If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
8IK1100 Rev 2 CPU Temps/Voltage Problems
Hello all, hoping for some help.
I just upgraded my 848 based Intel board to the Gigabye 8IK1100, Rev 2.0. My hardware is as follows: 8IK1100 Rev 2, Bios FG (latest) 4x512mb Infineon Cas 2.5 DDR 400 (dual) 120gb Seagate IDE (Baracuda V) 400W Fortron/Source Power supply P4 2.8 Northwood (800 bus) Sony 4x DVDRW Sony 16x DVDROM Nvidia 5200 Ultra Video Card Zalman 7000 heatsink/fan My issue is that the 2.8 has been running very very hot. I've made sure that the CPU heatsink/fan is securely fastened and working fine. This same cpu and memory were running in the on the 848 board in the same case at 40 deg cel but after changing the motherboard I am seeing 50 deg + cel (up to 60 under load). The box also isn't as stable as it was... The ram, board, bus, etc timings are all 100% stock (ie. no overclocking) and CIA is off. After much investigation I found that all utilities in windows (ie sandra) show the CPU getting way too much voltage -- like 1.6 instead of 1.525. I lowered the cpu voltage setting in the bios about 4 or 5 notches below "default" and now I'm seeing the correct voltage in Windows (1.525) and lower temps that are more within reason (45-50 deg cel at full load, 35 idle) Has anyone else seen this on the 8ik1100? it's very odd!! I had first tried a 3.2 gig Prescott on this board and could not get it stable and cool....I thought it was because of the Prescott's new design (well known for major heat)...now I wish I hadn't sold it because this might have been the root cause. Appreciate ya'alls insight. Z |
#3
|
|||
|
|||
|
#4
|
|||
|
|||
I'm no guru -- but here's my 2cents
Depending on the ram you are running, it might actually be very good for overclockability if you raise the memory voltage to +2 or +3. Geil and Mushkin and the like actually recommend it and test their ram at these levels for durability/stability. You might try setting your memory divisor to 1.66 instead of 2.0 in the bios. Then, bring the multiplier up to around 250 to make the memory run at about 400 again (up from 333). There's an extra cycle and bandwidth hit from the 1.66 divisor, but you should be able to push the system that much faster to offset this. It should at that point let you engage PAT....and you get a killer overclock and can still go faster. The nice thing about this is that even with the memory still running at native 400mhz (due to the 1.66 divisor), you can get a really stable overclock - my 2.8 is at 3.4 at this point. If i creep the multiplier higher I can get to nearly 3.7ghz -- almost a gig overclock and still be stable with a bit of overvoltage on cpu and mem. 242 for dual ddr 400 ram is really fast if you're still at 1:1 multiplier, btw. Both sets of ram I've tried (crucial and geil basic) aren't really stable much past 230 or so at 1:1. BTW: played some more this weekend with the new geil basic ram. With the timing at 2.5/3/3/6, PAT on, 1.66 multiplier I am able to run 3.65ghz dead stable with the memory voltage +2 and the cpu voltage *** down two settings **. 100% utilization yields about 60deg cel, which isn't too terrible. Zac ocnoob wrote in message . .. On 15 Apr 2004 21:38:00 -0700, (Zac) wrote: Well, every bone in my body tells me this board runs the voltage about 5 notches higher then it says in the bios....what out if you're running this one kids. However, least it's stable Right now I'm running the setup above at 3.4ghz, 5:4 FSB/DRAM, 2.5,2,2,8 bus 405mhz dual ddr and it's still 1 notch under the default CPU voltage in the bios (1.5125 I think)....I didn't up AGP/Memory either....and this is a dead stable config. FWIW, I'm not cluey about this stuff, but I do have my P2.6c running at 3.14 (ie, bus at 242) on this board - but maybe not so stable ;-) At lower voltages, Prime 95 torture test won't make it through first test without failing on hardware rounding error. With higher voltage (one notch above normal), I get to second test before it hits a rounding error. So probably only concern for daily use when I do heavy stuff for long periods (I think it crashed once before while I was running Tmpgenc). Idle temp (stock cooler) is 39 - not sure I want to set voltage any higher to try for more stable under stress (temp was 63 when Prime 95 failed). I set DIMM voltage at +0.1v cause someone said that would help with OC stability too? EasyTune reports voltage as 1.61, CPU-Z says 1.584, and bios setting is 1.5375 - huh? Anyway, I was running like this for a couple of months (242 with CPU and DIMM voltage at one notch above normal based on someone else recommendation) - but then board died and had to get it replaced under warranty. Hope that's not a result of overclocking - time will tell ;-) Not a bad overclock for an hour's work on a board everyone says can't clock and some cheap $179 unmatched crucial 2.5/3/3/8 memory !! Now If I can get it stable with 'performance mode' active on the dram settings....Alogether runs about 14% faster through Sandra then stock. CPUZ says memory "performance mode" disabled - I'm using DDR400 - can't get it to performance mode enabled with bus well over 200 can I? Zac (Zac) wrote in message . com... Hello all, hoping for some help. I just upgraded my 848 based Intel board to the Gigabye 8IK1100, Rev 2.0. My hardware is as follows: 8IK1100 Rev 2, Bios FG (latest) 4x512mb Infineon Cas 2.5 DDR 400 (dual) 120gb Seagate IDE (Baracuda V) 400W Fortron/Source Power supply P4 2.8 Norhwood (800 bus) Sony 4x DVDRW Sony 16x DVDROM Nvidia 5200 Ultra Video Card Zalman 7000 heatsink/fan My issue is that the 2.8 has been running very very hot. I've made sure that the CPU heatsink/fan is securely fastened and working fine. This same cpu and memory were running in the on the 848 board in the same case at 40 deg cel but after changing the motherboard I am seeing 50 deg + cel (up to 60 under load). The box also isn't as stable as it was... The ram, board, bus, etc timings are all 100% stock (ie. no overclocking) and CIA is off. After much investigation I found that all utilities in windows (ie sandra) show the CPU getting way too much voltage -- like 1.6 instead of 1.525. I lowered the cpu voltage setting in the bios about 4 or 5 notches below "default" and now I'm seeing the correct voltage in Windows (1.525) and lower temps that are more within reason (45-50 deg cel at full load, 35 idle) Has anyone else seen this on the 8ik1100? it's very odd!! I had first tried a 3.2 gig Prescott on this board and could not get it stable and cool....I thought it was because of the Prescott's new design (well known for major heat)...now I wish I hadn't sold it because this might have been the root cause. Appreciate ya'alls insight. Z |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
PSU problems with motherboard | BillyR | Homebuilt PC's | 6 | September 20th 04 12:40 AM |
Sapphire Radeon 9800SE + Shuttle machine = Problems | Jiri Kuukasjärvi | General | 2 | September 12th 04 11:17 AM |
Looking to solve problems before they happen | Will Carter | Homebuilt PC's | 5 | September 6th 04 04:13 AM |
Random reboot problems | Cory | Homebuilt PC's | 12 | August 21st 04 03:12 AM |
Major Computer Problems | Toronto Garage Door Company | General | 20 | November 13th 03 09:41 PM |