The plague and C1 chips... [Archive]

SBeaver

12-07-02, 11:04 AM

I know that C1's havn't been available for very long but still I ask:
Have anyone seen any confirmed cases of S.N.D.S on these?
It would be really good to know this, I mean, maybe they are better when it comes to taking high voltage?
We'll probably know if this is true as C1's become more widespread but if anyone of you have experienced anything that could be considered a symptom or if you, on the contrary, have found the C1 to be a voltage hungry beast that can take anything you throw at it, feel free to post here.

batboy

12-07-02, 11:37 AM

The basic chip design is the same on the C1 stepping, just mainly some improvements were made and slightly better yielding cores. I suspect they still will suffer if you overvolt them just like the B0.

SBeaver

12-07-02, 11:52 AM

Originally posted by batboy
The basic chip design is the same on the C1 stepping, just mainly some improvements were made and slightly better yielding cores. I suspect they still will suffer if you overvolt them just like the B0.

mm... your probably right, but there is allways a chance that they looked it over and found something in there that was too weak that could be responsible for the failures.
Just wait and see I guess.

crotale

12-07-02, 12:05 PM

I hope it was something in the B0 chips that burnt very easy...
And they've fixed it now! =)

Nails

12-07-02, 12:23 PM

Have you considered what the new variable voltage C1's will mean for SNDS?

AntmanMike

12-07-02, 01:26 PM

There wasn't anything wrong, in Intel's eyes. YOU were overvolting, going past Intel specifications, so it died. That is not their problem.

SBeaver

12-07-02, 02:05 PM

Originally posted by AntmanMike
There wasn't anything wrong, in Intel's eyes. YOU were overvolting, going past Intel specifications, so it died. That is not their problem.

Your post has no meaning.
Your sig is also rather lame.

Kasmodean

12-07-02, 02:54 PM

His post makes perfect sense. Why would Intel bother fixing something that according to their specs isn't broken. They are too busy on their road plan trying to beat amd in performance to worry how much voltage their chips can take beyond their defaults.

Overclocker456

12-07-02, 08:37 PM

I remember testing out a 2.4B C1 for a friend and I was only able to overclock it to 2.5GHz!!! WHAT!!! I think that is a quality control issue there.. shhessh.. I remember a 2.0GHz willy overclocked more than 100MHz.. god.. Some C1 overclock like crap.. I'm not sure what's to blame.. but surely a 2.4GHz CPU that can't do more than 2.5GHz is a sign of a poor quality chip going to the customer.

arklab

12-08-02, 09:06 AM

I wonder if you guys are maybe missing the point of early C1 overclocking results?

At least with the lower mutliplyer chips (esp. the 1.8A's, which is my experience) they don't need much over-voltage, if any!

The emerging patern seems to be a really nice OC at stock voltage, which is a little higher than the B0's anyway.

I got 3.0 from a 1.8A at stock volts and stock heatsink/fan.

The OMG chip had better cooling (and prob. a better chip). He got to 3.6GHz with slight over-volt, but things stopped right there.

If the pattern holds, the C1 will do just about everything they can do near stock voltage.

The question may now become how much of SNDS is voltage and how much is OC?

I hope we may see an end to Sudden Death with C1.

Hawk

12-08-02, 06:29 PM

Originally posted by Overclocker456
I remember testing out a 2.4B C1 for a friend and I was only able to overclock it to 2.5GHz!!! WHAT!!! I think that is a quality control issue there.. shhessh.. I remember a 2.0GHz willy overclocked more than 100MHz.. god.. Some C1 overclock like crap.. I'm not sure what's to blame.. but surely a 2.4GHz CPU that can't do more than 2.5GHz is a sign of a poor quality chip going to the customer.
Quality control issue?? How do you figure that? Did the chip run stable at 2.4? That is what your friend bought and there is no promise it will run faster only that it will run at 2.4 and remain stable and last through the warranty period.
Where do you see a QC issue? And how do you call that poor quality? If you buy a car that is rated at x HP and you don't get more is that a quality issue? Remember ANY thing you get over stock is a bonus. There are no promises from chip makers that they will run 1 MHz over stock. They actually tell you there is no warranty for any thing over stock.:cry:

Overclocker456

12-08-02, 06:43 PM

Originally posted by Hawk

Quality control issue?? How do you figure that? Did the chip run stable at 2.4? That is what your friend bought and there is no promise it will run faster only that it will run at 2.4 and remain stable and last through the warranty period.
Where do you see a QC issue? And how do you call that poor quality? If you buy a car that is rated at x HP and you don't get more is that a quality issue? Remember ANY thing you get over stock is a bonus. There are no promises from chip makers that they will run 1 MHz over stock. They actually tell you there is no warranty for any thing over stock.:cry:

yeah but the C1's are the future of the Pentium 4's now if the newest P4 stepping that just came out can't overclock 100 or 4.1% past it stock speed, it shows that the stepping isn't as evolved as we are lead to believe.

Hawk

12-08-02, 06:49 PM

Originally posted by Overclocker456

yeah but the C1's are the future of the Pentium 4's now if the newest P4 stepping that just came out can't overclock 100 or 4.1% past it stock speed, it shows that the stepping isn't as evolved as we are lead to believe.
Still not a QC issue.
I am just wondering if you could point me to where the manufacturer lead you to beleive the C1 stepping would overclock a given amount?
Thanks
Hawk

Overclocker456

12-08-02, 07:07 PM

Originally posted by Hawk

Still not a QC issue.
I am just wondering if you could point me to where the manufacturer lead you to beleive the C1 stepping would overclock a given amount?
Thanks
Hawk

The manufacturer never said anything about overclock and we know that..But Intel is going to release a .13micron 3.2GHz Pentium 4 based on the C1 stepping, They can't use the B0 because it can't reach speeds like that reilably. That's what I'm saying.. The C1 stepping is supposed to have a higher limit than the B0 but of course that isn't the case always as proves the 2.4B not getting more than 2.5GHz..

Hawk

12-08-02, 07:34 PM

Originally posted by Overclocker456

The manufacturer never said anything about overclock and we know that..But Intel is going to release a .13micron 3.2GHz Pentium 4 based on the C1 stepping, They can't use the B0 because it can't reach speeds like that reilably. That's what I'm saying.. The C1 stepping is supposed to have a higher limit than the B0 but of course that isn't the case always as proves the 2.4B not getting more than 2.5GHz..
You are avoiding the point. You said it was a quality control issue which it is NOT. If the chip performance is as stated thats all you are promised.
C1's do haVE A HIGHER STOCK LIMIT. tHE MANUFACTURER BASES THIS CALLS AN STOCK SETTINGS.

AntmanMike

12-08-02, 07:42 PM

Originally posted by SBeaver

Your post has no meaning.
Your sig is also rather lame.

No personal attacks allowed on the forum sir.

I have to agree with Hawk. That is not a quality control issue, since Intel never guaranteed it would overclock. The only reason they make new steppings is so CPUs can run at higher STOCK speeds more stable. Just because a B0 OCs more than a C1 isnt a quality control issue. Unless Intel markets a chip that they GUARANTEE a high overclock with, it is NOT a quality control issue. If you don't like the C1, don't buy it.

Me, I will be waiting until a find a C1 1.8 on E-Bay.

Hawk

12-08-02, 07:53 PM

Originally posted by AntmanMike

No personal attacks allowed on the forum sir.

I have to agree with Hawk. That is not a quality control issue, since Intel never guaranteed it would overclock. The only reason they make new steppings is so CPUs can run at higher STOCK speeds more stable. Just because a B0 OCs more than a C1 isnt a quality control issue. Unless Intel markets a chip that they GUARANTEE a high overclock with, it is NOT a quality control issue. If you don't like the C1, don't buy it.

