Core i9-7900X Skylake-X Review Shows Up

[EarthDog]

Some of us aren't very comfortable having our CPUs running that damn hot. I get scared when my CPU starts running hotter than 70c.

Ok, this is an exercise in insanity, continuing.

Good luck guys. Enjoy your cool running CPUs while spreading fud and showing ignorance through your opinion on the matter.

 

[Deleted member 172152]

Ive went over it before in the thread and my previous post (anyone reading them or has this degraded into trolling? Feels like i make points that are missed completely, lol!)...as ive said, ive run chips 100% at close to 90c for YEARS. That doesnt mean some wont fail, but its not even remotely close to the doosmday prophecy about it a few of you have running high temps the silicon is designed for.

For someone that doesnt f@h where cpu is running like that 24/7, in the vast majority of cases, it will easily last well past its warranted life.

Which cpu's though?

 

[notb]

Some of us aren't very comfortable having our CPUs running that damn hot. I get scared when my CPU starts running hotter than 70c.

But why does this scare you?
Why do you think 70*C is "hot" for a CPU?

 

[HTC]

But why does this scare you?
Why do you think 70*C is "hot" for a CPU?

I think the problem with that is that the temps inside the case will increase, leading to a temp increase in every component inside of the case. Sure: the CPU can take it easily enough but the rest of components may not have such a high tolerance.

It's the reason why reference GPUs, though generally noisier then AIB GPUs are sometimes preferred: because they expel the hot air directly off the case instead of into the case (to be extracted by case fans afterwards).

As an example: when i play a game for an extended period, my HDD's temps increase by around 8ºC, even though i have a 20 cm intake fan in front of the HDDs and 2 exaust top fans + 1 back exaust fan (all 12 cm). Why? Because my GPU is an AIB one and dumps the hot air inside the case, thus increasing the ambient temp inside the case.

 

[EarthDog]

Which cpu's though?

Nehalem(X58) on forward. Sandybridge...Ivybridge...Haswell...Broadwell...Haswell-E...Devil's Canyon...Broadwell-E...Skylake.

Obviously some of those arent past their warranty, but, Im guessing you get the picture.

I think the problem with that is that the temps inside the case will increase, leading to a temp increase in every component inside of the case. Sure: the CPU can take it easily enough but the rest of components may not have such a high tolerance.

It's the reason why reference GPUs, though generally noisier then AIB GPUs are sometimes preferred: because they expel the hot air directly off the case instead of into the case (to be extracted by case fans afterwards).

As an example: when i play a game for an extended period, my HDD's temps increase by around 8ºC, even though i have a 20 cm intake fan in front of the HDDs and 2 exaust top fans + 1 back exaust fan (all 12 cm). Why? Because my GPU is an AIB one and dumps the hot air inside the case, thus increasing the ambient temp inside the case.

True... but big deal!

It's not like those components are going out of their spec either! A few C warmer doesn't hurt most components...

 

[Kyuuba]

But how long can the chip withstand those temperatures? We've already seen that the cheap shitty TIM that Intel uses in their chips turns to clay after some time due to extreme temperature fluctuations. Eventually the TIM won't be able to conduct the heat properly and you're going to eventually lose the ability to overclock the chip due to the inability to get rid of the heat fast enough. It may be OK for the chip to run at 90c in the short term but in the long term it's going to be very bad news.

Don't worry, it will last more than 4 years running as is, i still have a 4770K that always ran 80+ degrees, still just fine man, these chips are designed for that, don't panic.

 

[HTC]

Nehalem(X58) on forward. Sandybridge...Ivybridge...Haswell...Broadwell...Haswell-E...Devil's Canyon...Broadwell-E...Skylake.

Obviously some of those arent past their warranty, but, Im guessing you get the picture.

True... but big deal!

It's not like those components are going out of their spec either! A few C warmer doesn't hurt most components...

While true, an increase of over 20% in mechanical HDDs is quite allot.

 

[Deleted member 172152]

Don't worry, it will last more than 4 years running as is, i still have a 4770K that always ran 80+ degrees, still just fine man, these chips are designed for that, don't panic.

Not quite nigh 100 degrees though is it? Intel paste has always been a slight issue, but now it's a major issue because of the even higher temps, and we're only at 10 cores of the 18!

