The Skylake-X Mess Explored: Thermal Paste And Runaway Power

[dudmont]

For your motherboard reviews with this CPU, might I suggest that you look up the rated lifetime of the PWM caps?

If those are 5k caps, they'll only last 5000 hours at their rated temperature. Lifetime usually doubles for each 10C under that rating.

These CPUs often end up in entry level workstations. If the intended usage pushes the caps to 75 C, and they are the more common 5k/105C, then conventional wisdom indicates that 10% of the caps should fail by the 40,000 hour mark. In those cases where the CPU is fully loaded most of the time, this will occur 4.5 to 5 years of age. This means that there isn't much room for mistakes in the motherboard layout, case design, or airflow requirements. It's something that potential consumers should take into account if they want to get the most out of this platform.

I'd have to check in more detail to give any estimates for the MOSFET lifetime, but that's another factor to account for if longevity is a priority.

I think you're right on with this issue here. The usefulness of these cpus suggest that they'll be more than powerful enough to handle most tasks assigned to them for many years(maybe 8-10), but will the hardware last that long, and that is an interesting point.
People on the forums(usually gamers) all the time are asking about whether to upgrade their I5 2,3, and 4 series CPUS to newer models and the answer is almost all their queries is no. Their cpu is not a problem yet. Your post is in a similar vein of thought. People may not need more cpu resources 8 years from now, but will their hardware still be running? I'm not sure the mobo makers and intel want their hardware to still be running by that point. Forcing people to buy new is much more profitable than building really quality parts now that last long enough to actually reach obsolescence.
This has been an interesting article to read and equally interesting forum thread to read, thanks to all involved. Very thought provoking.

 

[ahnilated]

It would have been very interesting to see you take the chip you used in this test and delid it and use some liquid metal TIM to see the differences. I have delidded 2 of my processors and I saw 12-30C drop in temps on mine (intel chips i7 3770K and i7 4790k).

 

[dudmont]

We have enough frame rates today in gaming and we dont need overclocking at all. Why are you focusing on overclocking so much ?

as for professional work , people never overclock workstations with ECC memory.

when is this overclocking hype going to end ?

Why overclocking? just to show off ? and say I got 20 more fps in a game that is already above 100 fps ?

move on people ... i you dont need to overclock there is no need to pay extra money just to hit useless fps that is not going to change anything.

Tinkering is fun, that is why people overclock. Simple as that.

 

[10tacle]

We have enough frame rates today in gaming and we dont need overclocking at all. Why are you focusing on overclocking so much ?

Why overclocking? just to show off ? and say I got 20 more fps in a game that is already above 100 fps ?

move on people ... i you dont need to overclock there is no need to pay extra money just to hit useless fps that is not going to change anything.

What is your major problem? First, don't tell other people what they need/want. That's not your business. Second, overclocking a mid-range CPU like mine yields faster results in both gaming and productivity applications, closing my i5 4690K in towards an i7 4790K in performance (hyperthread using apps withstanding). Third, people buy 144Hz G-sync and Free Sync monitors these days, and another 15-20FPS can mean the difference of utilizing its potential. Or not.

Finally, overclocking has long been done to get cheap performance out of what otherwise would be a much more expensive CPU purchased for the relative same performance. This goes for overclocking video cards as well. I had two GTX 970s that were overclocked to reference 980 performance. So just because *you* see no need for overclocking doesn't mean others don't.

 

[derekullo]

"Nevertheless, motherboard manufacturers aren’t blameless when it comes to the issues we encountered at launch and continue battling today. Using more thermodynamic expertise and less flashy plastic pieces would have paved the way for brawnier motherboards at the same price points."

How is this the motherboard manufacturer's fault?

The article even mentions:

"The CPU throttles before the motherboard’s VRM has a chance to hit 100°C."

What more could ASUS or MSI have done to make this more stable?

The heat issue lies within the cpu.

 

[Guest]

18 Core variant won't run more than 3.5GHz in Turbo. Haswell-E and Broadwell-E 18 cores CPUS i have, have same issue Skylake X will have, HEAT. As I said, given 14nm fabrication process these CPUs are built on that is your physical limit sort of speak. I think it is rather going overboard by complaining about heat 10/20 produces at 4.0 GHz+ clock speed. I noticed i got lot of dislikes which means i am right.

 

[Foeke]

I don't het the fear for deliding and put the cooler directly on the chip. Intel Pentium 2 and 3 and AMD Duron and Athlon from before 64 bit where naked. Never saw one that was destroyed. And these dies are probably even larger. By that it should be more stable.

 

[Guest]

If you think that AMD Thread Ripper will perform faster than Intel counterpart, think again. What is good about AMD Thread Ripper is $800 price tag for 16/32 CPU. Again a clock speed of it won't go over 3.4-3.5Ghz in Turbo (1-2 cores) only which makes it useless for gaming, still great for productivity. Same for 10+ cores Intel variants.

