20 maxium insertions in new intel boards

[Zebo]

More Lame wins for Intel

We start with a CPU that still chews up about twice the power of its Hammer competitor to perform tasks. It will now sit on top of a socket practically designed to break. To significantly overclock it, you almost need a degree in cryogenics.

Could matters get worse?

Why, yes. Just in case there were any overclockers foolish enough to buy Prescotts left, Intel has decided it is time to implement a lock on FSB overclocking!

Go over 10%, and it crashes.

 

[Sonic587]

It won't matter much to people who buy Dells and have never even inserted a CPU once in their life.

The only scary possibility is that MB makers might decide you're not going to get to RMA a board because you broke a pin with your own stupidity. These new sockets are a ticking time bomb waiting to break for those inexperienced with PC building.

 

[Zebo]

Originally posted by: Sonic587
It won't matter much to people who buy Dells and have never even inserted a CPU once in their life.

The only scary possibility is that MB makers might decide you're not going to get to RMA a board because you broke a pin with your own stupidity. These new sockets are a ticking time bomb waiting to break for those inexperienced with PC building.

The worst part is all of us have swapped out processors with the mobo still in the case and wires dangling everywhere. I did just the other day when swapping HSF..had to pull out processor to get a better handle and light on it to clean off AS and re-apply..just wait even experianced assemblers will break them first time.

 

[Sonic587]

If these sockets turn out to be as fragile as they say, who wouldn't be sweating bullets if they mess with their CPU socket? I'm hoping they are covering their butts for idiots and it's not as bad as people say. However, who here slams their CPU in the socket all the time? Another concern I have is what if you need a new heatsink and fan? That means you have to take the CPU out to reapply the thermal paste. I don't like the fact people will be worrying over something like that

 

[hifisoftware]

If you going to be building new PC you should know not to use Intel. Intel does make it as clear as they can I think.

 

[oldfart]

Thugsrook would use that up in a week!

 

[Budman]

Originally posted by: oldfart
Thugsrook would use that up in a week!

I think he's considering going A64.

 

[RalfHutter]

Intel's actually released a video to it's OEM dealers that gives detailed instructions on how to install a CPU on the Socket 775 boards. I don't ever recall seeing anything like this for any of the prior types of platforms so they must be concerned enough about potential problems to show people how to use the correct technique.

 

[StrangerGuy]

It is going to hurt the resale value of the LGA775 boards big time.

 

[LTC8K6]

Didn't one of the reviewers claim to have inserted and removed an LGA775 Prescott 40 times without a problem? I am sure I read that around here somewhere.

Intel has also replied to the overclock lock. Anandtech's article is updated with the reply.

"Intel has done extensive evaluation of our 915G/P and 925X chipset products and have designed them to run robustly at 800FSB. Any use or operation of these chipsets beyond their specifications, including overclocking, has not been evaluated or validated by Intel. If a board provider decides to overclock their platform, they do so at their own risk, since this action will void the warranty for the Intel products. Additionally, from time to time, Intel may choose to implement functionality that helps to ensure that the Intel product experience is not degraded by those who may try to run our products out of spec."

 

[LTC8K6]

The Intel spec sheet says it must pass a test which is 20 insertion cycles. There is no limit on the number of cycles.

Here is the quote they are misrepresenting:

The LGA775 Socket must withstand at least 20 insertion cycles (i.e., the package is removed at
the end of each de-actuation cycle and reinserted into the socket). Test is done per EIA-364, test
procedure 09. Contact resistance is measured when mated in 1st and 20th cycles.

Data sheet

 

[Zebo]

Ya right. Lock is still there, only a partial work-around adding 5% to 10%. But hey combine that with intels 10%, and 20% is all on could hope to get from the prescott without Dry ICE or liquid nitrogen anyhow. So not as bad as feared.

As far as the socket, the way Intel and other corps run thier PR releases though a battery of lawyers and PR firms to offset any liabity or negtivity, if Intel says "20 ", it's an extemly delicate proposition no doubt. Does it matter? It's cheap. It's designed to fail. It gives mobos resale zero value when ever customer fears socket comprimse.

 

[LTC8K6]

I am sure the sockets will be quite up to the task.

What does saying "you have to test it 20 times" have to do with resale value?

 

[LTC8K6]

Better not cycle that socket 940 too many times!

Seriously, these are just durability tests and aren't anything to make a big deal over.

Initial actuation/de-actuation cycles must be made by exercising five cycles per µPGA package. The final five cycles must be mated to the actual µPGA package test device.
Group 2 50 cycles per socket.
Group 5 5 cycles per socket.

 

[Mingon]

Topic Summary: Imagine your girlfriend telling you that.

So after 20 insertion's you have to get a new model, wheres the disadvantage there?

 

[jhu]

Originally posted by: Mingon

Topic Summary: Imagine your girlfriend telling you that.

So after 20 insertion's you have to get a new model, wheres the disadvantage there?

no, the disadvantage is that your pin breaks after 20 insertions.

 

[Zebo]

Yup yup