 

[Vlada011]

Some of us aren't very comfortable having our CPUs running that damn hot. I get scared when my CPU starts running hotter than 70c.

On my ex platform Ivy Bridge paste was under IHS and CPU was much hotter on 4.5GHz than i7-5820K on 4.2GHz.
I compare because they worked on same voltage 1.200V. But hot air from H100 (exhaust) is hotter with i7-5820K than with i7-3770K.
That mean that with thermal paste transfer of heat from CPU to H100 block/liquid/radiator was not well and a lot of heat stayed inside in CPU and H100 was able to remove only small amount of heat 40-50% example compare to 70-80% with i7-5820K. This numbers are not exactly accurate but example... And no matter what you do before change something under IHS or maybe even than you can't get better temps with much expensive coolers, I mean you get better temps but maybe 20C. No way to processor stay 60C on huge ammount of voltage and clock if liquid is cold in some expensive custom loop. Only way is to freeze CPU with LN2 or phase change or dice.

I think that's biggest mistake for X299 thermal paste.
Because some people are confused, they want to OC, but what to do, they remove IHS, and than what, how much paste, how to do that... You need to be far more precise than installatin paste on IHS and CPU block over that. Than questions after installation... Did I done everything nice, maybe die no contact with IHS any more, maybe is to much paste... What will happen if CPU die, people used on Intel OC warranty, to pay 30$ more and OC.

Positive thing is because Intel go with Intel Turbo Boost 3.0 and that mean CPU specification is 4.3 and 4.5GHz and everyone know with such high Turbo space for OC is 200-300MHz.
And many people will not OC, imagine Turbo 3.9 GHz and people try to reach 4.6-4.7GHz with paste.
That would be bad. Intel is not probably so stupid to say Turbo Boost 3.0 if CPU throttle and can't cool down with AIO cooler on fabric Turbo frequency.

From other side Ivy Bridge worked one night and more 15h test on more than 90-90C, he behave like 50C.
Nothing, no problems at all. I can't guarantee that now CPU could work on such high temperatures without throttle.
I didn't OC mine Haswell-X over 4.2GHz because I want to stay on 60-65C and don't want fans on high performance to work on more than 2000 RPM.
It would be bad if people figure out that i9-7900X work on 80C on 4.5GHz. I think they change something, because i7-5960X and i7-6950X are hot as hell on 4.5GHz.
Only people with watercooling and 280mm AIO with good samples are capable to keep on such clock and small voltage, and flux solder help a lot in that case.

 

[EarthDog]

While true, an increase of over 20% in mechanical HDDs is quite allot.

Instead of random statements, here is one backed up with data...

Overall, there is not a correlation between operating temperature and failure rates. The one exception is the Seagate Barracuda 1.5TB drives, which fail slightly more when they run warmer.

https://www.backblaze.com/blog/hard-drive-temperature-does-it-matter/

Not quite nigh 100 degrees though is it? Intel paste has always been a slight issue, but now it's a major issue because of the even higher temps, and we're only at 10 cores of the 18!

NOBODY here is saying run it at 100...Ive said over and over 90C.

I like how you keep repeating the same thing OVER AND OVER without a bit of data to support your thoughts... maybe people will believe it eventually...

 

[Fasola]

Instead of random statements, here is one backed up with data...
https://www.backblaze.com/blog/hard-drive-temperature-does-it-matter/

NOBODY here is saying run it at 100...Ive said over and over 90C.

I like how you keep repeating the same thing OVER AND OVER without a bit of data to support your thoughts... maybe people will believe it eventually...

Irrelevant, as you're not going to get those backblaze temperatures in your case with a toasty 90C CPU unless you're pushing some serious air. And what if you're using several drives and/or living somewhere with really hot summers.

 

[HTC]

Instead of random statements, here is one backed up with data...
https://www.backblaze.com/blog/hard-drive-temperature-does-it-matter/

NOBODY here is saying run it at 100...Ive said over and over 90C.

I like how you keep repeating the same thing OVER AND OVER without a bit of data to support your thoughts... maybe people will believe it eventually...

My HDDs reached 44ºC: a bit too hot for my liking.

Since then, the ambient temp has increased (expected since we're nearing the Summer): as such, should i game for an extended period, the temps would probably exceed that.

Disregard the 49ºC in the pic: there was something wrong with the temp sensor in the program @ the time.