 

[ElMojoMikeo]

I can't say I am totally surprised. I have a 3970X aka 'The Power Eater'. I have never dared to run it flat out at anything around 4.5GHz. The heat build up eventually creeps up towards 91C package temp. As the review points out. However, it is still a grade A slice of silicon as near perfect as Intel can get. Hence the price.
I am using a Corsair H100i and it is a process of trial and error to balance the clocks and keep temps in a safe place. But once this is done it is still a very, very fast six core processor. The 18 core variant is gonna need a fridge.

 

[g-unit1111]

Also i might want to add is to wait for second iteration of x299 boards. The first batch is a joke from cooling point of view. Evga is one of the companies which will get it right. X299 need copper based cooling for VRM and chipset and also 2x8pin CPU connectors with recommended PSU of 1000W+. That's how i would run x299 setup.

Are you serious? Most EVGA motherboards are repackaged Foxconn boards. And they've never really had a solid track record with Intel motherboards. There's much better out there you can buy, even for X299.

 

[ElMojoMikeo]

That might well be the best way to go. My board is X79 (Sabertooth). A really good MB but a very similar problem as in the test system. Once the VRMs heat up they seem to transmit heat towards the processor at a fairly unnerving rate. I have three alarms set starting at 80C and upto 91C. I don't see this as anything new! Until this beast arrived I don't think any cooler has had to deal with 105C. So given that a decent cooler can't handle my six core and keep it under 91C. So yes if this is a problem in board design it hasn't just happened!

 

[DSzymborski]

I'm starting to have Prescott/Tejas flashbacks. Please send help.

 

[ElMojoMikeo]

Haaaaa! Which one of us is Prescott?

 

[derekullo]

Prescott was Pentium 4.

Pentium 4 was intel's premier space heater of the 2000s.

https://en.wikipedia.org/wiki/NetBurst_(microarchitecture)#Scaling-up_issues

 

[bloodroses]

Haaaaa! Which one of us is Prescott?

My guess is it would be someone named Sydney. Athlon was the Ghostface killer back then.

 

[Malik 722]

after seeing the review i am thinking of using my car cooling system to cool down the intel new raging bull.

 

[dudmont]

after seeing the review i am thinking of using my car cooling system to cool down the intel new raging bull.

This could be the situation where setting up your rig to run out of a spare chest freezer makes sense again. :/

 

[JamesSneed]

If you think that AMD Thread Ripper will perform faster than Intel counterpart, think again. What is good about AMD Thread Ripper is $800 price tag for 16/32 CPU. Again a clock speed of it won't go over 3.4-3.5Ghz in Turbo (1-2 cores) only which makes it useless for gaming, still great for productivity. Same for 10+ cores Intel variants.

The 1950X Threadripper part is rumored 3.4 Ghz base. I agree Intel will have boosts on there 16 core part around 3.4-3.5Ghz. Which is exactly why Threadripper may outshine Intel as a workstation part.

 

[Crashman]

Although we're using some of the highest-end and most expensive cooling hardware available, we still measure up to a whopping 71 Kelvin difference between the cores' reported temperature and the heat spreader's top.

Kelvin eh? Now you're just trying to show off. 90% of your readers know that Kelvin scales identically to Celsius even though the zero point is different. Maybe only 80% of those who know this remember what absolute zero is. Still, it appears that you're just using alternative names to broadcast intellectual superiority, and it's only working on a bit more than 10% of your readers. As for that ego, stick a pin in it.

Or maybe I'm <cough> reading you wrong <cough> and you're just tired of using °C repeatedly...

 

[JamesSneed]

Although we're using some of the highest-end and most expensive cooling hardware available, we still measure up to a whopping 71 Kelvin difference between the cores' reported temperature and the heat spreader's top.

Kelvin eh? Now you're just trying to show off. 90% of your readers know that Kelvin scales identically to Celsius even though the zero point is different. Maybe only 80% of those who know this remember what absolute zero is. Still, it appears that you're just using alternative names to broadcast intellectual superiority, and it's only working on a bit more than 10% of your readers. As for that ego, stick a pin in it.

Or maybe I'm <cough> reading you wrong <cough> and you're just tired of using °C repeatedly...

What is funny is 71 Kelvin falls right in the range liquid nitrogen is a liquid since it freezes at 63K and boils at 77K. I just hate when my CPU hits a whopping -202C, terrible times indeed.

 

[Aspiring techie]

We have enough frame rates today in gaming and we dont need overclocking at all. Why are you focusing on overclocking so much ?

as for professional work , people never overclock workstations with ECC memory.

when is this overclocking hype going to end ?

Why overclocking? just to show off ? and say I got 20 more fps in a game that is already above 100 fps ?

move on people ... i you dont need to overclock there is no need to pay extra money just to hit useless fps that is not going to change anything.

Tinkering is fun, that is why people overclock. Simple as that.

Only if for free . but you will need to pay for water cooling , and more expensive RAM , and more expensive Motherboard that will put you in the higher range of the next GPU and/or the next CPU.

