Intel Core i9-10900K at defaults stressed, Power Consumption Reads 235W, Temperatures hits 93°C

Makes me wonder if a Custom watercooling job would improve anything? Or even a 360 AIO? Yeah I know that this is max torture test for the CPU and the temps will be lower in normal usage.

Also that user was using an inefficient motherboard so that problem could have been prevented. AMD and the websites that sold the FX9590 such as Newegg recommended a FX990 board and they even sold a Mobo CPU combo to make sure the user got the correct hardware.

 

Fair enough. The point is though.... AMD had very hot CPUs before Ryzen. They are doing better obviously, but things can always change. Which is why it is so stupid to have arguments over heat. And there are multitude of options to cool CPUs.

 

This is quite misleading. Just to be clear I think these new Intel CPUs are pretty lame and should not exist on 14nm.
But (partially!) the reason for the high temps is because the guy is also running a GPU stress test, which will contribute significantly to the CPU temps if it's inside a case. If it's an open testbench, dunno, never tried it.
Even so, good to keep in mind this is just AIDA64 FPU, which is hardly the most stressing test. prime95 will make it go to 300W with ease, likely above that if it's run with AVX2.

 

A true homage to The Pentium 4s of yore

 

Yeah, remember the Pentium 4, 3.04 ghz with hyper threading in early 2002? Amazing technology at the time, and it should have been included on every chip from that point out.

 

well to be fair to intel, they were in a rough spot, by 2003 the netburst roadmap evaporated and they had picked the pentium M uarch off the shelf and went on enhancing it inorder to compete with the amd, so probably would have been smt on every chip had that not happened.

 

So will they box this with a 240 AIO!? No, they will relax that turbo so it stays away from that temperature. If not it will probably kill it within 2 years.
Most of PC users don't even check their temps, imagine running that 4-5 hours per day gaming, it wont reach 93 with an AIO, but it will reach 93 with their stock junk of a cooler and a lot of pc users that game use that.

 

So you're saying that Intel have to "relax that turbo" or else these will die in 2 years? Let me get that straight.

You do realise that these temps were reached by stress testing the CPU and GPU both at the same time? And that under normal gaming load, most of those cores will do next to nothing and the package will chill in the 60C area?

 

I do, but i still think they are pushing them a bit to much..

 

how?

 

He posted a guy having problems. That is not the norm. Might as well link a thread where a guy first time built a ryzen PC and it's not booting then say "see? Ryzen cpus can't even boot"

Also temperature is a meaningless measure of anything. Power consumption tells you how much heat a chip is generating, not max Hotspot temperature of a 7nm chip. A static electricity spark is like 10000C. But rubbing your feet on the carpet is not going to heat your home up.

 

That still has nothing to do with him mumbling on about motherboard compatibility.

 

yeah especially fx 9590 was a desperate attempt of a chip pushed to 5 ghz from amd drawing 240 watts roughly and being a hot mess when companies are desperate to show that they competitive they do things like that.
I would highly disagree the temperatures are pointless and consumption . The more elaborate the demand on cooling and vrms are the higher the cost to get together everything and if we are talking about workhorse cpus that will be maxed out often if the consumption is only 10% different you get the top performer .if they are about equal in performance you get one that it's consumption is lower, why would you not at that point?

 

The "Intel Core i9-10900K" call a priest cause it's gonna get smokae'

I may want to play Doom where I go to Hell but I don't want my PC to actually take me there

 

They're not gonna die, people been using 7700Ks and 8700Ks at 5GHz, 24/7, at voltages around 1.3-1.4V. It's been more than 2 years for 8700K for example, and I haven't heard of anyone killing a chip with 1.4V if it's cooled adequately, i.e. it stays under 90-ish under load, although let's be honest, even at 100C it's not a big deal, it will throttle back as required.
If anything, this architecture is probably among the safest ever cause it's so endlessly rehashed. Even Silicon Lottery sold 8700Ks at 1.4-1.437V for 4.8-5.1GHz under AVX2, 5.3GHz without AVX. Those CPUs were stresstested in p95, and at 1.43V it likely sucked 300W with AVX2. And they're alive, you'll incredibly few complaints (if any) of Coffee Lakes dying because of OCs, even excessive ones like the SL sold, since most people recommended a top of 1.35V for safe operations.
So my guess is these CPUs are gonna be fine. They're still a product that should not exist and Intel should be ashamed, but they won't be doing much dying. And knowing the desperate amount of fanboys for specific corporations which is utter insanity, people will buy them like hotcakes just because of those 10 fps or w/e.

 

I disagree because reviewers need to do this stuff to tell others what the limits of the CPUs are so no-one pushes past the limits and kills their chip accidentally.

 

Jokes aside the 93C being referred to is under full stress I’m guessing? Still... anyone posted what the tjmax is for it? Would give a better idea what we’re looking at here

 

I guess cpu is always within specifications. Intel is not responsible for the temps inside your own case. Every rig is different. They just need to keep their cpus alive during stressing apps for reputation only not by obligation. I believe those harsh tests are a gray area regarding warranty.

 

Most newer Intel CPUs are 105. In BIOS, you can go up to I think 110 or 115.

 

These are sold-out on Newegg pre-order, $530.

Intel knows they will always sell.
I will be waiting for HH and the team to sort out the actual numbers we can expect.