| Thread Tools |
19th December 2004, 17:22 | #1 |
Posts: n/a
| Bad winchester?? Hi, I'm trying to OC my new setup. (MSI K8N Neo2 Platinum +3200 winchester, Memory is corsair xms3200 2-2-2-5, PSU is Enermax 485 CoolerGiant) I'm getting stuck at 2080 Prime stable, 2100 workable under windows. These are my test settings: Multiplier ....... x9 HT mult ........ x3 HTT 231 Mem set for 133 (is at 148 when max OC) V core 1.6V High perf ......disable Agr timing .....disable Dyn OC disabled C&Q disabled Vdim 2.75V Memory timings DRAM speed 133 2T CAS 2.5 Ras to Cas 4 Min Ras 6 Row precharge 4 I have been up to HT 260 with a x8 Multiplier setting. Also cannot get this setup Prime stable over 2020 Mhz on stock Vcore.... I must be doing something seriously wrong. (BTW the Vcore is measured with a DVM) Any help or suggestions is greatly appreciated. Yama |
19th December 2004, 17:33 | #2 |
Member Join Date: Mar 2003
Posts: 1,823
| Sure you're HTT multi is @ 3x , that is the thing that was holding me back.
__________________ |^^^^^^^^^^^^^^^^^^\||____ |....The BEER TRUCK.... ||||'""|""°\_,__ |___Just for You___l||__|__|___|) _____| |(@)(@)""""""""""""**|(@)(@)****|(@)| |
19th December 2004, 19:42 | #3 |
Posts: n/a
| Yep, it's on 3x just doublechecked.. |
21st December 2004, 20:51 | #4 |
Posts: n/a
| many people report problems with prime95 + winny... it doesn't mean your oc is unstable. test with some other benchmark/stress test programs |
21st December 2004, 20:55 | #5 |
Madshrimp Join Date: May 2002 Location: 7090/Belgium
Posts: 79,021
| rather a bug in P95
__________________ |
22nd December 2004, 01:15 | #6 |
Posts: n/a
| It seems no one is having much trouble OC'ing this combo. I checked the core voltage, 1.41V readout on DVM, checked the HTT multiplier using the AMD CPU info tool (shows 800 when at x4), have memory at 133MHz (4/6) at relaxed timing. CPU divider on x9. Disabled agressive timing and CnQ. Disabled all SATA ports. No overclock selected in CELL menu. Not using any PCI cards. Memory is Corsair XMS XL 3200 2-2-2-5. Power supply is Enermax Coolergiant 485W with a total of 31A divided over 2 separate 12V rails. I use clockgen to set and check clock, it confirms the CPU divider (CPU running at 1800) when cpu mup is set for 9. I can get up to about 2100 (HT= 230) then system crash in to blue screen. If I set the CPU multiplier to x10, I hit the blue screen again at about 2100MHz. (HT=210) Over 2020 I cannot get Prime 95 to complete more than 2 passes, except when I boost the CPU core to 1.6V. Then I get a little bit more, maybe 50MHz cpu clock. The system runs ok at standard settings(CPUx10,HT x5, Mem 200MHZ), even with agressive timing enabled and the memory at 1T 2-2-2-5. Another thing that is happening is that after a shutdown from the windows start menu, the system switches on the next time, and I sometimes get 3 messages telling me that windows has recovered from a serious error......send error report, etc. I don't know where to look for the source of this/these problems. Just did a check at HT x3, restart at HT=236 and 2130 MHz, So I don't think it's the HTT overloaded. Even though I thought prime95 was fixed, I'm now just using blue screens as reference. |
23rd December 2004, 02:00 | #7 |
Member Join Date: Jan 2004
Posts: 923
| I read that many people have problems with winchesters memory controller.
__________________ AMD Opteron 146@3ghz H2O Storm | DFI NF4 Ultra-D | 2*512MB Gskill LE | X800GTO˛@X850XT-PE | 2*250GB Hitachi | Modded Aopen H600B| Seasonic 600W | MX510@450HZ | 19\" TFT BenQ 937s+ |3DMARK2001 AMD XP2400+@2600+ | Epox 8rda+ | 2*256 Twinmos BH5 | Club3D 9600Pro | Maxtor 80GB + 200GB | server + moviestation | |
30th December 2004, 23:26 | #8 |
Posts: n/a
| No it's not the memory, nor the controller. It dies at around 2100Mhz, and is only dependent on the Vcore. Memory timings and speed and vdimm seem to be totally irrelevant. Tried it all, mem at 100 mhz, 166, 200. timing 2T 3-3-4-7. Everytime the same barrier. Also when I drop the CPU multiplier to something like 8, and set the HTT to 3x, it still craps out at 2100 at 1.7V (real 1.69V) vCore. Just a bad CPU, I think. Maybe AMD is screening better now. Thanks all for the help. |
30th December 2004, 23:29 | #9 |
Madshrimp Join Date: May 2002 Location: 7090/Belgium
Posts: 79,021
| very weird, might have been an early week Winchester, because all the recent ones are hitting 2500-2700 on stock AMD cooling
__________________ |
30th December 2004, 23:49 | #10 |
Member Join Date: Jun 2004 Location: mechelen(belgium)
Posts: 2,037
| i hope mine does srry to hear that Yama77
__________________ watercooling, chilled water,direct phase change, cascade, custom modding |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Battlefield: Bad Company 2 GPU Performance In-depth | jmke | WebNews | 1 | 8th March 2010 20:19 |
Bad Company 2 at 7680 x 1600 pixels | jmke | WebNews | 0 | 2nd February 2010 10:43 |
TSMC 40nm process is in bad shape | jmke | WebNews | 0 | 26th May 2009 10:08 |
What a Bad Hard Drive Crash Looks Like | jmke | WebNews | 0 | 17th March 2009 11:23 |
Intel Joins the Overclockers – Bad Axe 2 Motherboard | jmke | WebNews | 4 | 2nd December 2006 18:21 |
QX6700 Kentsfield w/ Bad Axe 2 Hits Over 3.7GHz On Reference Cooler | jmke | WebNews | 0 | 29th September 2006 21:11 |
PCs plagued by bad capacitors | Sidney | WebNews | 2 | 11th November 2005 16:01 |
Bad Boys, Bad Boys, Whatcha Gonna Do? SWAT 4 Reviewed | Sidney | WebNews | 0 | 9th May 2005 19:15 |
Bad OC mobo? | kristos | Hardware/Software Problems, Bugs | 18 | 14th September 2004 16:02 |
How to fix XP when it goes bad | jmke | WebNews | 0 | 13th September 2004 10:52 |
Thread Tools | |
| |