Intel Core i7-7740K Sets CPU Frequency And RAM World Records

[InvalidError]

Better do that in fairly high humidity. Styrofoam is horribly quick to get static buildup.

High humidity won't help you as the inside of the styrofoam cube will be all N2. You can spray the inside with conductive polymer.

 

[shrapnel_indie]

Better do that in fairly high humidity. Styrofoam is horribly quick to get static buildup.

High humidity won't help you as the inside of the styrofoam cube will be all N2. You can spray the inside with conductive polymer.

Point noted.

 

[rantoc]

24/7 stable overclocking, that's interesting. What can be done in the extreme, not so much really!

 

[10tacle]

https://www.youtube.com/watch?v=3nKQ5uDA8TM
AMD 8.4 Ghz in 2011. 7 Ghz, is it even worth mentioning?

Uhm, it was 7.5GHz, not 7.0GHz. Now considering that was six years ago and AMDs have different architecture meaning that direct GHz speed comparison with Intel is meaningless (a 3.5GHz i5 4690K destroys a 3.9GHz AMD A10 7870K), I'd say the worldwide attention deserves mention. Of course Tom's did include details:

Team AU broke the Kaby Lake CPU frequency record with a Kaby Lake-X Core i7-7740K clocked at 7.5GHz on an Gigabyte X299-SOC Champion. The team also broke the 3DMark03 and 3DMark06 world records, followed by taking the Aquamark world record.

 

[LilHammer99]

Who wants to bet that Intel paid for the fly-in to spin some positive press against the AMD Threadripper headlines?

 

[ddferrari]

Wow, what a veritable cornucopia of confusing responses!

"yea lets see some practical overclocking that the average consumer could do with water cooling"

There will be plenty of those. They have nothing to do with this. Two different things.

"All of these "records" are useless wastes of time."

Well, it's not your time so why do you care? Is everything that doesn't appeal directly to you a "waste of time", in your opinion?

"There isn't much of a point in beating OC records if the result is a CPU that performs worse than stock in most benchmarks and everyday uses."

Sure there is. The point is to get the highest clock speed at any cost for the sake of... the highest clock speed! I'd imagine that by most benchmarks and in every day use, a Camry would be ranked higher that a drag car that rides like a covered wagon and has no a/c, power windows, a sound system or airbags. Apples and oranges.

"I think this is useless in the real world."

It was never stated that this was meant for real world use. Neither is tightrope walking.

HONESTLY folks, it's not your equipment, not your time and not your money. Lighten up.

 

[bit_user]

"There isn't much of a point in beating OC records if the result is a CPU that performs worse than stock in most benchmarks and everyday uses."

Sure there is. The point is to get the highest clock speed at any cost for the sake of... the highest clock speed! I'd imagine that by most benchmarks and in every day use, a Camry would be ranked higher that a drag car that rides like a covered wagon and has no a/c, power windows, a sound system or airbags. Apples and oranges.

Yeah, most endeavors which move the state of the art usually have trickle-down benefits for the rest of us.

These overclockers are sponsored by Mobo vendors, and it's quite possible that what they learn influences products to make them overclock more easily & stably by less sophisticated users.

The car example is a good one. I'm sure auto racing has lead to many reliability & safety benefits in everyday cars.

"I think this is useless in the real world."

It was never stated that this was meant for real world use. Neither is tightrope walking.

HONESTLY folks, it's not your equipment, not your time and not your money. Lighten up.

Yeah, there's just a lot of cynicism floating about. Also, a fair amount of antipathy specifically towards Intel, some of which they did earn and some is just... haters gonna hate.

 

[bit_user]

Anandtech managed to answer a few questions I had & generally clarify things:

Firstly, the professional overclockers managed 7557 MHz encoding stable, with all four cores (no hyperthreading), on Kaby Lake-X.

For comparison, Kaby Lake-S CPUs manage around 7100 MHz. The absolute world record frequency is around 8800 MHz, held by Andre Yang on an AMD FX-8350 processor, although the KBL-X CPU wins here with IPC.

Secondly, Kaby Lake-X uses a thermal interface paste, rather than an indium-tin solder that we have seen on HEDT processors in the past. This raises a number of points, such that extreme overclockers will have to delid to get the best out of the processor, but also it offers a mixed message from Intel.

Source: http://www.anandtech.com/show/11463/intel-announces-kaby-lakex-processors-highend-desktop-getting-the-latest-microarchitecture

They also note:

This is a noticeable improvement over Der8auer’s previous record set earlier this year, when he overclocked the Core i7-7700K to 7328.3 MHz (in 4C/8T config) on an ASUS Z270 motherboard under liquid helium. Both CPUs were delidded and their thermal interfaces were replaced with a new compound.

Keep in mind that when Der8auer overclocked his Core i7-7700K in January, he did it in a regular room with normal temperature and humidity. The “Performance Matter” event was held outside at 26°C – 27°C with high humidity, which somewhat handicapped the new Core i7-7740X. Without such a handicap, the results could have been better.

