x299 confirmed housefires

>>61128939
Well, when you put hundreds of watts through a socket, that tends to happen.

>>61129063
this.
intel needed the CPU to work for show and under normal conditions, they didn't care for the OC people enough to raise the limit.

>>61129115
It still hits 100c on synthetic load (Prime95) even with an AIO, though. This platform is really just such a mess.

What are they even going to do about the 12-18 core CPUs? They're gonna have anemic clockspeeds, or they'll have to put out a new motherboard to handle it. Hell, sounds like they might need to put out a new motherboard anyway.

AMD Vega overclocked + Skylake-X overclocked when?

Make some charcoal

>>61128939
Oh god it's debauer again, you know this is gonna bring in a major shitstorm

>>61129181
I knew I bought that 1500 watt PSU for a reason.

>>61129181
Make sure to get the 375 watt watercooled version. 300 watt aircooled version is for pussies.

>>61128939

>Reports
>Youtube video

Brought to you by AMD

>>61129213
nah, water one will be less silent at 100%, I want WHIRRRRING noises, combined with the 400mm double rad I need for cooling skylake it's gonna sound like an airport.

>>61129218
It's debauer, he's a renowned intel overclocker.

pls think before you sitpost

>>61129245

>x299 confirmed housefires (plural)
>Reports (plural)
>It's one guy who intentionally overclocks shit past reason for youtube views

>>61128939
120-130C VRM TEMPS ON NON AVX

WHAT A DISASTER
I don't know which one will be a biggest shistorm come tomorrow, this or Vega

>>61129181
Time to buy stock in uranium mines I guess.

>>61129258
>4.5 1,2v
>past reason

Did you even watch the video?

>cables get so hot they might catch fire and blow up your PSU
Lmao

File: 1497287465725.jpg (158KB, 920x900px) Image search: [Google]

158KB, 920x900px

>>61129218
>Brought to you by AMD
t. fedora

>frequency jumping from 4.6GHz to 1.2GHz because everything is close to blowing up

This is crazy

>>61128939
>everyone designed their VRM heatsinks to look cool rather than perform their required function
How the fuck does this even happen?

>>61129258
>4.6GHz
>1.25V
>beyond reason

>>61128939
>assume it's some random shitty kiketube channel making a fake video for clicks
>it's actually der8auer

Hahahaha, Intel on suicide watch.

>>61129424
Intel rushing shit, this will be a chomp on the ass instead of a regular bite Skylake-X reviews brought them.

>$300+ motherboards throttling, blowing up VRM and generally having disgusting build quality

WHAT IS GOING THE FUCK ON HERE

>>61129508
Gonna be a repeat of this, but this time there's no easy fix, some motherboards are just shit and can't OC. http://www.legitreviews.com/intel-x99-motherboard-goes-up-in-smoke-for-reasons-unknown_150008

Honestly Skylake-X CPUs are crazy enough even without OC. I don't think anyone should purchase these with the intent of OCing, just run at stock if you're nuts enough to buy it.

>>61128939
we're gonna need 1000 watt thermal coolers for this one.

I'll just throw this in here too, video exploring the inconsistent performance of Skylake-X due to the various architecture changes (mesh, L2/L3 cache sizes) https://www.youtube.com/watch?v=3VV_It9kdrs

File: 1497882782692.png (758KB, 739x1221px) Image search: [Google]

758KB, 739x1221px

>>61129540
Just get a chiller/compressor bro

>>61129258
How did you complete the captcha?

File: i9.jpg (434KB, 1906x1536px) Image search: [Google]

434KB, 1906x1536px

>>61129530
Blame Intel. Motherboard manufacturers are trying to cope with a CPU that draws 300W with a mild overclock.

>can't even do a stable clock over 4.5gHz

wow intel fucked up hard

>>61129647
You can, but you need a 600 dollar motherboard

>>61129647
Not Intel! Mothebroad guys..

>>61129682
You can go cheaper than that, just watercool the mosfets, didn't you watch the video?

>>61129701
Gonna be a tight fit getting the water blocks on alongside the LN2 pot on the CPU.

File: [Mori] Kiniro Mosaic Pretty Days [720p].mkv_snapshot_25.43_[2017.03.05_15.04.33].jpg (129KB, 1280x720px) Image search: [Google]

129KB, 1280x720px

>>61129701
>just watercool the mosfets

>>61129701
I love the fact this is not told in jest.

>>61129559
GPUs draw 300W and above and manufacturers somehow manage to design adequate cooling for their VRMs, most of the time at least.

