Upgraded CPU in laptop, now it randomly shuts off???

[VirtualLarry]

Your understanding is incorrect.

Do you have documentation otherwise?

I've read some info that mentioned upgrading a laptop that used a lesser chipset, "for celeron and Pentium", with an i3, and it would then shut down, like clockwork, at the 15 minute mark.

NOT temp related.

 

[Phynaz]

Do you have documentation otherwise?

I've read some info that mentioned upgrading a laptop that used a lesser chipset, "for celeron and Pentium", with an i3, and it would then shut down, like clockwork, at the 15 minute mark.

NOT temp related.

You're asking me to prove a negative when other people in this thread have said you are wrong?

No.

You made the claim, you have to back it up. It's not my job to prove your claim false.

 

[jj109]

Do you have documentation otherwise?

I've read some info that mentioned upgrading a laptop that used a lesser chipset, "for celeron and Pentium", with an i3, and it would then shut down, like clockwork, at the 15 minute mark.

NOT temp related.

I've also read some info that VirtualLarry eats babies for breakfast.

Coincidence?


Insults are not tolerated here
Markfw900
Anandtech moderator

 

[Phynaz]

I've also read some info that VirtualLarry eats babies for breakfast.

Coincidence?

Prove it false Larry

 

[VirtualLarry]

You made the claim, you have to back it up. It's not my job to prove your claim false.

http://www.maximumpc.com/forums/viewtopic.php?f=36&t=241797

Edit: And our very own thread here:
http://forums.anandtech.com/showthread.php?t=2374689

Gillbot confirms the lock.

 

[Phynaz]

Ok, I'll give you that one Larry, first I've heard of it.

 

[Zodiark1593]

In many cases, the same board is used among laptops of the same family, and just use different processors between them. Theoretically, you should have no issue running an i5-2xxxM, and even an i7-2xxxM, though the quad core variants will push your cooling. (Example: I can put an i7-720QM quad core chip in my laptop and have it work, doesn't mean I should though.)

However, you've stepped up to an Ivy Bridge chip, which while it uses the same socket, is ultimately seen differently by the BIOS. If I remember correctly, desktop boards needed a BIOS update to work properly with an Ivy chip. As a new chipset also launched, it's unlikely there will be a BIOS revision to help you out in your case. I recommend going to a dual core Sandy Bridge chip instead. (2xxxM instead of 3xxxM numbering)

 

[ViRGE]

I'm surprised no one has suggested power delivery yet.

While both chips are nominally 35W, in practice the i5 is going to consume more power. The Pentium is likely not approaching 35W (due to the fact that it's clocked lower and has fewer features enabled), whereas the i5 would be.

If that laptop was truly built around the Pentium processor, then Toshiba may not have designed the power delivery system to handle any more power than what a Pentium requires. Which would certainly explain why it shuts off, since that would be a failsafe protection mechanism.

However, you've stepped up to an Ivy Bridge chip, which while it uses the same socket, is ultimately seen differently by the BIOS. If I remember correctly, desktop boards needed a BIOS update to work properly with an Ivy chip. As a new chipset also launched, it's unlikely there will be a BIOS revision to help you out in your case. I recommend going to a dual core Sandy Bridge chip instead. (2xxxM instead of 3xxxM numbering)

Zodiark, both chips are Ivy Bridge. Intel in all their wisdom has the Pentium chips a generation behind in numbering in order to make the products look more differentiated. Pentium 2000 is IVB, Pentium 3000 is HSW, and Pentium [3 digit] is SNB.

 

[ShintaiDK]

I'm surprised no one has suggested power delivery yet.

If you lookup the Toshiba model you see HM70 that is locked to Celeron and Pentium. And we had a thread about it previous as well with this exact behaviour when trying to use an ix CPU. Its the compromise for the OEM to save a few $.

 

[Maximilian]

Wow Intel suck D:

First ive heard of that behavior.

 

[ShintaiDK]

Wow Intel suck D:

First ive heard of that behavior.

Why? Because the OEM went cheapskate?

