PCDVD數位科技討論區
PCDVD數位科技討論區   註冊 常見問題 標記討論區為已讀

回到   PCDVD數位科技討論區 > 電腦硬體討論群組 > 系統組件
帳戶
密碼
 

  回應
 
主題工具
quizman
Major Member
 

加入日期: Feb 2001
您的住址: 台北市
文章: 263
第三段

/* 7th */
ok, i think i got to max that is possible with my hardware.
specs:

unlocked XP2200+ (shorted VSS and NC pins)
xp333-r rev 2.1 w/4/11 BIOS
Geforce4 MX 440 64MB
Thermaltake 6cu+ heatsink + fan, lapped, arctic silver 3
OCZ 256MB PC2700(333mhz) CL2.0 ram

CONFIG:

JP31 IS REMOVED.
FSB SETTING @ 167MHz
12x MULTIPLIER
AUTO PCI DIVIDER
AUTO VCORE SETTING
10%VCORE INCREASE - DISABLED

theoretical 2004mhz, realistic: 2001.67mhz
333mhz RAM @ 334MHZ ddr.

idle temp is at 38 degrees when fan is at 7300rpm (although it is kinda chilly outside and i did open a window since my previous temp measurement)...and with previous settings, the temp dropped 3-4 degrees after i opened window. so i assume this will be around 44 or 45 with previous ambient temp conditions. case is closed in both measurements.

13x multiplier setting now prevents comp from booting at all, when bus is at 167mhz....have to reset bios to make it boot again.

here are screenshots from wcpuid and CPUCool

http://www.cloud-9.net/Leeoniya/2200+3.jpg
http://www.cloud-9.net/Leeoniya/cpucool-3.jpg

if anyone figures out a way to make this faster, (without changing my ram to 3000 or 3200 or more, please tell) with faster ram, much higher may be possible for fsb and overall speed.

btw...it also stopped posting as a 2200+ anymore. now it posts as a 2004mhz athlon in bios....not like it matters.

/* 8th */
mini update:
43 deg. idle w/case closed
(and window closed this time)

if you look at this:

http://www.cloud-9.net/Leeoniya/cpucool-4.jpg

you will notice the ambiant temp inside case went up by 5 deg. after i closed window.
...so did the cpu temp 38 to 43.

but considering 43 was also idle temp for it when it was 1,826mhz, pretty good results.

/* 9th */
well, about 3 weeks of running or so. concluded that the comp was too unstable to be usable. with the last settings i had. although temp was stable at about 43 degrees, still kept freezing, even with no load. and it did so quite often. that made me go "hmmm" something's ain't right.
basically the higher i set the multiplier, the more often it decided to freeze, with no significant rise in temp. at 11x multiplier still froze once in a while, so i'm like: wtf.

upped my ram voltage to 2.7 with the VIO jumper (jp10) (do note that i'm running 167 bus and 11x multiplier with pc2700 333mhz ocz ddr ram)

also upped my vcore to 1.675v in bios.

temp stays the same, and comp seems to run with a lot more stability, so far been a day, no problems (knock on wood)

if it stays stable. i'll try to OC by changing the multiplier to 11.5x, 12, or maybe even 13. 14 makes unlocked 2200+ not post, or drops the bios multiplier auto back to like 10x, so wont go there.

since there was no temp increase when i changed the multiplier before, conclusion: ram and cpu speed difference caused freezing. so i'll run it like this for 2 or 3 days and if stable, up the multiplier while monitoring temp.

/* 10th */
here's my latest, and by far, greatest setup. (best combo of speed and stability) upping the ram voltage to 2.7 and vcore to 1.65v and upping the multiplier to 12x did the trick. no longer freezing, and much faster.
AMD Athlon XP Thoroughbred 2200+ (1.8GHz) @ 2001.68mhz
iWILL XP333-R (ver.2.1) 4/11 BIOS
SuperAudio
256MB PC2700 CL2.0 333MHz OCz DDR RAM (copper heat spreader on ram is shipped with it, standard)
GeForce4 MX 440
WinXP Pro SP1

setup:

unlocked 2200+ (shorted NC and VSS pins)
jp31 on board removed
12x multiplier in bios
167 fsb in bios
1.675v core in bios
2.7v jp10 setting (vmem/vio) (ram voltage)
thermaltake 6cu+, lapped heatsink, Arctic Silver 3.

ram @ 333.61mhz


43 deg. stable as all hell. much more so than my previous, not overclocked setup. ironic.

at 13x multiplier, it doesnt post with the above settings. sad but true.

still, getting an extra 200mhz out of a t-bred is pretty good with standard cooling and only 43 degrees operating temp at idle, and 45 deg. at full load. Also, t-breds are not actually hard to overclock well, much harder than Palominos, so i'm happy. still want to find a way to get that 13x multipler though...untapped speed.


here are shots of wcpuid and cpucool

http://www.cloud-9.net/Leeoniya/cpucool.png
http://www.cloud-9.net/Leeoniya/XP2200+.png

/* 11th */
correction. up the vcore to 1.675v. 1.65 is original

/* 12th */
bump (in case anyone needs it)
although it appears a 2200+ is a dying breed, now that 2400 2600 and 2700 are out., heh.
in any case....^
     
      
舊 2002-12-09, 12:03 AM #11
回應時引用此文章
quizman離線中  
halohalo
Senior Member
 
halohalo的大頭照
 

加入日期: May 2000
您的住址: 拉古奧爾的大地上
文章: 1,383
簡言之必須把2200+解倍頻限制﹐他使用的則是短路cpu針腳的方法...
然後用舊製程cpu開機進入bios後修改相關參數後﹐再安裝2200+﹐由於
無法使用高倍頻﹐必須以提高外頻至166Mhz以達到2200+...

(只是這樣子和那些用1600+超上來的實在沒太大差別﹐這是我個人的感想﹐而且似乎沒人嘗試用更高速的新製程cpu測試xp333﹐因此目前關於xp333的支援力還是個謎﹐但是沒有高倍頻支援力的話實在不樂觀)
 
__________________
舊 2002-12-09, 01:42 AM #12
回應時引用此文章
halohalo離線中  
Reich 唐
Golden Member
 
Reich 唐的大頭照
 

加入日期: Oct 2000
您的住址: 台北市
文章: 3,232
目前很多人的CPU都用的好好的(就是1600+這一批啦 )隨便在XP333上應該也是以2100+(166X10.5)或是2200+(172X10.5)在跑,破倍頻的跑個190外頻以上的也有,目前應該還夠用吧,等Barton出來應該比較妥當....當然,我可能會跳到P4去啦。回到正題,XP333送回去修改後應該是可以支援到2000MHz左右的.13製程2400+,再以上個人就沒聽說過了。
__________________
舊 2002-12-09, 02:07 AM #13
回應時引用此文章
Reich 唐離線中  


    回應


POPIN
主題工具

發表文章規則
不可以發起新主題
不可以回應主題
不可以上傳附加檔案
不可以編輯您的文章

vB 代碼打開
[IMG]代碼打開
HTML代碼關閉



所有的時間均為GMT +8。 現在的時間是08:17 PM.


vBulletin Version 3.0.1
powered_by_vbulletin 2024。