13 replies to this topic

[Vulpesveritas]

9ghz now.... From some thai guy running LN2 of course... but still.

http://valid.canardp....php?id=2385966
http://www.overclock...UB-lt-lt/page67

Just... wow. Even if it is only two cores, this is still rather wow on the clocks. This is nearly a full ghz above the previous champion before the 8150. On another Asus Crosshair V board too.

I'm somewhat horrified at the prospect of how much power that CPU must be eating...

Goes to show that FX chips seem to like overclocking... except for power consumption, and as a result heat output... Hopefully Piledriver will do better at that...

[bawchicawawa]

Vulpesveritas, on 28 May 2012 - 04:43 PM, said:

9ghz now.... From some thai guy running LN2 of course... but still.

http://valid.canardp....php?id=2385966
http://www.overclock...UB-lt-lt/page67

Just... wow. Even if it is only two cores, this is still rather wow on the clocks. This is nearly a full ghz above the previous champion before the 8150. On another Asus Crosshair V board too.

I'm somewhat horrified at the prospect of how much power that CPU must be eating...

Goes to show that FX chips seem to like overclocking... except for power consumption, and as a result heat output... Hopefully Piledriver will do better at that...

Yeah, it would be awesome if PD will get to 10ghz.

Also, the world record was done with liquid helium, i'm surprised that he got 9ghz with ln2.

[cipher]

Limited cores + just booting to get CPU-z to see 9GHz. Not stable at 9GHz, and obviously not practical since you can't keep it cooled like that forever. It might be fun for some, but to me it's just a number.

I'd be more interested in seeing highest overclock that ran stable for one hour of stress testing with something like Prime95. But there's no way to sync the two as proof, so perhaps apps like CPU-z should incorporate stress testing validation?

From that 9GHz overclock...

Edited by cipher, 29 May 2012 - 06:16 AM.

[Vulpesveritas]

cipher, on 29 May 2012 - 06:12 AM, said:

Limited cores + just booting to get CPU-z to see 9GHz. Not stable at 9GHz, and obviously not practical since you can't keep it cooled like that forever. It might be fun for some, but to me it's just a number.

I'd be more interested in seeing highest overclock that ran stable for one hour of stress testing with something like Prime95. But there's no way to sync the two as proof, so perhaps apps like CPU-z should incorporate stress testing validation?

From that 9GHz overclock...

Believe me, I would like it too, sadly there is more of a world record "wow" value for simply the bigger number.

[Zakatak]

Huh.

Graphene-based processors are expected to be 1THz by 2020 or so.

Edited by Zakatak, 30 May 2012 - 05:06 PM.

[Vulpesveritas]

Zakatak, on 30 May 2012 - 05:06 PM, said:

Huh.

Graphene-based processors are expected to be 1THz by 2020 or so.

Maybe, maybe not. Would be cool, but then again, so would flying cars.

mainly seeing as IBM has only started working with single nanometer carbon nanotubes and we're a bit away from single atom t thick transistors in my knowledge.

[Barbaric Soul]

2cores/2threads,nice DUAL core you have there, pffffft. That OC doesn't matter to me. OC that high with all cores/threads the CPU has inabled and you'll get my attention.

I just hate it when Intel or AMD, or anyone else for that matter, brags about a high arse OC, when they had to disable features of the CPU to obtain the OC.

Edited by Barbaric Soul, 31 May 2012 - 04:19 AM.

[Hruhai]

Pfft, I could do that in my sleep.... :'(

[Adridos]

That should handle at least ten million units in a strategy game, right?

[Aethon]

Vulpesveritas, on 30 May 2012 - 06:38 PM, said:

Maybe, maybe not. Would be cool, but then again, so would flying cars.

mainly seeing as IBM has only started working with single nanometer carbon nanotubes and we're a bit away from single atom t thick transistors in my knowledge.

People scare me enough with rolling cars...I hope flying cars never happen, lol.

[Skrapheap]

Vulpesveritas, on 30 May 2012 - 06:38 PM, said:

Maybe, maybe not. Would be cool, but then again, so would flying cars.

mainly seeing as IBM has only started working with single nanometer carbon nanotubes and we're a bit away from single atom t thick transistors in my knowledge.

IBM has already stated that graphene will not replace silicon as the primary material in CPUs, but rather act as various components in hybrid processors. Now the material that is gaining ground as a potential silicon replacement is molybdenite. However 2020 is still 8 years off, and a lot can happen in that time period.

[Vulpesveritas]

Wulffemein, on 31 May 2012 - 08:54 PM, said:

IBM has already stated that graphene will not replace silicon as the primary material in CPUs, but rather act as various components in hybrid processors. Now the material that is gaining ground as a potential silicon replacement is molybdenite. However 2020 is still 8 years off, and a lot can happen in that time period.

yet at the same time IBM has stated that carbon nanotube transistors show to be vastly superior to silicon transistors in sub-6nm processes....
Note that carbon nanotubes != graphene... technically.
pure graphene is a single atom thick molecule of carbon, usually in a flat sheet, and can either refer to the material, or just a flat structure of it.
carbon nanotubes are a tubular structure of graphene

The structure probably matters in they're effectiveness. But yes, 8 years can change a lot of things in the world of processors.
Also... IBM only said that single layer graphene won't... they never said anything about bi-layer graphene...

Edited by Vulpesveritas, 31 May 2012 - 09:20 PM.

[Chiyeko Kuramochi]

bi layer graphene would actually reduce it's features the whole special thing is all the things it can do once you got a single layer.

[Vulpesveritas]

Chiyeko Kuramochi, on 01 June 2012 - 01:01 AM, said:

bi layer graphene would actually reduce it's features the whole special thing is all the things it can do once you got a single layer.

The problem is though that single layer graphene isn't able to actually act as a semiconductor. Bi-layer graphene however, has already been shown to be a semiconductor, and an electrically tunable one at that, so not only is it able to be tuned to have a high on/off difference, but is also a possible good choice for LEDs.
Though single layer graphene should be good for interconnects given its high electrical capacity.