- Joined
- Dec 27, 2001
- Location
- In the lab
TysonM said:
I already posted about my 1.6A, and it died several weeks AFTER I removed the IHS. I had it on watercooling, too, with good contact with the waterblock.
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
The Official Sudden Northwood Death Syndrome (S.N.D.S.) Thread[Retired sticky]
- Thread starter Silversinksam
- Start date
I already posted about my 1.6A, and it died several weeks AFTER I removed the IHS. I had it on watercooling, too, with good contact with the waterblock.
I saw this on the front page of [H], would like to add some things.
This is a very strange comment, but do you guys use USB mice?
After about 2 months of my [email protected] (1.55V, actual 1.53), it wouldn't boot at anything but 1.6. Couldn't figure out what was going on for the life of me, I tried new RAM first, but I noticed my intellimouse blinking furiously when I tried to boot above 1600 (even at a FSB of 101). So, I found my PS2 converter, and plugged my mouse into my PS2 port... and voila. It now boots at 2400 again, running flawlessly.
I've had my 1.6A since April 1, and I even installed Windows2000 with it running at a FSB133 (which was probably stupid). I have a cheap-o Maxtop case, with the top blow-hole, side blow-hole, and two rear fans, with the processor cooled by a Swiftech with a 45CFM YSTech. Northbridge and HSF both have ArcticSilver 3. Running Corsair PC2400 256MB x2, 2-2-5-2, stock voltage.
Temps with YSTech:
Idle: 30C
Load: 42C (don't like this temp)
Temps with Delta
Idle: 22C (basically ambient)
Load: 26C
Too bad I can't sleep with my Delta on, it cools so well.
This is a very strange comment, but do you guys use USB mice?
After about 2 months of my [email protected] (1.55V, actual 1.53), it wouldn't boot at anything but 1.6. Couldn't figure out what was going on for the life of me, I tried new RAM first, but I noticed my intellimouse blinking furiously when I tried to boot above 1600 (even at a FSB of 101). So, I found my PS2 converter, and plugged my mouse into my PS2 port... and voila. It now boots at 2400 again, running flawlessly.
I've had my 1.6A since April 1, and I even installed Windows2000 with it running at a FSB133 (which was probably stupid). I have a cheap-o Maxtop case, with the top blow-hole, side blow-hole, and two rear fans, with the processor cooled by a Swiftech with a 45CFM YSTech. Northbridge and HSF both have ArcticSilver 3. Running Corsair PC2400 256MB x2, 2-2-5-2, stock voltage.
Temps with YSTech:
Idle: 30C
Load: 42C (don't like this temp)
Temps with Delta
Idle: 22C (basically ambient)
Load: 26C
Too bad I can't sleep with my Delta on, it cools so well.
My 2.53 has been running at 3ghz @ 1.575v since the end of August with no problems at all *knocks on wood*. It runs rock solid.
Ive got a Swiftech MCX-478 w/ 80mm Vantec Tornado that gives me load temps of 51-55C depending on ambient. Soon to be watercooled (my stuff is in the mail).
Ive got a Swiftech MCX-478 w/ 80mm Vantec Tornado that gives me load temps of 51-55C depending on ambient. Soon to be watercooled (my stuff is in the mail).
TysonM said:
P3 and Celeron Tualatins have the exact same heatspreader as a P4 don't they? (I've got a Celeron Tually but never seen a P4 in the flesh to know for sure)
I've been running my Celly 1.0A @ 1.4, 1.65V vid-pin mod, for nearly 7 months, mostly 24/7, ~45 full load with stock HSF, no probs so far *gulp* hope it stays that way!
- Joined
- Apr 6, 2001
- Location
- Mount Vernon, WA
After comparing the p3 tualatin in my other computer and the p4 hanging from my rearview mirror in my car, the p3 one is bigger and the p4 one is a bit thicker (looks that way anyway). But im half asleep so I cant see straightdog said:
Later today Ill take a picture of them side by side.
OP
- Joined
- Aug 8, 2001
- Location
- Sunshine State, USA
- Thread Starter
- #66
I've been following this SNDS and the jury is still out, many people have experienced this SNDS even with Extreme cooling using Promethia and similiar.
My theory is that the P4's just DO NOT like high voltage even for short periods of time. You know as well as I do that many people like to 'Test' the upper most plethora of what thier chips can do, usually using 1.85volts or above.
My theory is that perhaps P4's do not tolerate high voltage even for short periods of time. But Like I said the jury is still out on this theory.
But theory or not I personally won't give a p4 any more than 1.7volts just to be on the safe side
I started a new thread HERE so we don't clutter this informational thread.
My theory is that the P4's just DO NOT like high voltage even for short periods of time. You know as well as I do that many people like to 'Test' the upper most plethora of what thier chips can do, usually using 1.85volts or above.