 

[EarthDog]

Irrelevant, as you're not going to get those backblaze temperatures in your case with a toasty 90C CPU unless you're pushing some serious air. And what if you're using several drives and/or living somewhere with really hot summers.

relevent.. no correlation in temps.

If you are using several drives, do something about it cooling wise IF NEEDED.

Everything is to 'people's liking and preference'...which is fine, but still within operating parameters which seems lost. 90c is dubbed 'insane' yet its within operating parameters... 44c hd is too hot for someones liking...but within operating parameters...

I worry when i break those values for extended periods, but here nobody has even hit them.

 

[Fasola]

relevent.. no correlation in temps.

If you are using several drives, do something about it cooling wise IF NEEDED.

Everything is to 'people's liking and preference'...which is fine, but still within operating parameters which seems lost. 90c is dubbed 'insane' yet its within operating parameters... 44c hd is too hot for someones liking...but within operating parameters...

I worry when i break those values for extended periods, but here nobody has even hit them.

Irrelevant, because even the top temperature in their test is way below the point where warning signs pop-up for hard drives (HDD Sentinel says it's 43).

Yeah, do something, like get a CPU that runs cooler/can transfer heat properly.

I've hit 61 while doing a build in the summer with many hard drives. I was still testing stuff with the open case with no case fans mounted and I just forgot to check temperatures.

 

[EarthDog]

Relevant.. i dont care what 3rd party software says... read the white papers for the drives for operating temps.

61C... cool.. well, not cool, lol!? I mean, i can fight your side too, but, the reality is, fringe siutations (open air stack of drives no fans) need other methods.. for normal operations though, not so much.

 

[trparky]

90c may be within operating parameters but I can't imagine that that's very good for the processor over extended periods of time. Remember, heat is the enemy of electronics. Any time you're running that close to the limits you put undue stress on the components and that shortens the life of the component.

 

[notb]

I think the problem with that is that the temps inside the case will increase, leading to a temp increase in every component inside of the case. Sure: the CPU can take it easily enough but the rest of components may not have such a high tolerance.

They won't. AMD and Intel CPUs are emitting similar heat, it will result in similar temperature inside the case.
The only advantage you get from soldering is that the heat is transferred quickly from the chip to IHS. That's important for the chip, but fairly irrelevant for everything else inside the case.

Current Intel CPUs run hotter than AMD's (because of TIM), but tolerate much higher temperature as well.

As an example: when i play a game for an extended period, my HDD's temps increase by around 8ºC, even though i have a 20 cm intake fan in front of the HDDs and 2 exaust top fans + 1 back exaust fan (all 12 cm). Why? Because my GPU is an AIB one and dumps the hot air inside the case, thus increasing the ambient temp inside the case.

No. It's because of how the airflow looks in your case.
Your GPU is pushing a lot of hot air downwards, but - as you've said - the fans in your case suck air in front and push it out in the top-rear part.
Essentially, the hot air from the GPU is first sucked towards the HDDs - hence making them warmer.

 

[Fasola]

Relevant.. i dont care what 3rd party software says... read the white papers for the drives for operating temps.

61C... cool.. well, not cool, lol!? I mean, i can fight your side too, but, the reality is, fringe siutations (open air stack of drives no fans) need other methods.. for normal operations though, not so much.

Yeah, I'll still be sceptical of a study with temperatures nowhere near my own.

61C was just an example to your nobody hits 44 comment. The drive in question is an older one (operating range 0-60C) and does run several degrees hotter than the rest, even when not stacked and with a dedicated fan blowing. During the day it stays at around 43-45 and it hasn't been the hottest of summers. There's some tweaking to be done with the fans and I'll probably have to forego any pretense of silence for now.

Anyway, my point is that every little bit helps, and with paste, they've added another element into the whole cool/silent PC equation just to save a little money on very expensive hardware.

 

[notb]

90c may be within operating parameters but I can't imagine that that's very good for the processor over extended periods of time. Remember, heat is the enemy of electronics. Any time you're running that close to the limits you put undue stress on the components and that shortens the life of the component.

How do you know it's "that close to the limits"? It's below the safety threshold - that's all we should care about.

BTW: what's the threshold for Ryzen / Threadripper? I'm looking for the Intel's Tcase analogue.