>>61129734
Comparing the two is utterly retarded. Graphics card VRMs are generally hooked up right to the main fin array, and cooled by two or three fans. When you try to cool them with a tiny plate, you get EVGA tier temperatures and exploding cards. You can't have a fucking graphics card-sized heatsink and several 100mm fans on a motherboard. There'd be no room for the goddamn socket, RAM or anything else for that matter.

File: 1497288566148.png (23KB, 382x379px) Image search: [Google]

23KB, 382x379px

>>61129550
>$1000

>>61129701
>watercooling the vrm's is no longer a meme

>>61129773
You don't need a graphics card sized heatsink. If you watch the video, he's saying that the VRMs are literally better off with a little bit of airflow over them without the heatsink at all. So the heatsink in question is so ineffective and shit that you're better off removing it entirely.

>>61129822
probably due to the VRM design.

more than likely they will be designed to allow for better heatsink cooling if intel doesnt get their shit together

>>61129181
Skylake 18 core plus vega pro duo. All this connected with a car engine rad and fan. That beast would consume 1200w+ at full load.

>>61129895
Why even turn on your heater/AC during winter? You're practically getting a 2 in 1 deal here

>>61129550
this isnt even the 18 core model.

how the fuck will the 18 core model work at all even without overclocking?

>>61129822
He's putting a 120mm fan over them. That makes a huge fucking difference, heatsink or no heatsink. You cannot have any sort of fan in that position on a motherboard. It has to be passively-cooled, which is a huge fucking ask when dealing with a CPU that draws THAT much power.

You apparently don't have any concept of just how much difference airflow makes to cooling. It's why passively-cooled 750 Tis for example have giant fucking heatsinks on them and STILL run way hotter than the ones with tiny heatsinks and a single fan.

>>61129915
By running sub 3.2GHz clocks

>>61129915
Either Xeon level clockspeeds or new motherboards that can handle 200+ watts at stock.

>>61129931
>pump more power

Oh yeah I wonder how that will turn out.

File: the chernobyl of CPUs.png (214KB, 681x429px) Image search: [Google]

214KB, 681x429px

>>61129946

File: TCO2QFw.png (84KB, 653x726px) Image search: [Google]

84KB, 653x726px

>>61128939


Just put a heatsink on VRM and problem solved!

>>61129957
You still got the melting mosfets, but hey you're off to a start!

>>61128939
>temps are so crazy they run the risk of melting the PSU cables
I realize the housefire meme is just a joke, but I think Intel mighty actually burn somebody's house down with this shit.

>>61129773
>Graphics card VRMs are generally hooked up right to the main fin array
This is completely incorrect, the main heatsink usually cools the GPU alone, not even the VRAM is included. In any case all the heatpipes start from the GPU. The VRAM and VRMs are often either cooled by a plate or by smaller heatsinks underneath the big one for the GPU. They get (warm) airflow from the fans which first blow through the main GPU heatsink.

The EVGA burning VRMs did not happen because they were cooled by a plate, but because they weren't fucking cooled at all. The VRMs on the back were covered by the backplate but WERE NOT touching it. The "mod" EVGA distributed simply had people add a thermal pad between the VRMs and the backplate, which did in fact provide completely adequate cooling.

The issue on X299 can only be caused by utter shit thermal pads but more likely than not the real problem is that there's no airflow and the blocks of metal they slapped on top the MOSFETs have no fins and a very small heat dissipation area. If they designed an actual heatsink instead of a useless piece of painted metal meant to look cool the problem would not exist. Hell, if the problem were so severe they could've even used a small, quiet temperature-controlled fan for the power delivery. Those used to be a thing back in the day, mobos had a little fan and the VRM/northbridge/southbridge had real radiators and some were even linked together with heatpipes. They must've forgotten what a heatsink looks like since they learned about RGB decorations.

File: 1497565895125.jpg (112KB, 691x771px) Image search: [Google]

112KB, 691x771px

>>61129975

Shut up, goy!

File: 1424742573485.jpg (159KB, 900x900px) Image search: [Google]

159KB, 900x900px

>>61128939
Dumb Frenchman.

File: 1475211353488.jpg (108KB, 782x782px) Image search: [Google]

108KB, 782x782px

>>61128939
DELETE THIS VIDEO NOW

ISRAELS INTEL PLANT MUST STAY OPEN

SHREEEEEEEEEEEEEEEEEEEEEEEEEE

File: hitlfasg.webm (2MB, 900x660px) Image search: [Google]

2MB, 900x660px

>>61130021
He's German you fucking retard

>>61129999
>The EVGA burning VRMs did not happen because they were cooled by a plate, but because they weren't fucking cooled at all.
Wrong, it was a bad batch of capasitors. Gamers Nexus tested this quite accurate.