Me, I will be waiting until a find a C1 1.8 on E-Bay.
WOOOah! we agree on something?
;)

AntmanMike

12-08-02, 07:59 PM

:o. Naw, its just common sense. Intel never said that "The C1s will overclock better than the B0s". In fact, they never said that they will overclock at all. Anyone who argues otherwise doesn't understand that quality control is only what the company that makes the CPUs considers quality. Intel considers quality to be when the CPU runs perfectly at default speed, default voltage.

markodude

12-09-02, 06:28 AM

going back to something arklab said
"The emerging patern seems to be a really nice OC at stock voltage, which is a little higher than the B0's anyway."
Have to agree with my C1 experience, I can get my 2.5 to 2.8 at default, PRime95 stable....even with 1.825v it wont go any higher :(
Although I havent tried any extreme cooling yet....
BTW I wouldnt look on e-bay for a processor that OC's well, I mean who would sell a decent 1.8A C1 on Ebay, when they are seeming to OC better than 2.8's??? Generally the proc's on ebay either are bad overclockers or have been fried....sometimes you are lucky and get a good one but I dont think its worth the risk, anyway C1 1.8A's prolly wont be on there until mid/late 2003.

FIZZ3

12-09-02, 06:53 AM

Originally posted by markodude
going back to something arklab said
"The emerging patern seems to be a really nice OC at stock voltage, which is a little higher than the B0's anyway."
Have to agree with my C1 experience, I can get my 2.5 to 2.8 at default, PRime95 stable....even with 1.825v it wont go any higher :(

That experience is not exclusive to C1's at all. I have had the same pattern (stock voltage up to xx Mhz fully stable but 1.80v + magic dance xx+1 Mhz = crash) on several B0 chips.

I think this is actually related to the P4 NW design. When these chips were introduced I remember reading about how the chip would behave differently to voltage changes. This could be what we are seeing here.

xGRiMx

12-09-02, 09:27 AM

Intel has specified an absolute max Vcore of 1.75V. Every spike over that will damage the cpu and the effect will be much higher than on palomino's because of 0.13micron process compared to 0.18micron process. The same voltage for palomino is 2.2V so it usually wont be achieved without modding your motherboard unlike p4 motherboards that usually allow 1.85V Vcore or even higher.
The 2nd thing causing p4's to "die" is bad voltage regulating in some motherboards (especially abit's older motherboards, no info about the new ones)

SBeaver

12-09-02, 10:53 AM

Tamasha_Strife

12-09-02, 12:02 PM

To me there does seem to be an issue regarding C1 procs. You have to remember that CPUs are made then they test it and decide what speed it'll be sold as. If they dont have enough "low-end" chips, they'll grab some from the higher speed bins. With B0 stepping, pretty much every chip could make it over 2.6ghz and the top speed was 2.53ghz. That means all chips were good enough to be sold as high end CPUs but got thrown in the lower speed bins. With the C1 stepping we see something very different, some CPUs do great, some can't seem to OC at all. That means a lot more processors are sold at the actual speed they test at. To me this suggests that the C1 stepping production process is much more variable in the quality of the chips it produces than the B0 process. Of course, there is countinuous improvement made in the production process so things should change with time.

xGRiMx

12-09-02, 01:14 PM

Originally posted by xGRiMx
Intel has specified an absolute max Vcore of 1.75V. Every spike over that will damage the cpu and the effect will be much higher than on palomino's because of 0.13micron process compared to 0.18micron process. The same voltage for palomino is 2.2V so it usually wont be achieved without modding your motherboard unlike p4 motherboards that usually allow 1.85V Vcore or even higher.
The 2nd thing causing p4's to "die" is bad voltage regulating in some motherboards (especially abit's older motherboards, no info about the new ones)

I haven't checked amd's specs for tbred, but as its also produced on 0.13micron line, exceeding Absolute Vcore can cause the same kind of problems. Only time will tell :)
Me myself am going to buy a 2.66GHz northwood and overclock it with max Vcore of 1.65V. Hopefully i can achieve 3.2GHz+ at this voltage, because i dont want to go any higher.