300$ difference inside these will make you buy the better GPU.

What is the point ?

People overclock because it's a fun hobby. If someone wants to invest extra money into his hobby, (so long as he can afford it) it is perfectly fine and normal.

 

[Aspiring techie]

Although we're using some of the highest-end and most expensive cooling hardware available, we still measure up to a whopping 71 Kelvin difference between the cores' reported temperature and the heat spreader's top.

Kelvin eh? Now you're just trying to show off. 90% of your readers know that Kelvin scales identically to Celsius even though the zero point is different. Maybe only 80% of those who know this remember what absolute zero is. Still, it appears that you're just using alternative names to broadcast intellectual superiority, and it's only working on a bit more than 10% of your readers. As for that ego, stick a pin in it.

Or maybe I'm <cough> reading you wrong <cough> and you're just tired of using °C repeatedly...

What is funny is 71 Kelvin falls right in the range liquid nitrogen is a liquid since it freezes at 63K and boils at 77K. I just hate when my CPU hits a whopping -202C, terrible times indeed.

Yeah, such a pity when mine hits those temps. I prefer liquid helium myself...

 

[10tacle]

First , we are talking , no one has any problems.

Clearly you have a problem. You try to tell others what they do and don't need like a dictator.

second , this article is about a Very expensive CPU .. not about a cheap CPU you want to get better from it.

Your words:

you dont need to overclock there is no need to pay extra money just to hit useless fps that is not going to change anything.

Are you even aware that others can take that to mean overclocking ANYTHING?

and finally , this is also not about GTX 970 , it is about the new generation cards that destroyed the older generation .. and the coming cards will start at 11 Tflops to 25 Tflops are more than enough for 4K monitors.

What part of a 10-15% gain in FPS can mean the difference between smoother gameplay or not do you fail to understand?

yes in the past we needed overclocking. but today it is almost not that much of big difference.

Again, at higher resolutions above 1080p it means less. In most games (but not all).

and dont forget that Intel new Turbo 3 is more than enough as well .

^^Good luck with running an i3 on Ubisoft AAA games like Assassin's Creed series and The Division and other games that use big maps. I wish Tom's had a troll block option like Disqus.

 

[the nerd 389]

For your motherboard reviews with this CPU, might I suggest that you look up the rated lifetime of the PWM caps?

If those are 5k caps, they'll only last 5000 hours at their rated temperature. Lifetime usually doubles for each 10C under that rating.

These CPUs often end up in entry level workstations. If the intended usage pushes the caps to 75 C, and they are the more common 5k/105C, then conventional wisdom indicates that 10% of the caps should fail by the 40,000 hour mark. In those cases where the CPU is fully loaded most of the time, this will occur 4.5 to 5 years of age. This means that there isn't much room for mistakes in the motherboard layout, case design, or airflow requirements. It's something that potential consumers should take into account if they want to get the most out of this platform.

I'd have to check in more detail to give any estimates for the MOSFET lifetime, but that's another factor to account for if longevity is a priority.

I think you're right on with this issue here. The usefulness of these cpus suggest that they'll be more than powerful enough to handle most tasks assigned to them for many years(maybe 8-10), but will the hardware last that long, and that is an interesting point.
People on the forums(usually gamers) all the time are asking about whether to upgrade their I5 2,3, and 4 series CPUS to newer models and the answer is almost all their queries is no. Their cpu is not a problem yet. Your post is in a similar vein of thought. People may not need more cpu resources 8 years from now, but will their hardware still be running? I'm not sure the mobo makers and intel want their hardware to still be running by that point. Forcing people to buy new is much more profitable than building really quality parts now that last long enough to actually reach obsolescence.
This has been an interesting article to read and equally interesting forum thread to read, thanks to all involved. Very thought provoking.

What I meant was that these boards toe the line in terms of longevity, if my assumptions are correct, and if they were only using a single phase.

Since the VRM phases are not redundant (they divide the load up, and all phases are needed to handle the load), it only takes one failure to brick a board. The odds of a at least one phase out of a set of 12 phases failing in the first year (assuming the L90 is 4.5 years, as above) is a whopping 25%. Note that I'm using a pretty basic estimation procedure to get that number, so it's probably only accurate to +/- 10%. However, even a 15% failure rate in the first year is ridiculous for a motherboard. I don't think you could find a z270 board that bad if you crossed a Biostar with an ECS.

I'm not talking about these boards making it through an extended life. I'm talking about whether or not they'll make it through the first year. I'm not even using unrealistic figures for any of this. Aside from the thermal issues, these CPUs would be among the best values on the market for rendering until ThreadRipper comes out. The thermal issues unfortunately prevent you from using these to their full capacity.

 

[bit_user]

It'd be even more extreme to leave the die exposed and use a good torque screwdriver to minimize the possibility of mechanical damage from a non-uniform/excessive load.

Seriously?

Direct-die cooling is hardly new. Doing it correctly requires a specially machined shim to avoid excessive/uneven loading on the die. It's still not without risks, of course.