Huge P78xx Heat sink differences

Hi,
Preliminary results both ICD, 3.2 GHz Orthos small FTT's 8 minutes

OLD = Core 84c, TSZx 92c
New = Core 69c, TSZx 78c

 

Is the new heat sink from the quad core model? That is a nice improvement, wish I got one when I had my x9000.

 

It is the one we have seen on the quad core. Took some pic's I'l upload in a bit too.
my house is rather warm 84F so this is amazing to me, more below..........

Most importantly 3DMark06 at 3.7 used to give 100+c showing in throttle stop with the house cool, now just 87c. I am very happy............

Preliminary results both ICD, 3.4 GHz Orthos small FTT's 8 minutes

OLD = rebooted 3, min 42 seconds in from over 100c TSZx
New = Core 82c, TSZx 90c (ran 89c but one split second at 90c)

 

Those temps are really nice. anyone who is going to get an extreme CPU should really get one of those heat sinks.

 

Here are some pic's I took, click to open full images.........

Edit; Three more observations,

1.) The new HS has much shorter legs. When you tighten it down it has the appearance of placing more force from the heatsink to the CPU die. This could easily be a factor it heat transition.

2.) When converting the files from RAW to JPG I use the 3.7 GHz setting. Before Throttle Stop showed it running in the 80+c range which is still acceptable. Now Throttle Stop was showning 65-68c converting the files. This is more an observation as I don't have the exact numbers nor was I converting the same sequence of file. I can see though it is a huge improvement.

3.) Idle at 3.2 GHz has not really improved. I did not think it would however as I think where TDP starts geting pushed is where it matters. So it idles about 37c. Now at 3.7 GHz I used to idle about 47 to 51c but now it is 42-43C. So it appears as you really push the CPU is where this heatsink starts to shine.........

 

On your pictures is the old heatsink the one with the paste on? It looks like there is less copper on the new one then...

 

Tanware,

Where did you get the heatsink?

 

I want to know too, please share some more information about this!!

 

The heatsink, out of stock right now......

It does appear to be slightly smaller but apparently it is the heat pipe calibration for a higher TDP that counts..............

 

I just took a look at my heatsink and it looks like your new one, legs are short... the part number is 60.4I202.001.

 

You are most likely good then, I know most of the P7805's came with the long legged one's like my first pic up top............

Edit; pic of the new one, sightly different part number, not sure it matters...............

 

I'm pretty sure this is also the HS that comes on the 7815U. Can anyone else verify this?

Z

 

As I said in another thread BB will get manufacturers to use the cheapest parts. I think I have seen other models too with the other HS. Gateway is no slouch in using up various parts that are lying around to get systems assembled........

 

I got a 7811fx. Definitely want one of these heatsinks. Any differences to the graphics chip heatsinks for the 7815U compared to the original 7811fx?

 

This I don't know, Not many have pulled their systems apart and fewer posting the part number and or pictures. I know they have a replacement part and list for the P7815. I noticed it has .003 at the end so most likely it is the most recent revision........

http://www.laptoprepairparts.com/html/details.asp?ref=4477&category=misc

Edit; I would be carefull here. nVidia with the solder issue wants the GPU to run and stay warm. This lowered temp variation lets the card live longer before cold solder issues. This could also limit OC ability too. This can be accomplished by keeping the fluid in the heat pipe at a large volume but lowering the internal vacuum to raise the fluids vaporization point. This is pure speculation but people should be aware as this is what is attempted in the 9c.20.00 and above bios revisions (also reduces fan pulsing)...........

 

Heat pipe calibration? Count me stupid but I read it was the radiator size, matched for rated air flow.

edit

Well, Wiki explains it a bit; "Heat pipes must be tuned to particular cooling conditions. The choice of pipe material, size and coolant all have an effect on the optimal temperatures in which heat pipes work."

Both the pictured pipes looked about the same size so this would leave working fluid or a possible contamination problem?

 

No contamination issues. Both ICD and I tend to keep the system dust free with canned air removing the pannel to clear the HS from both sides.

I however can not discount an aged HS where the pipe has either lost fluid or the vacume to keep it well calibrated. I can tell you it worked fine with the stock P8400 and P9600.

I still run at my default 3.2 GHz but now 3DMark06 usually sees 60c and in the cpu test sections 65c where the CPU fan kicks in. I can run Orthos comfortably at 3.6 GHz and only see about 90c. I can even now run 3DMark 06 at 3.99 GHz but at very high temps to borderline reboot.

3.99 GHz is a bit hairy at 1.4500 vCore so I did it once to test but something that really scares the heck out of me. for some reason with TS I can get 14 but not 14.5 and does do the 15 as the multiplier so I can not seem to test 14.5.

The system even at 3.2 GHz would kick the fan at times under moderate CPU usage, now just the power of the NC2000 pad keeps the CPU cool enough except when I do something to stress out the load levels. Counting on ambient temps I usually idle between 35c to 40c. Idles were between 40c and 43c before. Not really a huge improvement but one just the same.

I really would have loved a Q9200 but other than that I am very happy. To be truthfull I rarely ever come close to being CPU limited now. I guess the grass is always greener on the other side though.