My theory is that perhaps P4's do not tolerate high voltage even for short periods of time. But Like I said the jury is still out on this theory.
But theory or not I personally won't give a p4 any more than 1.7volts just to be on the safe side
I started a new thread HERE so we don't clutter this informational thread.
After reading these forums for a few months I finally signed up and this is my first post.
I have experienced the SNDS twice although not complete failures. My first was with a 2.53 in Sept. I ran it at 2.85ghz for about 3 weeks at 1.65V and it started getting hot. The first week it ran around 35C. Second week temps went to a blazing 60C under load. Week three temps went to 75C at stock speed and would not overclock. I sent it back for a replacement. The second one started to get hot after a month. I was running it at 3ghz but now I feel it gets to hot at 60C under load. I'm running stock at 50C under load for the last month. As for the motherboards, I used a Gigabyte GA-8iexp and a Chaintech on the first proc and an IT7 on the second proc.
I've actually had worse luck with Athlons dying on me. I've had a 1700+ just up and die on me one day, non overclocked. I've had 2 2000+'s with problems as well. The first one I had Oc'd at 1.8Ghz with 1.8V. It decided it did not want to overclock one day no matter the voltage or the board so back to stock speed it went. The other 2000+ runs at 1250Mhz and not much more over that. It was at stock speed and stock voltage. It decided to crash on me a ton and then finally wouldn't boot. Pushing it back to 1250 it is stable but thats it. I use quality power supplies in everything as well as name brand and reputable boards. I guess you win some and lose some!
I have experienced the SNDS twice although not complete failures. My first was with a 2.53 in Sept. I ran it at 2.85ghz for about 3 weeks at 1.65V and it started getting hot. The first week it ran around 35C. Second week temps went to a blazing 60C under load. Week three temps went to 75C at stock speed and would not overclock. I sent it back for a replacement. The second one started to get hot after a month. I was running it at 3ghz but now I feel it gets to hot at 60C under load. I'm running stock at 50C under load for the last month. As for the motherboards, I used a Gigabyte GA-8iexp and a Chaintech on the first proc and an IT7 on the second proc.
I've actually had worse luck with Athlons dying on me. I've had a 1700+ just up and die on me one day, non overclocked. I've had 2 2000+'s with problems as well. The first one I had Oc'd at 1.8Ghz with 1.8V. It decided it did not want to overclock one day no matter the voltage or the board so back to stock speed it went. The other 2000+ runs at 1250Mhz and not much more over that. It was at stock speed and stock voltage. It decided to crash on me a ton and then finally wouldn't boot. Pushing it back to 1250 it is stable but thats it. I use quality power supplies in everything as well as name brand and reputable boards. I guess you win some and lose some!
hipro5 said:This S.N.D.S. will very SOON be S.A.D.S.(Sudden AMD Death Syndrome) as well and YOU WILL SEE IT....
My final words are :No more than 1,75VCore "MAX" at ALL CPU's made of 0,13ì (Intel - AMD)
QUOTE]
Maybe we could call it STD's since the it is a tbred core
1.6A Pack date - 06/12/02 Purchase date - 07/11/02
Epox 4G4A+ with stock Intel cooler.
Ran for several months at 150FSB/2.4Ghz with no problems. Decided to see if I could get a bit more out of it by upping the voltage. I used the .075 bios option increase and was able to run stable @ 158FSB/2.528Ghz for two or three months. Two or three weeks ago, W2K froze during boot... no BSOD. A couple of days later, got BSOD during boot. Lowered to 156FSB with same voltage. Ran for about a week and got another BSOD. Am now back at 150FSB with standard voltage. Looks like there may be something to this increased voltage thing, but I never took mine over 1.575. Temps were/are mid 50's under load.
Epox 4G4A+ with stock Intel cooler.
Ran for several months at 150FSB/2.4Ghz with no problems. Decided to see if I could get a bit more out of it by upping the voltage. I used the .075 bios option increase and was able to run stable @ 158FSB/2.528Ghz for two or three months. Two or three weeks ago, W2K froze during boot... no BSOD. A couple of days later, got BSOD during boot. Lowered to 156FSB with same voltage. Ran for about a week and got another BSOD. Am now back at 150FSB with standard voltage. Looks like there may be something to this increased voltage thing, but I never took mine over 1.575. Temps were/are mid 50's under load.
I do not mean to sound rude, but all of the problems you are all experiencing are your fault. It is a well documented fact (at least at anandtech) that it is dangerous to have the Vcore above 1.75. Most motherboards either overvolt or undervolt, so a safety standard is 1.70. It doesn't matter what type of cooling you have, whether it be Water Cooling, peltier, High Speed Fan, etc... when the Vcore is raised to 1.75 and above, a phenomena called electron migration takes place. In simple terms, what happens is the logic gates, and pathways carved by lithography break down. The processor can last from 1 second to a few months before the phenomena takes its hold, and destroys the processor. It doesn't matter whether the processor is at negative 20 degrees Celsius, or positive 200 Celsius (although the P4 will never reach this temp because of the onboard thermal protection, unlike AMD which topped over 380 Celsius on the review at tomshardware), the processor will suffer from electron migration. If you leave the voltage under 1.70, you should be fine, in theory. Just my two cents, if you want four, please respond.
