Intel Core i9-9900K Overclocked To 6.9GHZ On All Eight Cores

[PaulAlcorn]

With the help of liquid nitrogen, famed overclocker Splave pushed his Intel Core i9-9900K sample to 6.9 GHz on all cores at Intel's event.

Intel Core i9-9900K Overclocked To 6.9GHZ On All Eight Cores : Read more

 

[thecoolman101205]

at first i was thinking it was smoking

 

[dreamalladream]

Not practical. It's good for bragging rights nothing more.

 

[jdlawrence]

I wonder how many CPUs it took to find one that would OC at 5.3 on all cores. I am curious if this will be common on all chips or was this the most perfectly binned chip.

 

[Math Geek]

Not practical. It's good for bragging rights nothing more.

i agree but note the "Splave told Tom's Hardware the Core i9-9900K can hit 5.3 GHz on all cores with watercooling and a 1.4V Vcore" in the article. that's impressive and easily practical for the average pc enthusiast.

 

[stevensen76]

why do any of you reviewers bother with giving us benchmarks with LN2? noone uses it out here in the real world so don't bother. Give us "Real World" results please. Air cooled, water-cooled etc. Thats it. Thanks!

 

[Ninjawithagun]

Another Intel lie. AMD was the first to release an 8 core/16 thread mainstream CPU over a year ago. Seriously? Tom's needs to do a better job vetting these stories.

 

[brokenthorn]

Not much of a difference to justify the price, from previous 8700K. It's basically a tighter binned part.

 

[stdragon]

At the speed of light, once cycle @ 1Ghz travels the distance in 30cm. Theoretically at current CPU die sizes, that can clock up to 20 Ghz before you run into signal propagation issues of one cycle still trying to finish while another begins. Essentially the speed of light becomes the the physical limit.

That all said, I'm rather surprised they could clock it that high at 7 Ghz without running into excessive timings issues from the PCIe bus all the way to the CPU and back again.

 

[TerryLaze]

Not much of a difference to justify the price, from previous 8700K. It's basically a tighter binned part.

/s Yeah pfff 33% more cores ,big deal... /s

 

[mgallo848]

@NINJAWITHAGUN,

Did you even bother to read the article before posting? Please show me where in this article does Tom's even MENTION AMD? Please show me where in this article does Tom's say that Intel was the first 8 core/ 16 thread mainstream CPU?

If you want to hate Intel and thump your AMD chest that's your opinion but at least read the article first.

 

[bit_user]

Now hurry up and do the 28-core Xeon W-3175X! That thing better hit 5 GHz on all cores, or someone's pants are definitely on fire!

 

[bit_user]

why do any of you reviewers bother with giving us benchmarks with LN2? noone uses it out here in the real world so don't bother. Give us "Real World" results please. Air cooled, water-cooled etc. Thats it.

Because it's news! Some people like breaking records, and others like reading about them. I'd be curious how fast you could clock these chips with LN2, but not enough to try it. Now I know!

You could ask the same thing about any kinds of records that people try to break, really. Sometimes, it's useful to know where the boundaries are, in case there's ever a need or strong desire to push beyond the status quo.

 

[bit_user]

At the speed of light, once cycle @ 1Ghz travels the distance in 30cm. Theoretically at current CPU die sizes, that can clock up to 20 Ghz before you run into signal propagation issues of one cycle still trying to finish while another begins. Essentially the speed of light becomes the the physical limit.

First, the speed of light in semiconductors is far slower than a vacuum.

Second, the wires on these chips aren't run in straight lines. In fact, they go out of their way to avoid doing that, due to the antenna effect.

Third, why do you think the signal has to traverse the entire chip in one clock cycle? That's not how it works. I think they basically express the timing target in terms of a certain number of logic gates, and then break up pipelines so that no stage is longer than that.

Finally, I don't really know, but I'd guess one limiting factor has something to do with the switching speed of the transistors.

 

[bobba84]

The AMD fan bum hurt is incredible in here... Just be happy that both companies are actually competing well with each other for once. On a side note, it's about bloody time we don't have to delid a CPU aimed at overclockers (K series).

 

[alextheblue]

i agree but note the "Splave told Tom's Hardware the Core i9-9900K can hit 5.3 GHz on all cores with watercooling and a 1.4V Vcore" in the article. that's impressive and easily practical for the average pc enthusiast.

