Could use some opinions on upgrades

Below are just a couple of shots that will show how the thermal compound spread out. I had a bit too much but it didn't look to be any real concern. Will use a little less when I put it all back together. Cleaned up nice with rubbing alcohol.

dirtysink.jpg


You can see where I filled in the gaps between the copper pipe and aluminum to level it off and then leveled it off with a business card. But you can see there was a little too much and it pushed out on the left, right and bottom in this photo.

dirtycpu.jpg


I got a lousy photo of the CPU when I lifted the heat sink off. Used Photoshop to clean it up a bit but sharpened the pic a bit too much. However, you can see the thermal compound left a pretty good foot print. Too much on the left and a little too much on the lower right. Cleaned up well here too.
 
Okay this is weird. OCZ is out of stock on a 2Gig kit. They want to send me the 1150Mhz now and are willing to send the 4Gig kit instead. Since I'm running XP isn't that a hindrence? I've found that 1.5 to 2G seem to work well anything more seems to be wasted and even slow down XP.

This:

http://www.ocztechnology.com/products/memory/ocz_ddr2_pc2_9200_reaper_hpc_4gb_edition

Also my contact at OCZ says their engineers are testing their RAM with the Gigabyte boards. He spotted them recently.

Thoughts?
 
Okay this is weird. OCZ is out of stock on a 2Gig kit. They want to send me the 1150Mhz now and are willing to send the 4Gig kit instead. Since I'm running XP isn't that a hindrence? I've found that 1.5 to 2G seem to work well anything more seems to be wasted and even slow down XP.

This:

http://www.ocztechnology.com/products/memory/ocz_ddr2_pc2_9200_reaper_hpc_4gb_edition

Also my contact at OCZ says their engineers are testing their RAM with the Gigabyte boards. He spotted them recently.

Thoughts?

Not a hindrance, no. I run Windows XP x64 with 4GB and saw noticeable performance gains from when I upgraded from 2GB. You will only utilize the full 4GB if you have the 64-bit version of the OS, though. The 32-bit (x86) version can only utilize roughly 3-3.5GB.
 
Hmm...ya I'm running the 32 bit version.

Still shouldn't experience any loss in performance. Depending on how memory-intensive the programs you use are, you should probably notice an increase. It doesn't change the memory's clock speed or anything with 4GB on a 32-bit version, it just isn't able to use as much.
 
I have 4GB on one of my 32-bit XP machines... no performance loss from 2GB. Performance loss makes no sense since you simply cannot address above 3.3GB+ of the RAM.

Wow, you went from an 800 2GB kit to a 1066 2GB kit, and now to a 4GB kit?!? jeez. You definitely got your money's worth out of those sticks! Well, aside from all the headaches they've caused.
 
I have 4GB on one of my 32-bit XP machines... no performance loss from 2GB. Performance loss makes no sense since you simply cannot address above 3.3GB+ of the RAM.

Wow, you went from an 800 2GB kit to a 1066 2GB kit, and now to a 4GB kit?!? jeez. You definitely got your money's worth out of those sticks! Well, aside from all the headaches they've caused.

Yep, started with the PC 3200. By the way do you even use a swap file or just use RAM only?
 
Yep, started with the PC 3200. By the way do you even use a swap file or just use RAM only?

Can't answer for Enginurd but I still let the page file on when I had XP 32bit and 4GB to 6GB of RAM. There are some apps out there that don't work correctly without the page file enabled. So I left the thing on just in case.

So happy to see 3GB of RAM usage in Windows 7 64bit :)
 
Do you need to have the CPU installed to test the mb and RAM (to run Memtest)? Some how these feels like a stupid question. :rolleyes:

Off to bed.
 
Do you need to have the CPU installed to test the mb and RAM (to run Memtest)? Some how these feels like a stupid question. :rolleyes
Well, yeah, obviously. You need to start your computer to run memtest, and you can't do that without a CPU installed.
 
Okay...it's test time. Going right to Memtest. These are the 1150Mhz sticks and it auto detects at 800 which is interesting. The others were the 1066 and it detected them at 1066.

In any case I'm running them at 800Mhz with something like 2.4b that apparently matches the 333Mhz bus(?). This sound right? Says 1066 after setting this 2.xb option. Didn't want to stop the Memtest to verify that. I made few changes in the BIOS concerning mem tweaks. CPU was set to x9 for 3Ghz (used optimized defaults). I did set the RAM to 5-6-6-18 per the RAM and 2.2v.

I have not set some of the recommended changes earlier in regards to EIST and some other disable/enable functionality.
 
If I set the SMM to 3.2b that is referring to a 3:2 ratio correct? That sets this RAM to 1066 manually.
 
I've shorted up the list below to only reflect the categories I adjusted. All the rest I left at AUTO but are not posted here (sub categories). This RAM is 1150Mhz, 5-6-6-18 @ 2.2v.

