Pentium D's Suck for DC

[caferace]

My Pentium 805D (2.66x2 currently OC'ed to 3.4) has run VERY cool from day one. It does nothing besides DC though. Nada. According to various programs, it's almost running *too* cool, at ~42C at full load. That's with a Big Typhoon and arctic silver paste. Great airflow in the case as well.

My FX-55 on the other hand runs WAY hot, at a 10% overclock. One of these days, I need to dump the stock cooler. My bad.

-jim

 

[Smoke]

Hey Wiz, that looks like a neat forum. I need to visit there more often.

I guess you guys are just going to have to keep telling me things over and over ... that's your burden of being so young and having such old geezers as friends. :laugh:

 

[Crazee]

Originally posted by: Smoke
Hey Wiz, that looks like a neat forum. I need to visit there more often.

I guess you guys are just going to have to keep telling me things over and over ... that's your burden of being so young and having such old geezers as friends. :laugh:

I would visit those forums more often, but I have heard bad things about the webmaster there

 

[Wiz]

Nah, he's a class act :wine:

 

[Smoke]

Here's another update on the temperature and overclock status of my office machines.

The CPU temps for my overclocked PentiumD 2.66MHz system remain at the lower levels as reported earlier. My attempts to move the system back to previously overclocked levels has been a little disappointing. In fact the system seems to balk whenever I go any higher than 3.67MHz. So there are other factors involved beyond temperatures that have slowly degraded the overclocking abilities of this system.

This morning I cleaned and reapplied Artic Silver to the non-overclocked PentiumD 2.8MHz machine. The temperatures have shown a similar improvement. Currently with BOINC running 24/7 the CPU Temperature is ranging between 57C - 59C which is about 6C lower than before.

 

[Wiz]

CPU Temperature is ranging between 57C - 59C

Those are excellent temps!

I wonder about the 2.66 at only..(cough)... 3.67ghz

(that's still a really nice OC, nothing to be ashamed of there!)

It definitely could have other issues, memory, mobo, who knows?
I know you must at least feel better with it running the same speed but cooler!

 

[Assimilator1]

Smoke
The bulging capacitors isssue was a result of a massive batch of defective caps being sold to various mbrd manufacturers ,mainly around the 99-02 era IIRC.It was nothing to do with heat ,though of course if they blow then that can output quite a bit of heat
Check out this site ,makes for interesting reading ,with industrial espionage & all!

My main rig's still a Skt A ,I bought this mbrd nearly 4yrs ago (wow that's the longest I've ever kept a mbrd!:Q) ,I haven't noticed any general trend in increasing temperatures other than when I've upgraded CPUs ,upped the clock or vcore.

Have you re-applied the HS compound to see if this returns the temperatures to their as new temps? ,this seems the most likely answer to me mate

The increase in temperatures when running BOINC is amazing!
That is a lot! ,but don't forget that P4s are known for running hot ,more so the Prescott cores.
My XPM @2.5GHz ,1.68v running DPAD now is running at 39C (ext CPU temp sensor) ,55C (int CPU sensor) with 27C case temp.The CPU fan is temp controlled,& is currently at ~3900 RPM
With DPAD off for about 10mins it runs @ 27/36/48C ,fan @ ~3400RPM.
Of course the varying fan speed narrows the temp difference quite a lot ,which isn't helpful for us but I thougt it might be useful anyway.

Oh & inccidently the CPU runs at about the same temp whether running DPAD or BOINC ,it does run hotter with P95 though ,by about 2-3C.

Btw you owe me a gold star
(I see now someone else suggested the HS compound idea too)
Is that 4-6C drop comparable to 'as new' temps?

Something I do agree about though is the gradual lessening abilty of some CPUs (or mbrds?) to overclock at their 'as new' level.
I've been running this XPM @ 2.5GHz for a little over 2yrs now ,it used to run fine at that speed @ 1.66v ,but about a month ago I started to have problems with my rig crashing.
A P95 test showed that the CPU was no longer stable ,PC would lock after a few hours or so.It used to pass a 24hr P95 test ,I've upped the vcore to 1.68v & it seems to be ok now ,but something's fading!
Also ,many years ago I had a K6-2 350 ,for about 2mths it was fine at 420MHz ,then 1 day it just refused to load windows (BSOD) ,backed it off to 400MHz & it was fine.

Good to see you'll be crunching more & your rigs running cooler

Wiz
Wow you had a massive drop with new AS5!:Q

[edit]woah ,just realised that's the 1st long discussion I've had in DC for ages!

 

[BadThad]

Originally posted by: Assimilator1
Smoke
The bulging capacitors isssue was a result of a massive batch of defective caps being sold to various mbrd manufacturers ,mainly around the 99-02 era IIRC.It was nothing to do with heat ,though of course if they blow then that can output quite a bit of heat
Check out this site ,makes for interesting reading ,with industrial espionage & all!