 

[HTC]

They won't. AMD and Intel CPUs are emitting similar heat, it will result in similar temperature inside the case.
The only advantage you get from soldering is that the heat is transferred quickly from the chip to IHS. That's important for the chip, but fairly irrelevant for everything else inside the case.

Current Intel CPUs run hotter than AMD's (because of TIM), but tolerate much higher temperature as well.


No. It's because of how the airflow looks in your case.
Your GPU is pushing a lot of hot air downwards, but - as you've said - the fans in your case suck air in front and push it out in the top-rear part.
Essentially, the hot air from the GPU is first sucked towards the HDDs - hence making them warmer.

But the HDDs are between the intake fan and the GPU: how can the air be sucked in the direction of an intake fan? Shouldn't it go in the other direction?

 

[FR@NK]

I think the problem with that is that the temps inside the case will increase, leading to a temp increase in every component inside of the case. Sure: the CPU can take it easily enough but the rest of components may not have such a high tolerance.

True... but big deal!

Heat output is measured in watts not degrees.

The same cpu running at the same speed and workload will put out the same amount of heat regardless if the cooling system can get it down to 60C or as high as 90C. The difference in degrees between ambient temperature and the die temperature is just a measure of how efficient the cooling system(TIM or solder included) is at removing the heat from the die. Its not a measure of heat output. And thus wont heat the inside of the case any more or less.

 

[trparky]

How do you know it's "that close to the limits"? It's below the safety threshold - that's all we should care about.

Historically 100c was the limit at which the chip automatically throttles. Running at 90c under load doesn't leave much room for any error in possible cooling setups. And we're not even talking about AVX instructions that are known to really heat the chip up.

 

[claes]

Heat output is measured in watts not degrees.

The same cpu running at the same speed and workload will put out the same amount of heat regardless if the cooling system can get it down to 60C or as high as 90C. The difference in degrees between ambient temperature and the die temperature is just a measure of how efficient the cooling system(TIM or solder included) is at removing the heat from the die. Its not a measure of heat output. And thus wont heat the inside of the case any more or less.

So much this!!!

A 95W CPU running at 100* is going to put out the same heat as a 95W CPU running at 15* (assuming both are consuming 95W/at load).

(Also, 200mm fans are useless in a PC - dunno why so many TPU users are under any other impression)

Instead of random statements, here is one backed up with data...
https://www.backblaze.com/blog/hard-drive-temperature-does-it-matter/

While this is true, Backblaze's disks, on average, run at a maximum 31*, with a few disks running as high as 38*, so it doesn't really tell us what happens to a disk at >40*

If you look to Google's and Microsoft's results you'll still find little correlation (except at cold temperatures; see links at BB source), but you will see that drives at high temperatures fail sooner than drives at low temperatures in the case of Google, and in Microsoft's case, where drives were allowed to run closer to their operating limit, there's a more obvious correlation at high temperatures.

 

[EarthDog]

Heat output is measured in watts not degrees.

The same cpu running at the same speed and workload will put out the same amount of heat regardless if the cooling system can get it down to 60C or as high as 90C. The difference in degrees between ambient temperature and the die temperature is just a measure of how efficient the cooling system(TIM or solder included) is at removing the heat from the die. Its not a measure of heat output. And thus wont heat the inside of the case any more or less.

+1.. tell me something I don't know!!! Great post!

90c may be within operating parameters but I can't imagine that that's very good for the processor over extended periods of time. Remember, heat is the enemy of electronics. Any time you're running that close to the limits you put undue stress on the components and that shortens the life of the component.

LOLOLOL.. cant take it anymore...

"within operating parameters..."

Can't imagine that's good for the processor...


SAY WHAT?!


Since the temps there were low... here is a wiki article from google's data showing the same trends with higher temps.
https://en.wikibooks.org/wiki/Minimizing_Hard_Disk_Drive_Failure_and_Data_Loss/Environmental_Control

 

[trparky]

OK, let's stick you in the middle of Death Valley with a 140 F temperatures. Sure, humans can survive at higher temperatures so long as we remain hydrated but it will be pushing our systems to the critical point. You won't be a very happy camper. Or if you run an engine at high temperatures, sure the engine can handle the temperatures but eventually the heat will cause metal fatigue and it'll blow itself apart.

I tend to err on the side of caution, I never push things to the breaking point.