Intel Increases Overclocking Potential in Haswell Processors

[Ajay]

Agreed, BLCK overclocking replaced FSB overclocking. Now Intel is back to BLCK overclocking like the Nehlem days.

But wait, it's Revolutionary (TM)! :whiste:

 

[Atreidin]

This much of your post as been proven false by a member here (IDC) who replaced the Intel stock TIM with AS Diamond with no change in temps.

Are you talking about the thread where he got a -20 deg C change in temps when overclocked? He used NT-H1 in that. Are you referring to something else? I could see how when not overclocked the difference would be lower but I think he showed pretty definitively that Intel could have picked much more thermally conductive paste (although describing their choice as low-quality is probably going too far).

 

[IntelEnthusiast]

But wait, it's Revolutionary (TM)! :whiste:

Lets wait a bit and see what our next generation of processors before we get carried away with talking about what they may or may not do.

 

[Smoblikat]

Agreed, BLCK overclocking replaced FSB overclocking. Now Intel is back to BLCK overclocking like the Nehlem days.

If I remember correctly BCLK on nehelm wasnt at fixed integers. You could raise it and lower it just like an FSB, If haswell is like SB-E then we have three choices of BCLK and little to no deviation.

Lets wait a bit and see what our next generation of processors before we get carried away with talking about what they may or may not do.

I dont suppose you know if it will behave more like a variable FSB or a fixed integer BCLK (SB, SB-E) that can be switched to a few different selections.

 

[Ajay]

Lets wait a bit and see what our next generation of processors before we get carried away with talking about what they may or may not do.

That was just a joke, based on some website byline - hence the whistle emoticon. Evidently, I failed in my attempt at humor.

 

[WhoBeDaPlaya]

If I remember correctly BCLK on nehelm wasnt at fixed integers. You could raise it and lower it just like an FSB, If haswell is like SB-E then we have three choices of BCLK and little to no deviation.

Yup, it was completely variable. The single-bin turbo was kind of pathetic though

 

[Concillian]

The real takeaway is that locked chips may have some amount of OC tweaking like the 3820 does on 2011.

It's a pretty big thing if you can have SOME level of OCing on an Intel CPU under $200.

 

[Phynaz]

Are you talking about the thread where he got a -20 deg C change in temps when overclocked? He used NT-H1 in that. Are you referring to something else? I could see how when not overclocked the difference would be lower but I think he showed pretty definitively that Intel could have picked much more thermally conductive paste (although describing their choice as low-quality is probably going too far).

You misread the thread. Results are independent of the TIM used.

Here's his results http://forums.anandtech.com/showpost.php?p=34001126&postcount=496