Any recommendations on these settings for stability, performance or etc?

Okay I've adjusted the following for OCing in the BIOS:

MB Intelligent Tweaker(M.I.T.)
Robust Graphics Booster ...............: Auto
CPU Clock Ratio .......................: 9x
Fine CPU Clock Ratio...................: 0.0
CPU Frequency .........................: 4.05 GHz (450X9)

Clock Chip Control
Standard Clock Control
CPU Host Clock Control.................: Enabled
CPU Host Frequency (Mhz) ..............: 450
PCI Express Frequency (Mhz) ...........: 100Mhz
C.I.A.2 ...............................: Disabled

DRAM Performance Control
Performance Enhance....................: Turbo
Extreme Memory Profile (X.M.P.)........: Auto
(G)MCH Frequency Latch.................: Auto
System Memory Multiplier ..............: 2.40B
Memory Frequency (Mhz) ................: 1080
DRAM Timing Selectable ................: Manual

Standard Timing Control
CAS Latency Time.......................: 5
tRCD ..................................: 6
tRP'...................................: 6
tRAS...................................: 18

Motherboard Voltage Control
CPU
Load-Line Calibration..................: Disabled AUTO
CPU Vcore..............................: 1.30v
CPU Termination........................: 1.20V AUTO
CPU PLL................................: 1.50V AUTO
CPU Reference..........................: .76v AUTO

MCH/ICH
MCH Core...............................: 1.2V
MCH Reference..........................: 0.760v AUTO
MCH/DRAM Ref...........................: 0.90 AUTO
ICH I/O................................: 1.500V AUTO
ICH Core...............................: 1.100V AUTO

DRAM
DRAM Voltage...........................: 2.2v
DRAM Termination ......................: .9 AUTO
Channel A Reference....................: .9 AUTO
Channel B Reference ...................: .9 AUTO
 
Use the latest beta version of Real Temp to monitor your temperatures.

Where's the beta at?

Is it normal to see the Gigabyte ET6 temps so different from the 3rd party software?

Also, any thoughts on the BIOS settings just above? Not sure I have these set right.

Thanks!!
 
Keep the load temps of the cpu cores under 60°C and you'll be safe.

Nope. The built-in thermal protection features will make sure that does not happen.

hopefully. lol. does that scare you? boo!
 
Keep the load temps of the cpu cores under 60°C and you'll be safe.



hopefully. lol. does that scare you? boo!

Nice.

Apparently it needs tweaking. Once I started loading some software and I rebooted it rebooted twice. I went into the BIOS and got the small red screen warning about voltages not being correct. Bitch to get back into the BIOS some times. Have to clear CMOS. PITA process to do the OC thing. For some of you it's like crossing the street. Don't have a lot of time to read the details like I'd like to. I thought I could pull it off.

Apparently when loading some software this will test the system in an OCing setup. So the voltage needs to be increased somewhere? CPU or MCH?

As for temps it typically runs at 41C at idle according to OCCT. BIOS says about 32/33C. This is running with the OC mentioned recently.

Great another f'ing reboot loop. I find it interesting this stuff is made to OC and it's been going on for the # of years it has but there's still these stupid hassles like this that come up.
 
Keep the load temps of the cpu cores under 60°C and you'll be safe.

hopefully. lol. does that scare you? boo!
It's not hopeful. I've run my CPU up to 100C before and had the thermal protection kick in and underclock my CPU. The only way the thermal protection would not be functional is if the CPU is defective.
I find it interesting this stuff is made to OC and it's been going on for the # of years it has but there's still these stupid hassles like this that come up.
That's because on an electronic level, getting a CPU to run faster than it's designed to is a lot more complicated than it sounds, and as CPUs get faster and processes get smaller, making a board that can overclock a CPU reliably becomes progressively more difficult. That's also one of the reasons that modern boards have so many fine-grained overclocking options, like GTL voltages and clock skews.
 
It's not hopeful. I've run my CPU up to 100C before and had the thermal protection kick in and underclock my CPU. The only way the thermal protection would not be functional is if the CPU is defective.

That's because on an electronic level, getting a CPU to run faster than it's designed to is a lot more complicated than it sounds, and as CPUs get faster and processes get smaller, making a board that can overclock a CPU reliably becomes progressively more difficult. That's also one of the reasons that modern boards have so many fine-grained overclocking options, like GTL voltages and clock skews.

:mad:

Back to trying to get into the BIOS again. Pulled battery for a while and put on jumper. After having done this twice was finally able to get in and hit Optimized Defaults. Rebooted and am back in the reboot loop again. Holy shit!
 
Does your PC work fine at stock speeds? If so, the PSU is fine.

I've have it set to 3Ghz. Not sure what to set the 1150 RAM to. It AUTO detects as 800. RAM voltage for the RAMS says 2.2v. But had to move it to AUTO as it was the only thing left to change that seemed to be hanging the BIOS. Odd!

