Does the motherboard affect the extent of overclocking performance?

[consptheory77]

It will always my impression that overclocking performance was solely a matter of the "silicon lottery", that some chips could stretch further in GHz than others. But I just read a review of the Gigabyte Z87-D3HP in the UK PC magazine PC Format, October 2012, and it says the following:

"The real bummer is that the Z87-D3HP does that old trick of booting at that higher speed but then throttling down much lower as soon as you give it something to ponder. The same goes for hand-tuned overclocking. It boots successfully at 4.7 GHz, which is up there with the best, but give it some multi-threaded number crunching to think about and the clocks drop to 4GHz."

 

[TJ Hooker]

Yes, the motherboard plays a part in how well your CPU will overclock.

 

[Onefurrybanana]

Yes for example the pc mate motherboards are very basic and won't overclock well where as the asus hero VII has extra power delivery features so you can give your cpu more power allowing a better overclock also better heatsinks on the vrms.

 

[filippi]

That is usually why someone will spend twice as much money in a top ASUS ROG motherboard.

 

[geofelt]

The motherboard does play a part, but unless you are seeking maximum overclocks, not so much.
Every motherboard comparison review that I have read shows minimal differences in performance across the board.
That said, you are likely to do better with a Z97 based motherboard.

If you are looking at haswell, I think the silicon lottery is the chief factor.
Moreover, it is not the cooling that limits the overclock, but your tolerance for high levels of Vcore.

If you are looking at a new build, think skylake.

 

[consptheory77]

The motherboard does play a part, but unless you are seeking maximum overclocks, not so much.
Every motherboard comparison review that I have read shows minimal differences in performance across the board.
That said, you are likely to do better with a Z97 based motherboard.

If you are looking at haswell, I think the silicon lottery is the chief factor.
Moreover, it is not the cooling that limits the overclock, but your tolerance for high levels of Vcore.

If you are looking at a new build, think skylake.

The board I have for my build is a GA-Z97-HD3, the only reason I bought it was because it was actually the cheapest motherboard I could get at the time, I got it for an even US $50. I put a G3258 in it, and overclocked it by a GHz. I don't plan on changing my motherboard anytime soon, but I am planning to replace the G3258 with the i5-4690K, so after reading that review, I wondered if there will be limitations imposed by my board.

 

[geofelt]

Not to worry.
4690K is a big upgrade, even at stock.
When you OC, just keep the Vcore under 1.30v.

I would imagine that something in the 4.0-4.4 range would be the minimum you can do.

Remember, the G3258 is also haswell, so you should be able to do at least as well with an oc as you are now doing with the G3258.

 

[consptheory77]

Not to worry.
4690K is a big upgrade, even at stock.
When you OC, just keep the Vcore under 1.30v.

I would imagine that something in the 4.0-4.4 range would be the minimum you can do.

Remember, the G3258 is also haswell, so you should be able to do at least as well with an oc as you are now doing with the G3258.

I got up to 4.2 GHz on the G3258 by going up to 1.254V, so if I can get the 4690K up to 4.5 GHz with the same voltage, I guess that would be an excellent result. But I am curious as to what dynamic within the motherboard itself makes this possible, or limits it, such as in the case of the Z87 boards mentioned, or if it's just the difference between the Haswell architecture and the Ivy Bridge architecture.

 

[TJ Hooker]

I imagine the biggest factor on the mobo would be the VRM circuitry, in terms of quality, number of phases, heatsinks, etc. I don't know why Z97 would tend to perform better that Z87 in this regard.