It's impressive for watercooling. But I'd hardly call a rig built around a somewhat aggressively (for non-chilled liquid) overclocked 9900K "average". It's pretty much the definition of a top-of-the-line build using consumer-tier enthusiast-aimed components. Not HEDT-tier, but also faster than HEDT in typical hobbyist measures like frames per second (all else remaining equal).

it's about bloody time we don't have to delid a CPU aimed at overclockers (K series).

I guess Intel was tired of getting bad press for their 5 cents of paste. AMD played an indirect role here as well, by using solder fairly extensively. Now the question is... will this extend to more upcoming K series chips? For example an unlocked i3 with soldered IHS is more interesting to me. I understand if the locked chips don't, I mean who cares, they're locked.

 

[gmoonfrog17]

Yeah but how does any of this effect Arma 3?

 

[SSA-Ed]

Liquid nitrogen cooling? Wouldn't that cause moisture to accumulate around the CPU? Let's just soak the mobo in cold water.

 

[geof2001]

How about a story of how Intel is just trying to bury AMD with all this bogus news coverage or does Tom's really have that hard-on for Intel that you cover your front page with their "announcements and introductions".

 

[mgallo848]

How about a story of how Intel is just trying to bury AMD with all this bogus news coverage or does Tom's really have that hard-on for Intel that you cover your front page with their "announcements and introductions".

Judging by your logic I guess you should be equally mad at Tom's Hardware article about how AMD claims a 47% boost in Threadripper when in reality that number is only in BF1

Oh and by the way, trying to bury the competition is how the real world works. Sorry you're mad about it.

 

[Christopher1]

at first i was thinking it was smoking

Do not feel bad about that. Anyone who did not parse or comprehend the "Liquid N!" cooling would have thought that as well.

 

[cryoburner]

Liquid nitrogen cooling? Wouldn't that cause moisture to accumulate around the CPU? Let's just soak the mobo in cold water.

Cooling a processor with liquid nitrogen like this is only for attempting to break records, since its not practical for real-world usage scenarios. They typically put insulating material around the cpu socket to reduce condensation, and are only running the system temporarily, just to see see how high the clock rates can go, not for anything resembling long-term use. When performed at an event hosted by the manufacturer, it basically amounts to little more than fluff marketing.

 

[mgallo848]

not for anything resembling long-term use. When performed at an event hosted by the manufacturer, it basically amounts to little more than fluff marketing.

It doesn't diminish the accomplishments.
If we were to have that sort of attitude towards everything then we should say the same thing about everything that serves no consumer purpose. For example:

Should we get rid of Nascar, F1 racing, Quarter mile funny cars and MotoGP bikes because they aren't built for the average consumer? Nor are they built for longevety?

Sometimes it's about achieving new heights in performance and seeing what things are capable of doing. The results may not affect us today but they may sometime in the future.

 

[stevensen76]

why do any of you reviewers bother with giving us benchmarks with LN2? noone uses it out here in the real world so don't bother. Give us "Real World" results please. Air cooled, water-cooled etc. Thats it.

Because it's news! Some people like breaking records, and others like reading about them. I'd be curious how fast you could clock these chips with LN2, but not enough to try it. Now I know!

You could ask the same thing about any kinds of records that people try to break, really. Sometimes, it's useful to know where the boundaries are, in case there's ever a need or strong desire to push beyond the status quo.

Well, I guess now you know how fast someone else clocked a cpu chip with technology you will never use. Not sure how useful that knowledge is, but you now possess it. Didn't mean to get you sideways by writing my opinion, as I'm more interested in seeing how far you can push this cpu with tech I already own or can easily obtain and implement. But hey...6.9GHz...that is pretty amazing.

 

[JamesSneed]

Oh Intel and its PR. Here is the playbook.

1) Around 2 months out announce the CPU's are coming but give no oficial pricing.
2) Run some unofficial overclocked testing and let the results post into a testing suites database without providing any OC'ed details.
3) Get some review site to early post test results showing a much larger than expected performance increase over your competitor. Make sure its a 3rd party so it's not false advertising.
4) Give the CPU's to some overclockers to see what they can do with them then. Then tell everyone about the high clocks they hit on liquid nitrogen. (Not trying to diminish overclockers, but the timing Intel does this with each release.)
5) Place the CPU's on sale the same day as reviews to reap pent up demand from prior steps.