AMD A10-6800K APU Overclocked to Over 8.0 GHz

[N.Broekhuijsen]

An overclocker has managed to overclock the AMD A10-6800K to ever-so-slightly past the 8.0 GHz mark.

AMD A10-6800K APU Overclocked to Over 8.0 GHz : Read more

 

[WithoutWeakness]

That's an absurd overclock. I know some people find overclocking like this pointless because it's not attainable for normal users and therefore seems like a waste of time. However, I think it's a great exercise to see just how far an architecture and process can be pushed if you can take the thermal and power limits out of the equation. As technology moves forward and transistors shrink we will continually see more cores and lower power consumption but it's amazing to see that same design be able to go the other direction and offer huge increases in clock speed and performance. It's impressive to see that the APU design that was thought to be destined for low-mid range "gaming" laptops is also scalable to the very high end of consumer desktop use (excluding workstation needs).

 

[bombebomb]

Wonder how the GPU was running on it.

 

[nforce4max]

AMD is achieving what Intel didn't didn't with the Slugtiam 4, give it another year to two years and we will see someone bragging about hitting 10ghz.

I would like to see this benched
Doubt it could be benched beyond 5-6ghz

 

[JPNpower]

These days aren't here to stay. Hurry up and OC on!
More transistors and shrinking architecture inherently creates massive amounts of heat. Maybe water cooling will be the standard cooler boxed with CPUs some day soon!

 

[acadia11]

Been trying to figure out what I was going to do with this liquad nitrogen i had in the freezer for a while now.

 

[acadia11]

liquid.

 

[Andrew Boult]

Cool I heard it almost equaled the i7 4770k at stock. AMD are back

 

[ojas]

Well, after all those "Haswell's reaching 5GHz @ 0.9v" rumours, i refuse to believe another extreme overclocking news piece. :|

 

[groveborn]

"These days aren't here to stay. Hurry up and OC on!
More transistors and shrinking architecture inherently creates massive amounts of heat. Maybe water cooling will be the standard cooler boxed with CPUs some day soon! "

Wrong. Smaller transistors mean they use less power, which inherently creates less heat. Today's CPUs are hitting 5Ghz on air cooling, where if we tried to push a 300 mhz CPU much past its design it would catch fire. The APU in this test was set at 2 volts, but five years ago the CPUs were all above 3.3 volts.

 

[Andrew Boult]

hah yeah you right i Actually bought a 4770k for my birthday and although i've managed to get it down to .09v at stock speeds i had to raise it to 0.95 to make it stable, still impressed though

 

[InvalidError]

Wrong. Smaller transistors mean they use less power, which inherently creates less heat.

Not really.

Yes, smaller transistors use less power (~40% less per shrink) but you have 2-3X as many of them within a given surface area so power per square millimeter when unconstrained by power and heat budgets is still increasing by 20-50% per shrink and that makes smaller processes increasingly difficult to cool down.

 

[nebun]

how stable?....were all programs able to run without crashing???....these are very important variables...overclock is noting without stability.....we all can overclock our machines but at what expense?...something has to give

 

[ahaywood]

@acadia11 "Been trying to figure out what I was going to do with this liquad nitrogen i had in the freezer for a while now."

Given that Nitrogen boils at −195.79 °C, −320.33 °F, it wouldn't really last in your freezer for too long.

 

[Andrew Boult]

very, used prime 95 for and hour and it was fine on 0.95. no longer using that though doing a small overclock on 4Ghz at 1.05v now and it will stay that way till i can get better than the stock cooling on it. i I did have successful overclock on 4.7ghz but the cpu temps scared me into not doing that again

 

[milktea]

Yes, smaller transistors use less power (~40% less per shrink) but you have 2-3X as many of them within a given surface area so power per square millimeter when unconstrained by power and heat budgets is still increasing by 20-50% per shrink and that makes smaller processes increasingly difficult to cool down.

Agree, Intel needs to shrink both the transistor as well as reducing the transistor count, in order to lower the power. But I think they should also do away with the TIM and IHS. The TIM that Intel is using for the IHS is making it difficult to cool the processor. They should have just do away with the IHS and just let the die surface exposed. Or maybe add copper over metal to cover the bare die surface, that should make the cooling much more effective with aftermarket coolers.

 

[JPNpower]

Where do people get this nitrogen from?

 

[InvalidError]

The TIM that Intel is using for the IHS is making it difficult to cool the processor.

The TIM is not the real problem from what I have read.

Someone who de-lidded his Haswell measured the substrate-to-top-of-die and substrate-to-IHS-top heights, subtracted the IHS thickness from the difference and concluded that there is a ~60 micron gap between the IHS and CPU die - the IHS is not making physical contact with the die at all. By replacing the TIM with known high-quality stuff and replicating the gap using paper shims to achieve the same substrate-to-IHS thickness, he found out that Intel's TIM appears to be just as good as enthusiast stuff since the enthusiast stuff actually performed a few degrees worse than Intel's TIM. Removing the gap (shims) is what produced a 20-30C drop in OC temperatures.

Based on that, it seems Intel's TIM would be perfectly fine were it not for the surprisingly large gap between IHS and die.

 

[InvalidError]

Why keep shrinking the transistors?

Benefits of smaller transistors and wires:
- lower gate capacitance and trace resistance/inductance = faster switching and lower power
- less die area to pack a given amount of features = more dies per wafer = cheaper production
- faster switching and shorter wires = faster propagation = ability to cram more logic between D-flops = more work per clock and lower power
- larger transistor budget = more features moved on-chip = cheaper system cost

This is no by means an exhaustive list. Those are merely the more obvious things I can think of.

 

[slomo4sho]

This thing seems to be capable of 5GHz+ on air.

 

[Shin-san]

I never thought I would ever see a CPU running with a 63x multiplier. I also had no idea that the multiplier could go that high

 

[Guest]

I would love to see benchies of said overclock.

 

[nebun]

I would love to see benchies of said overclock.

wishful thinking, lol

 

[JPNpower]

So is it any cheaper to use thermal paste and a poorly constructed IHS compared to whatevertheyusedbefore?

 

[UVB076]

All the people suggesting having fewer transistors in the name of higher clocks should try to learn what makes a CPU perform well... It's not just clock speeds.