Interesting. Checked OCCT v3 and the temp is the same it was at 4Ghz. WTF?

By the way is CPU-Z accurate. I see the CPU speed vary. It was around 2ghz..for a moment was at 3ghz then dropped to 2 again. Other utils show 3ghz. Too many inconsistencies to know what is right and wrong here.

It's back to rebooting again. When I FINALLY get into the BIOS I get that error message about voltage and the hardware. I don't know what the hell to do.
 
CPU-Z is accurate. Intel has something called Speedstep that downclocks your CPU when it's doing anything requiring that much CPU power. CPU-Z is providing realtime CPU speeds hence why you seen the CPU speed vary.

Some questions:
- What does the error message say about the voltage and the hardware?
- How many times does it reboot before you can finally get into the BIOs?
- What does RealTemp report your load temps as?

Next time you reset the CMOS, do NOT use optimized defaults. In fact I think I already warned you already about doing that.

Also, idle temps don't really matter that much. What really matters is your LOAD temperatures which should not exceed 60C to 65C. So use Realtemp to monitor your temperatures as well to see if it agrees with what OCCT is reporting as your load temps are correct.
 
Some questions:
- What does the error message say about the voltage and the hardware?
- How many times does it reboot before you can finally get into the BIOs?
- What does RealTemp report your load temps as?

Next time you reset the CMOS, do NOT use optimized defaults. In fact I think I already warned you already about doing that.

Also, idle temps don't really matter that much. What really matters is your LOAD temperatures which should not exceed 60C to 65C. So use Realtemp to monitor your temperatures as well to see if it agrees with what OCCT is reporting as your load temps are correct.

I have not written the error down. But it refers to a mismatch in voltage with the installed h/w.

How many reboots. If I let it I think it would go on and on. Some times it seems to lock up and I reboot it myself. Yet if I wait sometimes I may be able to get into the BIOS. Other times it looks like I'm going to get into the BIOS and it hangs.

I do not have load temps yet. Just more idle temps. I do believe Real Temp (beta) and OCCT have matched.

At one point I disconnected about 5 USB items and then was able to get going again. Coincidence? Who knows. I pulled everything apart again and have the mb sitting on a box and connected to power, video, RAM, optical drives and the main hdd.

I know it's crazy but it worked for my bro where he installed the cardboard washer between the standoffs and the bottom of the mb and then used another between the screw and the mb and his problems went away. He had wackier stuff going on also. He makes a point that the PSU supplies grounds to everything and the PSU is attached to the case.

Is the PSU faulty? Is the mb actually shorting out? CPU? Sorry to be a pest. I've done this work before but don't do it that often. It seems that this would be pretty straight forward. But I get varying info from sources.

I assume Memtest doesn't create much of a load. Correct?

Is anyone willing to walk through step by step with me on this? I'd truly appreciate it.
 
I know it's crazy but it worked for my bro where he installed the cardboard washer between the standoffs and the bottom of the mb and then used another between the screw and the mb and his problems went away. He had wackier stuff going on also. He makes a point that the PSU supplies grounds to everything and the PSU is attached to the case.
He's wrong. I don't mean to belittle him, but that's simply how it is. The cardboard washers are completely useless and are actually counterproductive.
 
He's wrong. I don't mean to belittle him, but that's simply how it is. The cardboard washers are completely useless and are actually counterproductive.

I don't think you're belittling him but only speaking your opinion. I'm fine with that. But how do you explain all the flaky things he had going on with a P35 Gigabyte mb and then he adds the washer and suddenly it runs solid. Odd.

Yet everyone I've talked to: local pc shop, Gigabyte tech support, this forum ALL say not to bother with the washers.
 
I don't think you're belittling him but only speaking your opinion. I'm fine with that. But how do you explain all the flaky things he had going on with a P35 Gigabyte mb and then he adds the washer and suddenly it runs solid. Odd.
I can't explain that without having had direct contact with it. However, I suspect that the actual problem was something else, and that it was somehow solved indirectly while he was installing the washers. The washers themselves almost certainly had nothing to do with it, unless he had been installing the board incorrectly to begin with and that was causing shorts.
 
- Memtest does not create much of a load
- To determine whether or not the PSU is the problem, you will need to use another PSU. Try to buy one from a local Best Buy or something, use it to test the system, and then return it after you've figured out whether or not the PSU was the problem
- As for the cardboard washer idea, meh. Try it as an absolute LAST RESORT. As for why your bro's problems disappeared, we can't make any guesses since we weren't there when he built his PC. He may have wired up the PC wrong, had a loose screw/wire/cable, etc. More than likely, as people do, your bro may have left out or did not notice certain changes he made to the system when he added those cardboard washers.
- Did you reset the CMOS yet?
 
Back
Top