Does raising FSB speed use more power?

[farina]

This may be a stupid question, but when you're overclocking a Core 2 Duo and you push the FSB...does that in turn increase the chip's power demand? Assuming I'm not manually increasing the Core Voltage.

Thanks.

 

[Starcrosser]

Yes, but not as much as if you changed both the VCore and the FSB speed. Faster the processor operates the more energy it must use.

 

[farina]

Interesting, so for example, a Core 2 Duo E8600 (TDP 65W) would actually use more power than a Core 2 Duo E6400 (TDP 65W)?

Maybe in this case it's not the wattage that's the important part? This question relates to my other question here: http://forums.anandtech.com/showthread.php?t=2159012

Thanks again!

 

[Lonyo]

You can't compare different CPUs, and especially not ones from different families.

Any given single CPU will use more power when the FSB is increased in a somewhat linear relationship with the clockspeed increase, in theory.

Two CPUs with different clockspeeds will not necessarily have such a relationship, because most CPUs run at slightly different voltages etc. The TDP value for a CPU is usually the given maximum power use of that CPU, but every CPU will be slightly different. TDP sets a ceiling.
A 3.33HGz E8600 could use less power than a 2.93GHz E8400 (or w/e the model number is), but it could also use the same or more, if both are rated at 65w TDP.

 

[farina]

So, in the case of my problem, I need to find out the difference between an E6400 (2.13GHz clocked to 3.5GHz) and a E8600 (3.33GHz) processor.

How do I determine if the new chip is killing my PSU...or if it's a software issue?

Thanks

 

[maniac5999]

A higher clockspeed will use more energy. (of course) An interesting question here, however (and I'm no Intel expert, so I don't know) is what if you increase the FSB, but decrease the Multiplier to counter? Assuming the same piece of silicon, what uses more power: 266 X 10 or 133 X 20? Or is it a wash?

 

[farina]

Ya, that's also a good question...need to find someone who knows about this crap 🙂

 

[veri745]

A higher clockspeed will use more energy. (of course) An interesting question here, however (and I'm no Intel expert, so I don't know) is what if you increase the FSB, but decrease the Multiplier to counter? Assuming the same piece of silicon, what uses more power: 266 X 10 or 133 X 20? Or is it a wash?

With the difference in FSB, everything that's running off of the FSB clock would draw more power (including the core PLL), but the core itself would be drawing the same amount of power.

The increase is probably marginal though.

 

[fffblackmage]

Interesting, so for example, a Core 2 Duo E8600 (TDP 65W) would actually use more power than a Core 2 Duo E6400 (TDP 65W)?

Maybe in this case it's not the wattage that's the important part? This question relates to my other question here: http://forums.anandtech.com/showthread.php?t=2159012

Thanks again!

Those two CPUs don't quite make an apple to apple comparison. The E6400 was based on 65nm Conroe, while the E8600 is a 45nm Wolfdale, basically a die-shrunk Conroe. Because those two CPUs have different size transistors, they don't use the same amount of power. This is how the Wolfdales can have a higher clockspeed without increasing power consumption (TDP).

Raising the FSB does increase overall power consumption, since the north bridge will use more power. And if the RAM is sync'ed to the NB, then the RAM will also use more power.