Too bad they held it outside, but I guess that leaves room for even further progress.

They also have an impressive shot of a delid/relid job that's filled with paste. The claim being that the PCB actually dissipates heat, so you want better conductivity between it and the heat spreader.

 

[elbert]

Being it broke the 7700k 7Ghz record by 500Mhz mark me impressed.

 

[PaulAlcorn]

Anandtech managed to answer a few questions I had & generally clarify things:

Firstly, the professional overclockers managed 7557 MHz encoding stable, with all four cores (no hyperthreading), on Kaby Lake-X.

For comparison, Kaby Lake-S CPUs manage around 7100 MHz. The absolute world record frequency is around 8800 MHz, held by Andre Yang on an AMD FX-8350 processor, although the KBL-X CPU wins here with IPC.

Secondly, Kaby Lake-X uses a thermal interface paste, rather than an indium-tin solder that we have seen on HEDT processors in the past. This raises a number of points, such that extreme overclockers will have to delid to get the best out of the processor, but also it offers a mixed message from Intel.

Source: http://www.anandtech.com/show/11463/intel-announces-kaby-lakex-processors-highend-desktop-getting-the-latest-microarchitecture

They also note:

This is a noticeable improvement over Der8auer’s previous record set earlier this year, when he overclocked the Core i7-7700K to 7328.3 MHz (in 4C/8T config) on an ASUS Z270 motherboard under liquid helium. Both CPUs were delidded and their thermal interfaces were replaced with a new compound.

Keep in mind that when Der8auer overclocked his Core i7-7700K in January, he did it in a regular room with normal temperature and humidity. The “Performance Matter” event was held outside at 26°C – 27°C with high humidity, which somewhat handicapped the new Core i7-7740X. Without such a handicap, the results could have been better.

Too bad they held it outside, but I guess that leaves room for even further progress.

They also have an impressive shot of a delid/relid job that's filled with paste. The claim being that the PCB actually dissipates heat, so you want better conductivity between it and the heat spreader.

Their coverage was from the ASUS event, this was from Gigabyte's. Though I'm not positive, I dont think they submitted the clocks from the ASUS event to HWBot.

Intel is also using the disabled GPU as dark silicon so they can push a higher TDP. It will be interesting to see how that equates to 'normal' overclocking.

 

[InvalidError]

These overclockers are sponsored by Mobo vendors, and it's quite possible that what they learn influences products to make them overclock more easily & stably by less sophisticated users.

The problem with most research at cryogenic temperatures is that it doesn't apply to room temperature. Most materials behave too differently. If you want to do real research relevant to everyday overclocking, a -40C two stages phase-change cooler would be more useful.

Another problem with using liquid helium to break non-sustainable records is that helium is a non-renewable resource since the Earth's gravitational pull isn't strong enough to stop it from escaping to space.

 

[tank20101955]

the new record holer is the amd fx 8370 8 core chip running at 8722.8 MHz using liquid nitrogen, and the fastest intel is the celeron D 352 running at 8542.7 MHz on liquid nitrogen according to the hwbot.org website's cpu hall of fame.

 

[Malik 722]

it will be a record when they will make processors with standard speed of 7 or 8 GHZ.

 

[bit_user]

Another problem with using liquid helium to break non-sustainable records is that helium is a non-renewable resource since the Earth's gravitational pull isn't strong enough to stop it from escaping to space.

Oh, come on - helium-filled balloons have got to be a much bigger waste of the stuff! Helium-based OC is like the F1 racing of computing. Sure, the cars burn a lot of gas, but there are only a few of them and they aren't driven so much.

Anyway, if it's sufficiently rare & correspondingly priced, then it shouldn't be much trouble for them to capture most of what boils off.

 

[InvalidError]

Anyway, if it's sufficiently rare & correspondingly priced, then it shouldn't be much trouble for them to capture most of what boils off.

Helium has a specific heat of 20J/mol and a density of 0.18g/L at standard conditions. If the LHe-cooled chip dissipates 180W, it'll boil about 18g of LHe per second which will expand to about 100L/s or ~1600GPM if you let it warm up all the way to 0C. Putting helium back into a reasonably sized tank at that rate is going to require a massive multi-stage monster. A closed loop LHe cryo-cooler would be far more efficient as the helium re-enters the compressor at a fraction of the expansion volume thanks to still being at deep cryogenic temperatures. Still going to be a large multi-stage contraption, but a more realistically manageable one.

 

[bit_user]

A closed loop LHe cryo-cooler would be far more efficient as the helium re-enters the compressor at a fraction of the expansion volume thanks to still being at deep cryogenic temperatures. Still going to be a large multi-stage contraption, but a more realistically manageable one.

Closed-loop was my first thought, but I figured you were going to say it would be more expensive than capture.

In any case, I think we can agree that it's good helium-based OC is extremely uncommon.