[EDIT]
The reason that the Northwoods are not able to tolerate such high voltages is because of the .13 micron process within the core. Older Williamettes, as most of you can remember, could take around 2-2.2 Volts, and not hiccup. The only problem with running the Williamettes this high was because they ran very hot.
[/EDIT]
[EDIT]
The reason that the Northwoods are not able to tolerate such high voltages is because of the .13 micron process within the core. Older Williamettes, as most of you can remember, could take around 2-2.2 Volts, and not hiccup. The only problem with running the Williamettes this high was because they ran very hot.
[/EDIT]
Last edited:
i agree with KraziKid on most of what he says.
i haven't seen any issue pertaining to this. most nw's that i've had a chance to build/work with that have overclocked have been working great. none of them are pin modded though. i'd like to address a different issue though.
most of you are completely going against the manufacturer's spec. the spec is there for a reason. if you overclock, be prepared to accept the consequences. these parts are binned for a reason. there are test patterns and procedures intel goes through to validate certain cpus at certain frequencies and voltages. there's no guarantee that you're going to not fry your cpu by doing this. when you increase the voltage on the cpu, it changes the entire characteristics of it. dynamic power and static power consumption skyrocket from the smallest bump in core voltage. you're talking about something that probably has something close to 45+ amps running through it, more at full utilization. power is going to increase by a lot, put more stress on the mainboard that may have cheap mosfets, and you're asking for trouble. you may be talking about a good part at it's rated frequency that is marginal at higher frequencies. and some of you sending back mainboards and cpus that you have already voided the warranty on? you should eat the cost. if you're not prepared to do so, don't overclock. sorry for the rant, but use some common sense.
i haven't seen any issue pertaining to this. most nw's that i've had a chance to build/work with that have overclocked have been working great. none of them are pin modded though. i'd like to address a different issue though.
most of you are completely going against the manufacturer's spec. the spec is there for a reason. if you overclock, be prepared to accept the consequences. these parts are binned for a reason. there are test patterns and procedures intel goes through to validate certain cpus at certain frequencies and voltages. there's no guarantee that you're going to not fry your cpu by doing this. when you increase the voltage on the cpu, it changes the entire characteristics of it. dynamic power and static power consumption skyrocket from the smallest bump in core voltage. you're talking about something that probably has something close to 45+ amps running through it, more at full utilization. power is going to increase by a lot, put more stress on the mainboard that may have cheap mosfets, and you're asking for trouble. you may be talking about a good part at it's rated frequency that is marginal at higher frequencies. and some of you sending back mainboards and cpus that you have already voided the warranty on? you should eat the cost. if you're not prepared to do so, don't overclock. sorry for the rant, but use some common sense.
OP
- Joined
- Aug 8, 2001
- Location
- Sunshine State, USA
- Thread Starter
- #75
Ange062 said:
woohoo, [H]ers rule!!
The Inquirer also picked up this story
http://www.theinquirer.net/
I have a 1.6A since Feb this year, it is running @ 1.7V @ 2.24G and it always bsod under WinXP in Nov. But I am very lucky, I returned it to the distributor and have refund because they don't have 1.6GA anymore.
Now, I am using Celeron 2.0G @1.525V running 2.88G. I totally agree the C1 stepping don't like high voltage. The temp. goes up rapidly if I increase the voltage.
Have fun with your Intel CPU !
Now, I am using Celeron 2.0G @1.525V running 2.88G. I totally agree the C1 stepping don't like high voltage. The temp. goes up rapidly if I increase the voltage.
Have fun with your Intel CPU !
- Joined
- Sep 18, 2001
- Location
- Rhode Island, USA
My 2.0a ran 5 months, overclocked it occasionaly, but the stepping codes on it blew, so it would not go over ~2.4 stable, so I left it at 2.2. One day, it wouldn't boot, rma'd both the CPU and memory and now it runs 2.8g, no problems. The reason why I think it died is because I had a TON of static electricity in my room, everytime I walked in, I touched my case and BANG, I think it back-fed thru the case into the cpu/memory.
Jim
Jim
borealiss said:
That one gets me too. If my 1.6A dies from overclocking, it is my responsibility... not the vendor. I wouldn't think of trying to return a processor that may have run for twenty years at stock settings.
Those three IBM 75GXP's are another story altogether...
- Joined
- Sep 21, 2001
no way
i found this this link that claims to end this problem.
wow, this may be the solution that works)
i found this this link that claims to end this problem.
wow, this may be the solution that works
)Similar threads