My main rig's still a Skt A ,I bought this mbrd nearly 4yrs ago (wow that's the longest I've ever kept a mbrd!:Q) ,I haven't noticed any general trend in increasing temperatures other than when I've upgraded CPUs ,upped the clock or vcore.

Have you re-applied the HS compound to see if this returns the temperatures to their as new temps? ,this seems the most likely answer to me mate

The increase in temperatures when running BOINC is amazing!
That is a lot! ,but don't forget that P4s are known for running hot ,more so the Prescott cores.
My XPM @2.5GHz ,1.68v running DPAD now is running at 39C (ext CPU temp sensor) ,55C (int CPU sensor) with 27C case temp.The CPU fan is temp controlled,& is currently at ~3900 RPM
With DPAD off for about 10mins it runs @ 27/36/48C ,fan @ ~3400RPM.
Of course the varying fan speed narrows the temp difference quite a lot ,which isn't helpful for us but I thougt it might be useful anyway.

Oh & inccidently the CPU runs at about the same temp whether running DPAD or BOINC ,it does run hotter with P95 though ,by about 2-3C.

Btw you owe me a gold star
(I see now someone else suggested the HS compound idea too)
Is that 4-6C drop comparable to 'as new' temps?

Something I do agree about though is the gradual lessening abilty of some CPUs (or mbrds?) to overclock at their 'as new' level.
I've been running this XPM @ 2.5GHz for a little over 2yrs now ,it used to run fine at that speed @ 1.66v ,but about a month ago I started to have problems with my rig crashing.
A P95 test showed that the CPU was no longer stable ,PC would lock after a few hours or so.It used to pass a 24hr P95 test ,I've upped the vcore to 1.68v & it seems to be ok now ,but something's fading!
Also ,many years ago I had a K6-2 350 ,for about 2mths it was fine at 420MHz ,then 1 day it just refused to load windows (BSOD) ,backed it off to 400MHz & it was fine.

Good to see you'll be crunching more & your rigs running cooler

Wiz
Wow you had a massive drop with new AS5!:Q

[edit]woah ,just realised that's the 1st long discussion I've had in DC for ages!

Agreed once again. I've had a LOT of overclocked PC's that just seem to gradually fade in OC ability. I keep backing them down until I end up back as the stock speed, lol.

Glad I made a thread that finally got you talking again! LMAO

 

[BadThad]

Well, let's see how it holds up now. I moved the mobo/cpu into a different case and installed the Artic Freezer cooler....a microATX, LOL. I'm just going to leave the side off. Idle temp is about 36°C and 46-49°C with Rosetta running two processes. It's a tight fit in this microATX case:

Pic 1
Pic 2
Pic 3
Pic 4

Another funny thing, it's using an old 350W TSP power supply with a whopping 15A on the 12V rail and only a 20-pin connector to the mobo. So far, it's running fine.

 

[Wiz]

Wiz

Wow you had a massive drop with new AS5!

Yes, this made a believer out of me (yet again)

Now I run my 3.0 OC'd to 3.46ghz and it does almost as well in Rosetta as the 3.8ghz machine I have at work.

Best of all, no corruption while running OC'd means no downtime.

Nice pics BadThad, that's a great cooler - I really like the design.
Sucker is big though!

 

[BadThad]

Originally posted by: Wiz

Wiz

Wow you had a massive drop with new AS5!

Yes, this made a believer out of me (yet again)

Now I run my 3.0 OC'd to 3.46ghz and it does almost as well in Rosetta as the 3.8ghz machine I have at work.

Best of all, no corruption while running OC'd means no downtime.

Nice pics BadThad, that's a great cooler - I really like the design.
Sucker is big though!

Tight fit in a microATX case. This machine is nothing more than a full-time cruncher now. It's going to stay just like the pics show. The Pentium D is just not a good choice for my daughter's PC, her room get's pretty hot in the summer and I know I'd be back in the same boat again eventually.

I ordered her a Gigabyte DS3 Wednesday, I'm going to put that in her PC with an E6300. That's a much better CPU for a hot environment compared to a Pentium D, lol.

 

[Assimilator1]

I've never had to drop a CPU back to standard speed
Though once I did kill a S7 Cyrix! (shows how crap they were!)

I've been around ,but just don't seem to bump into you guys so much these days ,I guess I'll just have to start running R@H soon.......especially when I go for a Core 2 upgrade in a few months (know of any good overclocking mbrds for them?)

 

[BadThad]

Originally posted by: Assimilator1
I've never had to drop a CPU back to standard speed though
Though once I did kill a S7 Cyrix! (shows how crap they were!)

I've been around ,but just don't seem to bump into you guys so much these days ,I guess I'll just have to start running R@H soon.......especially when I go for a Core 2 upgrade in a few months (know of any good overclocking mbrds for them?)