File: 1454230755807.jpg (20KB, 418x359px) Image search: [Google]

20KB, 418x359px

>>61130095
<he took the bait

File: 1496948335603.png (234KB, 800x612px) Image search: [Google]

234KB, 800x612px

>>61130154
>Wrong, it was a bad batch of capasitors.
>tfw, bad capacitors make your motherboard pull 300 W through an 8 pin

File: 1348764844816.png (2KB, 244x226px) Image search: [Google]

2KB, 244x226px

>>61130212

>>61130215
How is that even comparable?

>>61130154
EVGA literally distributes thermal pads for the back side of the PCB where VRMs are located and provides instructions on how to install them and have also released a VBIOS version which adjusts fan control in order to keep the MOSFETs cooler. The manufacturer quite clearly explains the issues they found and how they've been addressed on their own, official site.
https://www.evga.com/thermalmod/

File: 1473093430877.jpg (123KB, 800x800px) Image search: [Google]

123KB, 800x800px

>>61130238

>>61129999
Yeah, it's funny how copper heat sinks for increased cooling performance used to be a selling point and now it's just some abstract aluminium extrusions with a smooth painted surface because who needs heat transfer anyway.

File: 1488441421203[1].png (162KB, 633x900px) Image search: [Google]

162KB, 633x900px

>>61129258
Hahaha, this guy.

File: 1483378412042.jpg (64KB, 800x800px) Image search: [Google]

64KB, 800x800px

>>61129258

File: intel ceo JUST.jpg (119KB, 1000x1000px) Image search: [Google]

119KB, 1000x1000px

>>61128939
>90c cables potentially catching fire
>downclock to 1.2 GHz due to VRM temps
>gaymur bling instead of properly designed heatsinks

X299 just keeps getting better and better.

Prime95 is not a valid benchmark. No real world program will ever stress a CPU that hard.

>>61130371
Prime95 wasn't a problem until Skylake started melting faces.

File: 1484577998946.png (66KB, 533x414px) Image search: [Google]

66KB, 533x414px

>>61130371
>Multi-core doesn't matter!
>Price/performance doesn't matter!
>Power usage doesn't matter!
>Temperatures don't matter!
>Soldered dies don't matter!
>*NEW* Locked platform features don't matter!
>*NEW* Synthetic loads don't matter!

File: total_meltdown.jpg (40KB, 391x254px) Image search: [Google]

40KB, 391x254px

>Intel earlier: Don't overclock i7-7700K
>Intel soon: Don't overclock i9-7900X

>>61130254
They just had to show that they acknowledged the problem. The temps were fine. Look it up.
http://www.gamersnexus.net/guides/2691-final-evga-vrm-thermal-torture-test-and-analysis

>>61130421
Those temps look like exactly what EVGA said, "marginally within spec" in a bad thermal environment, like a hot case. The original cards with the original VBIOS had the absolute minimal yet still adequate cooling for the VRMs, which the update and thermal pads greatly improved upon. Hitting 110C+ PCB temperature because of 40C air isn't great at all since a case with shitty airflow could very well get as toasty as that, even with a single card.

>>61130254
>fan control
>to keep mosfets cooler

I assume they tweaked the cpu fan profile then because the mosfets dont have fans. also goodluck using an aircooler, you need an aio which has its own fan curves makes that change irrelevant

>>61130596
The discussion is about EVGA GTX 1080 graphics cards.

File: rEMq5pM.webm (1MB, 352x640px) Image search: [Google]

1MB, 352x640px

>>61130610
oh, nvm then

File: 1471180292258.png (370KB, 759x422px) Image search: [Google]

370KB, 759x422px

>>61130371
God forbid people use demanding software

I've been out of the loop for a while. What other problems has been going on with X299?

>>61131859
EVERYTHING IS FINE

>>61131859
They aren't soldered anymore, this is leading to serious temperature issues even at stock.

>>61131859
Everything
Will
Be
Fire

So how hard would it be to make a waterblock on the caps, Vrm, and Mosfets?
i have a VRM block for looks basically.

Hold up
What's stopping us from making a giant motherboard water block and place it with thermal pads behind the motherboard?

I have a 120mm fan blowing on the back of my CPU socket that reduced temps by about 3C alone.

