• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Questions on CPU Cooling

B

blade52x

2[H]4U
2FA
Joined
Sep 9, 2006
Messages
3,224
I'm writing up a report for research related to electronic cooling, and I'm looking at addressing some things in my report. It's not something that is being published, so references are not a major issue (I'm having difficulty finding some of this stuff through Google Scholar). Some things I know just from reading [H] articles, Anandtech, TechPowerUp, and etc. But anyway...

1) Are we near the peak on air cooling? It seems like the high end air coolers today aren't much better than the TRUE from a few years ago. Corsair, Antec, and others seem to be addressing this issue now with their easy install water-coolers.

2) I read online on one of these threads that the i7-3930K is 6-core as opposed to 8-core because the full 8-core version has too high a TDP, and two cores were disabled to bring it down. Anyone know if there is validity to this? I sent an email this morning to Intel about this. This question is not meant to bash Intel in any way, it's just more related to the increasing TDP we have seen over the years. (Besides Intel will have 8-cores with the same TDP as the i7-3930k at some point anyway).

3) Moore's Law and TDP : "As the transistor density doubles, the overall design becomes roughly 40% faster, while keeping power consumption the same". If this is true, why are we increasing in TDP? Are we increasing more than 40% every generation (no references to the BD disaster please). Or is this more due to parallel processing?

Thanks.
 
1.We're reaching the peak with heat-pipe solution for CPU cooling.But if they switch to vapor-chamber solution then the boundary of air cooling can still expand more.

For the 2 and 3 question,i really have no solid answer so i'll leave it for others.
 
2. They're either disabled for TDP reasons, or disabled so that Intel can better market their Xeons which will have all 8 cores enabled (or so I've heard). I haven't been keeping up on that area, really.

3. It's hard to apply such a general law to a field as large and varying as CPU design is over small time-frames. We are constantly increasing performance and transistor density while keeping power consumption at reasonable levels, but there's a lot of variables introduced in each new generation.
 
Yeah, if you lower the power consumption by 40% per transistor, but increase the amount of power using transistors by 40%, is there any gain?
 
3) Moore's law has nothing to do with TDP, and it has nothing to do with the relationship between transistor density and performance.
 
ob1 said:
Yeah, if you lower the power consumption by 40% per transistor, but increase the amount of power using transistors by 40%, is there any gain?
Click to expand...

This. Compare the TDP of, say, Prescott chips (single core, 3.4GHz, 115W) with that of Sandy Bridge (quad-core, 3.4GHz, 95W) or even SB-E (6 core, 3.3GHz, 130W) - so I'm now getting basically 6 Prescotts in the same TDP (actually much more if you consider it from a performance standpoint).
 
You must log in or register to reply here.
Back
Top