I have the P5B Deluxe, great mobo so far! ....but save your money, they are expensive. Another good choice is the Gigabyte DS3 and it's a bit cheaper. No matter what, you've got to get one of these. I love Core 2 Duo!

 

[Assimilator1]

Originally posted by: BadThad

Originally posted by: Assimilator1
I've never had to drop a CPU back to standard speed though
Though once I did kill a S7 Cyrix! (shows how crap they were!)

I've been around ,but just don't seem to bump into you guys so much these days ,I guess I'll just have to start running R@H soon.......especially when I go for a Core 2 upgrade in a few months (know of any good overclocking mbrds for them?)

I have the P5B Deluxe, great mobo so far! ....but save your money, they are expensive. Another good choice is the Gigabyte DS3 and it's a bit cheaper. No matter what, you've got to get one of these. I love Core 2 Duo!

Oh I will ,I will! ,I can't wait.........well actually I am & I will ,lol

Does the DS3 have 2xIDE sockets?

(btw you guys are slacking ,I mentioned I might do R@H soon & no one tried to reel me in!?:Q )

 

[Smoke]

Originally posted by: Assimilator1

.......

(btw you guys are slacking ,I mentioned I might do R@H soon & no one tried to reel me in!?:Q )

We know how shy you are and didn't want to spook you.

You know you would certainly be appreciated. Why, I might even stop referring to you as Ass1! :laugh:

 

[Assimilator1]

lol ,well in that case I won't call you grandad again

 

[Insidious]

yeah, yeah, I'm late to the party..... again

I wish I could come up with a link, but I once read an article talking about RAM (and memory in general) degrading over time. If I remember correctly, static charges/residual energy degraded the doping of the semi-conductor material over time.

The article had some evidence that down-time tended to allow some dissipation of these charges within the molecular matrix of the semi-conductor and some of the degradation (but not much) could be reversed. It also pointed to more evidence that this was one of the very few components in a modern PC that continuous operation seemed to exaccerbate. (conventional wisdom tends to support an argument that most electronic components have a longer life if thermal cycling of powering on and off is reduced or eliminated)

So anyway, this may address the difficulty of maintaining stable OCs over time.

But of course this is ALL YOUR FAULT for not upgrading often enough

Love,
Sid

 

[Assimilator1]

lol

Interesting though ,got more info about the memory degredation issue?

Btw I must really be in the doghouse with my main PC! it's got a m/brd that's 4yrs old in March!:Q ,I swear I haven't done that before! (*scratches head* ,nope previous record was my old Slot 1 Soyo mbrd which I had for just over 2yrs)

Good old skt A ,getting close (if not past) to being past it now though

 

[Insidious]

This paper talks about the degradation mechanism, but doesn't address how/if it is affected with age or operation of the memory module.

(you asked for it.... :laugh: )

-Sid

 

[Wiz]

This all brings to mind the webcam I saw a year or two ago of some light bulb in a firehouse that's been burning for something like 105 years straight.

Just to illustrate that some components benefit from the lack of thermal cycling, while others - not so much.


Ah - here it is: Centennial LightBulb

 

[BadThad]

Originally posted by: Smoke

Originally posted by: Assimilator1

.......

(btw you guys are slacking ,I mentioned I might do R@H soon & no one tried to reel me in!?:Q )

We know how shy you are and didn't want to spook you.

You know you would certainly be appreciated. Why, I might even stop referring to you as Ass1! :laugh:

LOL! Yea, that's the ticket!

 

[BadThad]

Originally posted by: Insidious
yeah, yeah, I'm late to the party..... again

I wish I could come up with a link, but I once read an article talking about RAM (and memory in general) degrading over time. If I remember correctly, static charges/residual energy degraded the doping of the semi-conductor material over time.

The article had some evidence that down-time tended to allow some dissipation of these charges within the molecular matrix of the semi-conductor and some of the degradation (but not much) could be reversed. It also pointed to more evidence that this was one of the very few components in a modern PC that continuous operation seemed to exaccerbate. (conventional wisdom tends to support an argument that most electronic components have a longer life if thermal cycling of powering on and off is reduced or eliminated)

So anyway, this may address the difficulty of maintaining stable OCs over time.

But of course this is ALL YOUR FAULT for not upgrading often enough

Love,
Sid

Bizzare, I would have never guessed that. I always considered memory modules dasm near invincible.

Oh, and I'm a firm believer in leaving a PC on 24/7. Power cycling is evil!

 

[Insidious]

I'm with you Thad

I turn 'em off to clean them, and that's pretty much it. OK, I turn them off to sell them too

-Sid

 

[BadThad]

Originally posted by: Insidious
I'm with you Thad

I turn 'em off to clean them, and that's pretty much it. OK, I turn them off to sell them too

-Sid

LMAO

Besides, they can't crunch DC when they're off!

 

[Insidious]

Protien bendin' hunks of burning LOVE! :thumbsup:

-Sid