>>
>>56348348
>I already answered your post
No, you didn't. There is no answer to my post. Worrying about power consumption is completely valid. You can't contest this unless mommy pays the bills and you do literally nothing with your equipment.
>>56348351
>People are not dropping their blades for a box full of atoms because of power consumption.
Why the fuck would you even say some stupid bullshit like this? No, no one is going to replace their C7000 with a single Xeon-D board, you fucking retard.
>shit my home lab pulls about 275-300w constant and that isnt even much.
OK? Good for you, nigga.
>people are making fun of you because anyone who thinks a fucking atom is a server replacement is a fucking moron, full fucking stop.
Xeon-D != Atom. Full fucking stop. Don't reply to me with this dumb shit ever again.
>>
>>56348363
Nah, atoms are pretty much DOA and intel has mostly given up on them. We Xeon-d and carrizo/bristol ridge APUs now.
>>
File: Screenshot_2016-08-31-12-13-22.png (273KB, 1080x1920px) Image search: [Google]
Screenshot_2016-08-31-12-13-22.png
273KB, 1080x1920px
>I3 6100 is 170% faster than a z8700
Does that mean 5 of those atoms would actually outperform the i3?
Not bad when you consider it's 10w vs 54w.
>>
>>56348415
>Worrying about power consumption is completely valid.
Will this be the third time you fail to understand that having a shitload of atoms encoding video will not magically do the same work for less power?
If you're worried about power consumption you don't buy a shitload of atoms for video encoding. Do I need to clarify it a 4th time?
>>
>>56348415
sorry kiddo, but if your parents ask what that noise is and why their power bill is $20 higher you need to find an alternative hobby. power consumption is only an issue if you are talking like 300w+ constant, and even that is conservative

get a job
>>
File: Screenshot_2016-08-31-12-15-58.png (339KB, 1080x1920px) Image search: [Google]
Screenshot_2016-08-31-12-15-58.png
339KB, 1080x1920px
>>56348439
I was expecting a quad core skylake to be more than 2.5x faster than a quad core Intel atom...
>>
>>56348439
It's not 10w retard. You need additional hardware. see >>56348280
>>
>>56348449
Once again. Xeon Ds are not Atoms.
>>56348456
Incoherent bullshit. Go back to pretending to be a sysadmin in your basement, now.
>>
>>56348473
I've never mentioned Xenons ITT dummy
>>
>>56348489
Then fuck off.
>>
>>56348473
using a xeon-d for 'encoding video' makes no sense. and since time is money you would better off buying something like a 2 cpu 12 core box and spinning it up to rock out a bunch of encodes quickly. but you live with your parents and are concerned about power draw
>>
>>56348439
If you're going to use a shitty synthetic benchmark then at least use passmark you cuntsickle

https://www.cpubenchmark.net/compare.php?cmp%5B%5D=2380&cmp%5B%5D=2585

But yeah atom performance is fucking shit no matter what. intel basically gave up on them due to how embarrassingly shit they performed irl.
>>
>>56348502
>be wrong
>someone points out that you're wrong
>yeah but [irrelevant thing]
>that's irrelevant
>fuck off
lol
>>
>>56348504
>live with your parents
What difference does that make? Literally who gives a shit? Power consumption is an even bigger issue for servers than desktops and you're going on about who I live with?
Nice misdirection you degenerate fuck
>>
>>56348518
Dam this makes me miss middle school a little. I mean the whole experience was shit but the way middle school kids did damage control make me chuckle and not want to kill myself at least for that day.
>>
>>56347696
Why don't you compress that jpg?
>>
>>56348519
>what difference does it make

fucking confirmed. power consumption is an issue for servers when you have a room full of them. ONE machine that barely cracks a couple hundred watts under load is a non issue.
>>
>>56348519
Not him, but calling your shitty toy computer grown up words like "server" doesn't change anything. You plug it into the wall in your room your only "concern" is if mommy thinks you leech to much
>>
File: 1472606349831.webm (206KB, 1487x2048px) Image search: [Google]
1472606349831.webm
206KB, 1487x2048px
>>56348545
>not using a static webm like a normal human being
>>
File: 66520pwloadtotal.png (29KB, 600x550px) Image search: [Google]
66520pwloadtotal.png
29KB, 600x550px
>>
File: Capture.png (118KB, 689x1100px) Image search: [Google]
Capture.png
118KB, 689x1100px
>>56348818
>>
>>56348834
jesus...
>>
>>56348865
Now imagine the 12c/24t part with the same 45w TDP.
The 16c/32t parts are only 65w.

Atoms cannot compete in perf/clock with the Core i line. Xeon-Ds have throughput and insane perf/watt. That Atom X2750 is made for a glorified NAS, not real performance.
>>
>>56348834
Damn, that Xeon would be great on a Workstation laptop. Same TDP with double the cores, no useless IGP.
>>
>>56347696
>4 cores, no ht
>1.8ghz
>twelve 2015 intel gpu cores
>2w "scenario design power" I assume means 5w heavy load
>$21 unit price

Ok, what's the catch?
>>
>>56349288
It reaches a maximum of 2.5W for CPU only loads on all 4 cores, full turbo-boost. Once the GPU comes into question like playing games, it hits 10W.

Tested on my Surface 3 with readings from HWMonitor.

Really fucking slow for Single threaded tasks, the IPC is similar to Pentium 4s.
>>
>>56349288
1 PCIe 2.0 lane
2GB RAM. One channel.

http://www.cpubenchmark.net/compare.php?cmp%5B%5D=2585&cmp%5B%5D=2630
>>
>>56348834
>c2750
What year is this
>>
>>56349441
They work with 4GB too though.
Thread posts: 65
Thread images: 11


[Boards: 3 / a / aco / adv / an / asp / b / bant / biz / c / can / cgl / ck / cm / co / cock / d / diy / e / fa / fap / fit / fitlit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mlpol / mo / mtv / mu / n / news / o / out / outsoc / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / spa / t / tg / toy / trash / trv / tv / u / v / vg / vint / vip / vp / vr / w / wg / wsg / wsr / x / y] [Search | Top | Home]

I'm aware that Imgur.com will stop allowing adult images since 15th of May. I'm taking actions to backup as much data as possible.
Read more on this topic here - https://archived.moe/talk/thread/1694/


If you need a post removed click on it's [Report] button and follow the instruction.
DMCA Content Takedown via dmca.com
All images are hosted on imgur.com.
If you like this website please support us by donating with Bitcoins at 16mKtbZiwW52BLkibtCr8jUg2KVUMTxVQ5
All trademarks and copyrights on this page are owned by their respective parties.
Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
This is a 4chan archive - all of the content originated from that site.
This means that RandomArchive shows their content, archived.
If you need information for a Poster - contact them.