Jump to content

Talk:Soft error

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

1

[edit]

The table at the bottom listing decade of error-correction method introduction is misleading re LDPC codes; while invented/discovered in the 1960s it was not until the 1990s and their rediscovery that technology caught up such that an LDPC system could actually be implemented. Also, LDPC codes and Turbo codes aren't used for soft error detection/correction in the sense that this article implies. I think the table could be removed entirely and nothing be lost from the article.

example of a 1 bit error in memory causing a segfault

[edit]

maybe this would be a good reference. its a very detailed analysis of a segfault, that turned out to be a 1 bit error in the memory caching a binary file. http://blog.ksplice.com/2010/06/attack-of-the-cosmic-rays/ --194.36.2.117 (talk) 14:52, 19 April 2011 (UTC)[reply]

hard vs soft

[edit]

Is there such a thing as a "hard" error? If so, it would be very natural to mention it, and contrast hard- and soft-errors. For example, the article on Software does something similar:

The term (Software) was coined to contrast to the old term hardware (meaning physical devices). In contrast to hardware, software is intangible,...

I think this would strengthen the article. Lavaka (talk) 18:10, 2 August 2011 (UTC)[reply]

radioactive material

[edit]

This can't be right: "take a typical computer with a large memory capacity at least 10 years before the radioactive elements of the chip's materials begin to decay." Radioactive materials don't "begin" to decay; they undergo Exponential decay. Unfortunately I'm not sure what the correct statement is in the context. Is there an expert in the house? ALloydFlanagan (talk) 20:42, 22 September 2012 (UTC)[reply]

I'm glad I'm not the only one to notice that; it's just plain wrong. Unfortunately I'm not an expert on memory hardware either. I've removed the problematic sentence. Lower down in the article things get more technical, and it's annoying to have something obviously wrong in the introduction. Wyvern (talk) 15:45, 4 April 2013 (UTC)[reply]

Some inaccuracies?

[edit]

A soft error will not damage a system's hardware; the only damage is to the data that is being processed. What if a soft error triggers a killer poke? That's (albeit it very rare) a possibility!

As for Soft errors typically can be remedied by cold booting the computer.. Maybe someone better qualified should add unless corrupt data was permanently saved?

2A02:A453:A18F:1:E59A:11B4:2069:47B4 (talk) 00:24, 13 February 2018 (UTC)[reply]

Alpha particles from package decay

[edit]

The reference https://ieeexplore.ieee.org/document/5722063/ might support "perhaps lasting a fraction of a nanosecond" but papers like https://ieeexplore.ieee.org/document/6068271/ indicate that alpha particles can have significant impact on combinatorial circuits. Pulse stretching, also called “propagation-induced pulse broadening” (PIPB) could explain this change in understanding, see Ferlet-Cavrois V, Paillet P, Mcmorrow D, et al. New insights into single event transient propagation in chains of inverters—evidence for propagation-induced pulse broadening. IEEE T Nuc Sci, 2007, 54: 2338–2346 — Preceding unsigned comment added by Mark56435 (talkcontribs) 07:11, 6 September 2018 (UTC)[reply]