• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Surfrat's 651 Results!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

TechtonicPC

Member, formerly Surfrat
Joined
Nov 1, 2004
Location
Florida
Well finally she is up and running! Before I go into results I have to give a HUGE thanks to Sent! He drove an hour to bring a chip so I could flash my bios! You da man!!!

Ok for the nitty gritty!

Intel 651
SL94W - Malay

Ok, first problem we had was the temperature. The IHS on this thing is complete garbage. I took a pick that Sent is going to post in another thread, when you see it you will know what i mean! Ok so after 2 hours of experimenting with AS5 and Heatsink placement, we got it to load at 40c idol. Not great, but not bad.

FYI this is on Air Cooling: XP 120 w/ Panaflow 118cfm
MOBO: Asus P5WD2- Premium

Here is where i have ended up at this point:

4.5ghz
1.325 volts

Appears to be 100% stable, but have not had time to Prime it yet. Here are some pics.

PCMark04
4.5ghz run 1.JPG


3DMark03
3dmark03 run 1.JPG


SuperPI
superpi run 1.JPG


Idol temps at this speed are 43c. Load temps are 60c.

So what do we think?
 
Sorry pictures were a little to big, click on them to get a clear image.

Also, I have not had time to really tweak anything to much. Ram timings and what not are stilla t default, so I hope to get some better benchmark scores eventually. Will probably burn in onight and install a fresh copy of windows tomorrow.
 
Nasgul said:
Not trying to be anal about screen shots but CPU-Z 1.31 should be more accurate on the CM readings.

Thank you. :)

4.5ghz on air......nice!


Sorry I used what I had and hadn't even noticed the old version.

4.5ghz cpuz.JPG
 
xTrEmEoVrClOcKr said:
Nice Surf, Sen told me your IHS is totally screwed up ..imagine how well it would clock w/ a nice IHS? :).

Yea it is just pitiful!!!

A picture is worth a thousand words rigth!!

IHSOMG.jpg
 
very nice results. load temps are a bit high, but i'm sure you can work on getting them down, then you can go for some more ;)

you should post that pic of the IHS in here too, just for reference, as it is your IHS, and this is your thread ;)

*edit* you are way ahead of me. that is terrible, i would say a good lapping is in order, but its a brand new chip. I would email that pic to someone at Intel, and see if you get a response.

you know, they spend SOOO much on R&D, you would think making a square piece of aluminum covered copper FLAT wouldn't be too much of a problem. :rolleyes:
 
The IHS are exactly how Intel designed them to be. The last 3 intel cpus I've owned are the exact same as the one pictured above. Notice on the stock cooler, the TIM is in a small square, so when it heats up it spreads out, leaving about a 1/8th inch perimeter around the edge of the heatspreader. Every Intel(socket478,775) heatspreader is almost concave. You cleary covered the entire HS with TIM, and since outer rim of the HS is higher, that leaves that ring, where there is no contact on the center. Put a dab of AS5 in the center of the heatspreader, and then put the heatsink on an that will give you the spread you need. The outer maybe 1/8th" does not have to be in contact with the heatsink at all. All the heat is transfered through the center of the heatspreader. Sent and surf you can argue, but I'm telling you, Intel would not release all these procs from different generations with a defective heatspreader.
 
RangerXLT8 said:
Sent and surf you can argue, but I'm telling you, Intel would not release all these procs from different generations with a defective heatspreader.
I beg to differ, my 640 has an excelent IHS and we were getting 36C idle on his board to flash for CM support (this in stark contrast to the initial 63C boot)
 
RangerXLT8 said:
The IHS are exactly how Intel designed them to be. The last 3 intel cpus I've owned are the exact same as the one pictured above. Notice on the stock cooler, the TIM is in a small square, so when it heats up it spreads out, leaving about a 1/8th inch perimeter around the edge of the heatspreader. Every Intel(socket478,775) heatspreader is almost concave. You cleary covered the entire HS with TIM, and since outer rim of the HS is higher, that leaves that ring, where there is no contact on the center. Put a dab of AS5 in the center of the heatspreader, and then put the heatsink on an that will give you the spread you need. The outer maybe 1/8th" does not have to be in contact with the heatsink at all. All the heat is transfered through the center of the heatspreader. Sent and surf you can argue, but I'm telling you, Intel would not release all these procs from different generations with a defective heatspreader.

Actually I spread it out that time so it was even all the way across, just to show how bad this IHS is. It is obviously hitting the edges and not in the middle. Placing a dab in the middle produced the same temps as spreading it. Now in the end we placed a LARGE dab in the middle. Probably 3 X what you would normally use. This provided the best temps, because the excess AS5 filled in the very large gap.
 
Very nice.

I'd push my 630 more but for whatever reason, anything over 280 on this board gives really weird results. 290 ended up at 251, 285 ended up at 262...
 
For an update. 2 Instances of Prime running for an hour now no problems thus far! I will let it run through the night.
 
Really nice Surf.

You reached a 4.5GHz with only 1.32v which is a 32% overclock.
On par with Ross's 661. Once you get the HSF straightened out 4.8 should be hit with ease. I wish someone had the cajones and cooling to see how CM runs @ 1.6v
 
xTrEmEoVrClOcKr said:
Damn ..4.5ghz stable on a bad IHS ... id guess 4.7ghz would be a snap w/ a lapped IHS / flat IHS.
Yes, shame Intel cant get em right :( (is not suggesting you lap the IHS, do what we discussed Surfrat)
 
hUMANbEATbOX said:
out of curiosity, what did you discuss? secret ihs tactics are intriguing ;)
He and I are banking on limping our gimp chips along for a second binned batch (one that will clock a great deal better). If we are gonna RMA them (which we rightfully can do) Im sure as hell going to use it to my advantage.

That being said for the record his chip did 4.4 on default volts on the XP120 with ridiclous temps.
 
Surfrat said:
Actually I spread it out that time so it was even all the way across, just to show how bad this IHS is. It is obviously hitting the edges and not in the middle. Placing a dab in the middle produced the same temps as spreading it. Now in the end we placed a LARGE dab in the middle. Probably 3 X what you would normally use. This provided the best temps, because the excess AS5 filled in the very large gap.

That large dab, bet is about the same amount of TIM as the stock Intel hSF has on it.
 
Back