E8500 overclock result..so far !!

[ctal]

Originally Posted by hansje

Only CPU 4.2 and memory at default 61FPS at the start of a race.
CPU 4.0 and memory like above 72FPS..so eleven frames more.

I thought that with cpu 4.2 gHz and vcore 1.43V (or 1.41V per post 1 above) you were able to run the memory at 442 mHz (= ddr2-884), at 4-4-4-10-1T. What are the "default" memory settings that are resulting in 61 fps? CPU-Z is one way of determining that. My suspicion is that the "default" settings are running your memory unnecessarily slowly.

Regards,
-- Al

[ctal]

Of course, you could always throw in a phase change cooler and get 5.2 gHz, or a liquid nitrogen cooler, and get 5.3 GHz!!!!

MADSHRIMPS - Hardware Reviews ,Crazy Projects, Modding Tutorials and Overclocking

-- Al

[TQ]

Originally Posted by ctal

Of course, you could always throw in a phase change cooler

Didn't Dr. Emmett Brown have one of those installed in the DeLorean?

"Wait a minute, Doc. Ah... Are you telling me that you built a time machine... out of a DeLorean?

[ctal]

Originally Posted by TQ

Didn't Dr. Emmett Brown have one of those installed in the DeLorean?

"Wait a minute, Doc. Ah... Are you telling me that you built a time machine... out of a DeLorean?

Sounds right!

My wife and I have seen reruns of that on cable numerous times in the last few years, but I haven't paid serious attention to the alleged technology.

Completely OT (apologies to Hansje) -- Since my youth I've felt that one of the all-time (pun intended) great works of literature that also happens to be science fiction was "The Time Machine" by H. G. Wells. It was made into a pretty good movie in 1960, with Rod Taylor and Yvette Mimieux, and I think it was re-done more recently (don't know how well).

Regards,
-- Al

[TQ]

Mmmmmm....Yvette Mimieux.

(Actually it was a flux capacitor...I was just being...you know, me.)

[Rickwell]

flux capacitor ?Asus made that didn't they?

[hansje]

Woow, just out of bed and already a lot of reading to do.

Thanks for the replies...soon back with answers I hope

Hans

[hansje]

Results:

CPU 4.0 and memory 444.5 with 4-4-4-10-1T >> 72 FPS at the start.

CPU 4.0 and memory 333.4 (default) also 4-4-4-10-1T >> 63 FPS...with CPU 4.2, same settings, also 63 FPS !!
So to overclock memory seems very important to gain some FPS !!

I like to use CPU 4.0...because I need less Vcore only 1.328 and the TEMPS stays MAX at 58C after stress-testing.

Thanks for the links TQ


Hans

[TQ]

Originally Posted by hansje

Thanks for the links TQ

You're welcome, sir.

[ctal]

Originally Posted by hansje

Results:

CPU 4.0 and memory 444.5 with 4-4-4-10-1T >> 72 FPS at the start.

CPU 4.0 and memory 333.4 (default) also 4-4-4-10-1T >> 63 FPS...with CPU 4.2, same settings, also 63 FPS !!
So to overclock memory seems very important to gain some FPS !!

I like to use CPU 4.0...because I need less Vcore only 1.328 and the TEMPS stays MAX at 58C after stress-testing.

That all seems very expectable. A 33% increase in memory speed (444/333), with the same latencies, produces a 14% (72/63) increase in fps.

I think you are doing the right thing running at 4.0 gHz; as I said earlier I would allow some margin relative to the 1.45V absolute maximum rating for Vcore, and your margin at 4.2 gHz is very small.

Regards,
-- Al

[hansje]

Indeed not bad, only try to get it stable...Memtest one hour no problem.

Using Prime blend/test, after 40 min. 1 error !!!!

Only CPU -test Orthos, few hours no probem ???....looks like RAM problem ??

Raised the Mem/v to 2.275 V now, and Vcore 1.34 instead of 1.33

TEMPS are great I think >> MAX 58C
Hans

[hansje]

Strange can't give my FSB Volt. >> 1.5V BIOS...system won't start at all.

1.4 working fine ????

When I raise my FSB volt, sometimes my Vcore goes one higher by ityself at the same time !!??

Hans

[ctal]

Modern integrated circuits do not like to be exposed to voltages on their signal pins that are higher than the power supply voltage to them. I would not set the fsb voltage to any higher than Vcore. There are resistors between the fsb supply voltage and output pins of the cpu, but the current flowing through them is very small when the output is in a high (logic 1) state. The very small current results in a very small voltage drop across the resistors; therefore under that condition (logic 1) the cpu output would see the fsb voltage pretty directly. If that is higher than Vcore, it could be unhealthy.

Regards,
-- Al

[hansje]

Interesting...thanks, just the info I need

Hans.

[hansje]

Two hours and 10 min. PRIME blend/test so far ...before only 40 min !!

Had to raise Vcore 1.35 (CPU at 4.0) and mem. 4-4-4-12-1T with 2.275 V
MAX temp >> 62C but taken in consideration a hot room (think 20C) ...its not bad !!!!

Hope to get some hours without errors

Keep you guys informed

Hans.