You know, the same OEMs that forces nVidia and AMD to rebrand GPUs, just to try and sell the same thing twice.

 

[SPBHM]

yes, this restriction is well known and 100% ridiculous.

Why? Because the OEM went cheapskate?

who cares, he bought the NB with a Pentium just like Intel wanted, now if he wants to DIY upgrade the CPU for one 100% physically compatible it should work fine, like it used to with older "pentium/Celeron but not core" chipset (like GL40), and it works fine for some time, until Intel decides to shutdown your computer.

 

[sm625]

I have no trouble believing Intel would do something like this, but I must point out that this sort of problem requires nothing so sinister. An i5 can produce much larger voltage and current swings on a power supply compared to a 2020m. It could be that one of those swings results in a distortion large enough to crash the cpu. To verify this you need to compare the laptop motherboard with the same version that ships with an i5. If you compare the two boards side by side I bet you'll see that the 2020m version has lower value capacitors, lower power regulators, missing components, etc etc.

 

[ShintaiDK]

he bought the NB with a Pentium just like Intel wanted,

And here I thought it was Toshiba that designed, manufactored and sold the laptop without any relevance to Intel in the process besides being a component supplier for 2 parts in the laptop. :hmm:

Clearly its evil Intel that is at fault. How dare they give more options to OEMs in a free market economy depending on what you wish to pay, that OEMs on their own can choose from. Specially when some are directly OEM requested as a new SKU.

 

[VirtualLarry]

I have no trouble believing Intel would do something like this, but I must point out that this sort of problem requires nothing so sinister. An i5 can produce much larger voltage and current swings on a power supply compared to a 2020m. It could be that one of those swings results in a distortion large enough to crash the cpu. To verify this you need to compare the laptop motherboard with the same version that ships with an i5. If you compare the two boards side by side I bet you'll see that the 2020m version has lower value capacitors, lower power regulators, missing components, etc etc.

But... same TDP? Shutdown in 30 minutes, even at idle?

 

[Lakeshorecoupe]

But... same TDP? Shutdown in 30 minutes, even at idle?

Exactly.. voltage on the i5 was similar to the 2020m as well,they both sat around .85-.90v. Only time the i5 went higher was under 100% load and the turbo was overclocking, in which the max it went was .98-.99v . The i5 also only ran about 5-10c warmer depending on what you were doing. Only time the i5 would run significantly hotter is if you were at 100% overclocked load.

 

[Ketchup]

Lakeshorecoupe, I don't know if you looked at this, but thought I would mention it:

According to the support page for this model, there are three BIOS updates since the original:

http://support.toshiba.com/support/modelHome?freeText=1200005637

Since many new CPUs are fully supported by way of BIOS updates, I think it would be worth a shot.

 

[cbrunny]

First things that come to mind are BIOS updates and chipset drivers.

 

[Maximilian]

Why? Because the OEM went cheapskate?

You know, the same OEMs that forces nVidia and AMD to rebrand GPUs, just to try and sell the same thing twice.

Yes.

 

[Lakeshorecoupe]

Lakeshorecoupe, I don't know if you looked at this, but thought I would mention it:

According to the support page for this model, there are three BIOS updates since the original:

http://support.toshiba.com/support/modelHome?freeText=1200005637

Since many new CPUs are fully supported by way of BIOS updates, I think it would be worth a shot.

I did a toshiba automatic bios update when i did the cpu, didnt help

 

[Zodiark1593]

Zodiark, both chips are Ivy Bridge. Intel in all their wisdom has the Pentium chips a generation behind in numbering in order to make the products look more differentiated. Pentium 2000 is IVB, Pentium 3000 is HSW, and Pentium [3 digit] is SNB.

Damn, way to make me look dumb.


Well, you could try running a Linux Live USB/CD for awhile and see if it shuts off. If it doesn't, I'd probably blame a driver (chipset driver?). If your Linux Live session does shut down, then it likely is a BIOS/Mobo incompatibility. Not sure what you could do about the BIOS though since it's already updated.