>>61131989
>>61132010
All you'd need to do is blow a fan on them. But the point is the cooling is not properly designed in the first place.

>>61131989
>>61132010
You'd still have to make sure to get a board with more than an 8 pin CPU power connector.

>>61130371
Pretty sure prime 95 is bullshit, I had a FX-4100 that ran stable at 4.8Ghz but if i tested it on prime95 always got bad results compared to other bench mark tools that actually did something instead of just run at 100%

>>61132080
>pretty sure my anecdotal evidence isn't bullshit

Let's just ignore that practically every review site, every decent overclocker and most stability testing guides tells you to use P95

File: watercooled i9-7980XE.jpg (104KB, 1280x720px) Image search: [Google]

104KB, 1280x720px

>>61129981

>>61132115
That won't do shit, heat is in a small area the size of a few fingernails.

It need a water jet cutter

>>61132137
Aim a fire hydrant at it?

>>61128939
>he uses hardware with the ME
HAHAHHAHAHHAHAHHAHHAHAHAHHAHAHAHHAHAHHAHA

>>61132137
Make sure not to turn up the PSI too high. https://www.youtube.com/watch?v=VfgbYSuZC2U

>>61132145
It'll explode before I connect the hoses, also you need more than 1 person for that job.

>>61132101
Isn't it common knowledge that prime95 shouldnt be used on anything post haswell, causes overvolting or something?

>>61132194
No.

File: intel persuasion kit.jpg (460KB, 2083x1405px) Image search: [Google]

460KB, 2083x1405px

>>61132375
lul.

>>61132405
Yeah I got a $10 iTunes GC which is literally useless for me. Intel is shitting themselves HARD. Don't know anyone in the district that uses Intel's iPOS or RXT for selling. AMD doesn't even need a shilling site to get increase their sales.

>>61132194
They patched it last year or so. It's completely viable for Skylake and beyond.

File: s-l1600.jpg (233KB, 1600x1066px) Image search: [Google]

233KB, 1600x1066px

What happened to the enthusiast hw industry?
Just look at the heatsink on the vrms on this x79 motherboard.

>>61132508
Too much bling, not enough brains.

File: s-l1600 (1).jpg (146KB, 600x800px) Image search: [Google]

146KB, 600x800px

>>61132508
Another x79 motherboard, look at the vrm heatsink, is a real fucking heatsink.

File: s-l1600 (2).jpg (287KB, 1000x933px) Image search: [Google]

287KB, 1000x933px

>>61132525
Looks at this x79 Dell guys, a fucking fan in the vrms.
Seriously fuck this gaymen shit, they ruined the whole industry, now the heatsink is just a fucking piece of plastic.

>>61131989
A passive heatsink would be enough if all boards didn't gone with the muh max gsymer branding meme where heatsinks don't have fins and just are rgb blocks with plastic on top

>>61132194
If you use it with avx it's avx all time on so it's burdens a lot on heat, especially after haswell. In the video he clearly states that he run without avx so it's clearly hot.

Also look at comments. D8bauer edpecifically say the socket alone can draw up to 400w

File: gigabyte-x79-up4.jpg (471KB, 1920x1094px) Image search: [Google]

471KB, 1920x1094px

>>61132550
>>61132525
>>61132508
X79-UP4 had a reasonable heatsink too.
It was also one of the first boards using IR's PWRStage.

>>61128939
Not at all to defend Intel and their rushed launch, but it seems the concrete problem is that the motherboard vendors consider the gaymen design on their MILITARYCLASS heatsinks more important than their actual thermal performance. If this makes them realize that they're making retards out of themselves with their heatsink and soldermask so-called "designs", I'd almost thank Intel for it.

>>61132525
But where are the rgbs? Rgbs can drop 20c

>>61132593
They probably didn't realize just how much of a power monster the CPUs were.

>>61132602
Did you watch the video? He got like a 20 °C improvement from *removing* the heatsink.

>>61132643
And blowing a fan on them.

>>61132648
Which was blowing on the heatsink before he removed it.

>>61132550
Leave it to dell to make a motherboard that looks better than 90% of the gamer garbage you get now. How hard is it to make a motherboard now a days that is just built to work, stay cool, and OC well, looks be damned

>>61132660
Those fans look like failbait.
There is a reason we stopped using fans on mainboards, they routinely fail.
Not to mention they sound like shit.
Big passive heatsinks is the way to go for VRMs, too bad the x299 boards use tiny heatsinks made from alu.

Bring back the copper monsters of time past.

