Core i9-13900K at 9GHz
-
ASUS has managed to overclock the i9-13900K to 9GHz. During the test, the processor was submerged in liquid helium at -250ºC.
-
What a barbarity, to beat that you would have to aspirate the helium with a large vacuum pump, or better yet you put it all in an ultra freezing industrial chamber.
What I'm telling you, in a distant future in which helium is being extracted for fuel on the moon, records will be broken there, but on the far side, oh boy, on the other side it's still hot. -
And I remember when people used to mount peltir sheets without thinking about the price of electricity....
Regards -
Take that! No one beats this anymore. How long did the record of the celeron cedar mill at 8Ghz last? And it's been a while since then. Mention apart from this spectacular article from Anandtech from the year 2000 where they said that by 2005 netburst would reach 10GHz at 1V: https://www.anandtech.com/show/680/6
I wonder how many kW the TDP of the cpu in this test is. -
@palotes said in Core i9-13900K at 9GHz:
Mention aside this spectacular article from Anandtech from the year 2000 where they said that by 2005 netburst would reach 10GHz at 1V:
Well if they said that thinking about the computing capacity that would be needed.... divide it among the cores that were used around 2005 and the same thing wasn't that far off.
@palotes said in Core i9-13900K at 9GHz:
I wonder how many kW the CPU's TDP is in this test.
You meant W not? it won't reach even one kilowatt.
-
What a monstrosity!
Cluster MPI de tres nodos formado por tres SERVIDORRES HP PROLIANT DL360P 1U 2X E5-2680V2 CPU'S, 20 cores, 40 hilos a una frecuencia base de 2,8GHz. 192 GB de memoria RAM DDR3 ECC (8 x módulos de16 GB 800Mhz)´+ 1U 2X E5.-2697 V2 a una frecuencia base de 2,8 Ghz 192 GB de Ram ECC (8 módulos de 16 Gb 800 mhz) 24 Cores 48 hilos ( 44c/88T)+ 1u 2xE5 2603v2 con una frecuencia base de 1,9ghz 64 Gb DDR3 ecc (4x módulos de 16)
-
@defaultuser I'm afraid the article is crystal clear about the prediction mess. They start by saying that the p6 architecture began at 150MHz in 1995 and that 5 years later it had reached 1GHz. The mindset of people in 2000 is the same as in the 90s, that everything is solved by increasing the frequency. And the filthy crap of netburst is good proof that at Intel they still thought like that. In the article they even talk about physical issues to reach 10GHz like that a 70nm node will be necessary.
The mindset has changed a lot and that's why one might think they were referring to a kind of equivalence like the Palominos nomenclature in the AthlonXp where the equivalent performance of a P4 was expressed but at a lower frequency. This was in 2001. I think Intel changed the paradigm with the Yonah in 2003 but it wasn't until the Conroe and the Core 2 Duo that GHz definitively stopped being the main means to increase performance.
We are talking about an article from 2000, before the Conroe, the Yonah and the Palomino. When they said 10GHz they were literally referring to that frequency.
About the kW it was just an exaggeration.
-
@palotes I hadn't read it all.
Yes, back then there was a lot of speculation and dreaming, but well, those were the beginnings, now a lot of water has flowed under the bridge of IT and everything but everything has changed a lot.
Look, a Palomino was my first micro, and the truth is that now that I think about it, we all thought that micros would come with more and more transistors and higher frequencies, and chinpun.
Now that I look back, it did lag, yes, waiting for silicon at 10Ghz in 70nm was not. -
@portada-hl @Xevipiu new need unlocked.
