"Cosmic Rays" make computers fail?

[klah]

Has anyone read this article: www.research.ibm.com/journal/rd/421/ziegler.html discussing the correlation between cosmic ray intensity at different geographic locations and the soft-fail rate of computer systems?

The last column, the relative cosmic ray intensity of the site, can be used to scale the soft-fail rate of a computer system from the datum, New York City. For example, Leadville, Colorado, in the United States will have a fail rate about 13x of the datum. This number has been verified experimentally at the IBM laboratory in Leadville

Does this article seem plausible and/or accurate ? Do i need a tin-foil hat for my PC?

 

[dkozloski]

I wouldn't get too excited about a completely random event that occurrs once in a coons age.

 

[Armitage]

Yes, but it's primarily of concern to space-based electronics.
Google on "Single Event Upset"

Take a look at this link as well: Single Event Upset
And this one: Landsat

 

[CTho9305]

Originally posted by: klah
Has anyone read this article: www.research.ibm.com/journal/rd/421/ziegler.html discussing the correlation between cosmic ray intensity at different geographic locations and the soft-fail rate of computer systems?

The last column, the relative cosmic ray intensity of the site, can be used to scale the soft-fail rate of a computer system from the datum, New York City. For example, Leadville, Colorado, in the United States will have a fail rate about 13x of the datum. This number has been verified experimentally at the IBM laboratory in Leadville

Does this article seem plausible and/or accurate ? Do i need a tin-foil hat for my PC?

Yup. That is why you use ECC memory in important systems - to fix the approximately one bit every six months that gets flipped in ram.

 

[globalstud]

Cosmic rays hitting your DNA can cause mutations as well. You better put on the tinfoil hat too.

 

[RossGr]

Why is this considered News, olds is more like it. This has been a know phenomena since the first solid state computers.

 

[SuperTool]

It used to be an afterthought, but now it's becoming a bigger concern in circuit design.
Because devices and capacitances are getting much smaller, there is less capacitance and it takes much less energy to switch a state of a memory cell, or to discharge a dynamic circuit.

 

[KenGr]

Wait a minute - stop and think. If computers failed 13 times as much in Colorado as in New York, wouldn't someone have noticed? The key here is that although the article may be correct, it only addresses failures due to cosmic radiation. Since computers have metal cases and are usually inside buildings, the radiation level is very small. Even if one out of a thousand computer failures in New York is caused by cosmic rays (which is probably a very high estimate, the failure rate in Colorado would only be 0.1% higher.

So, don't waste your time on the tinfoil (or lead lined underwear).