>>61132525
Even my obnoxiously named ASRock X99M Killer has a real heatsink on the VRMs. It's not a very good heatsink, but it doesn't have to be.

>>61132691
That is the one failpoint I see, but your right bring back the monster copper VRM heatsinks from Socket 775 days if your actually serious about VRM cooling

>>61132691
You don't need a massive copper turd on your VRMs, they don't get that hot. Just a decent fin array is enough. Doesn't even have to be dense, in fact it's better if it isn't for passive cooling.

File: images (16).jpg (31KB, 512x288px) Image search: [Google]

31KB, 512x288px

>>61132525
Asus x99 ws

>>61128939
I honestly couldn't give a fuck about Intel's 10% faster than common sense CPUs we have atm housefires. What I'm interested in is their future projects and the overall future of CPUs. Intel isn't behind in performance per say but they got caught with pants down since they made no progress in 5 years. I doubt AMD will make any progress in near future too. Is there anything groundbreaking in 7nm department or are we all destined to use 15 year old CPUs from now on.

>>61132835
10nm will be slower than current gen. As per Intel's roadmap.
But will consume less power. But by how much is probably just in line with scaling.

>>61132835
>AMD will make any progress in near future too
There's a lot to improve on Ryzen. Zen 2/+ or what ever the fuck it's being referred to now may also have a 15% IPC increase. Going by previous slides from AMD.
However that may be less now due to Ryzen coming out over the IPC projected.

>>61128939
when the 18 core realeses there will be nuclear bombs eveywhere

>>61133206
Zen 2
Zen+ is what Epyc and Threadripper are using, it's just Zen with more interconnects, since Ryzen was never for 2S or more.

>>61133272
What? Zeppelin always had GMI-links. And EPYC forms inter-socket link using PCI-E lanes.

File: ARM-Artemis-10nm-1.png (90KB, 678x380px) Image search: [Google]

90KB, 678x380px

>>61133142

>>61133784
And they'll lose the clock advantage too.

>>61132525
I find it amusing $100 ebay chinkshit motherboards have better VRM cooling than the new shit coming out.

>>61132835
>Intel isn't behind in performance per say
Increasing power draw by 35% to gain 25% performance is not "ahead". It's just making a housefire. In terms of multicore they are definitely without a doubt behind.

The only crown they still hold is high clockspeed 4 cores that still run under reasonable temps/watts (when delidded, anyway).

>>61128939
Oven?!

Oy vey, it's anuddah shoah!!

AMD fags confirmed for anti-semites.

>>61130042
OOOOOYYYYYYY VVVVVVEEEYYYYYYY

>>61132525
Taiwan was a mistake, they fucked up the Ryzen motherboards as well

>b-but muh RBGs and military grade decals

The damage control is already beginning. Got this guy in another thread claiming he's going to replace the wires going to the 8-pin CPU power to prevent them from overheating. >>61139223

>shitty Dow Corning thermal paste makes it impossible to cool, even with AIOs
>ridiculously high power draw
>VRM heatsinks don't work
>cable temps so high they might catch fire
Jesus Christ Intel

>>61139980
It keeps happening. http://www.overclock.net/t/1631319/skylake-x-binning/150#post_26189935

File: SUPER HOT.jpg (95KB, 796x731px) Image search: [Google]

95KB, 796x731px

>>61129181

>>61142238
>>/pol/

File: nigger1.webm (3MB, 640x360px) Image search: [Google]

3MB, 640x360px

>>61142323
>muh fwheeeewws
>can't redirect properly
>>>/out/

>>61142323
Nigger.

>>61132602
m8, they get engineering samples months before launch with parts progressively getting closer to the real deal. there are some ES chips out there that are within a couple hundred MHz of their end product counterparts.

File: file.png (131KB, 1010x188px) Image search: [Google]

131KB, 1010x188px

>>61132375
Is this for real? They even typo'd their own products.

>>61144157
pooinloos wrote that page for (((intel))), give them some slack

>>61128939
water cooled x299 mobo when?

>>61132375
>let's conviniently forget about Ryzen 5
The Jew strikes again.

>>61144285
>water cooled x299 mobo when?
When cooling systems for nuclear reactors become available for the consumer market.

>>61128939
Stupid Nazi should have set the LEDs to blue.

>>61144157
>intel word salad naming convention
>not even their pr can keep shit straight

Can you blame them?

>>61140982
>AMD literally naming their cards after large stars aka. ginormous nuclear fusion reactors.

I mean, who thought this was a good idea?