Archived 2025-06-15 at the Wayback Machine > 자유게시판

Usually, ECC memory maintains a memory system immune to single-bit errors: the info that's learn from each word is at all times the identical as the information that had been written to it, even when one of the bits actually saved has been flipped to the flawed state. Most non-ECC memory cannot detect errors, Memory Wave although some non-ECC memory with parity help permits detection however not correction. ECC memory is utilized in most computers where knowledge corruption can't be tolerated, like industrial management applications, essential databases, and infrastructural memory caches. Error correction codes protect against undetected data corruption and are used in computers where such corruption is unacceptable, examples being scientific and financial computing functions, or in database and file servers. ECC can even scale back the number of crashes in multi-consumer server purposes and maximum-availability methods. Electrical or magnetic interference inside a pc system can cause a single little bit of dynamic random-access memory (DRAM) to spontaneously flip to the other state.

It was initially thought that this was mainly resulting from alpha particles emitted by contaminants in chip packaging materials, however research has shown that almost all of 1-off smooth errors in DRAM chips happen because of background radiation, mainly neutrons from cosmic ray secondaries, which can change the contents of one or more memory cells or interfere with the circuitry used to learn or write to them. Therefore, the error rates improve quickly with rising altitude; for example, in comparison with sea stage, the speed of neutron flux is 3.5 times higher at 1.5 km and 300 instances increased at 10-12 km (the cruising altitude of business airplanes). As a result, methods operating at high altitudes require special provisions for reliability. For example, the spacecraft Cassini-Huygens, launched in 1997, contained two identical flight recorders, every with 2.5 gigabits of memory in the form of arrays of commercial DRAM chips. As a result of built-in EDAC functionality, the spacecraft's engineering telemetry reported the number of (correctable) single-bit-per-word errors and (uncorrectable) double-bit-per-phrase errors.

During the first 2.5 years of flight, the spacecraft reported a nearly constant single-bit error price of about 280 errors per day. Nevertheless, on November 6, 1997, throughout the primary month in area, the number of errors increased by greater than an element of four on that single day. There was some concern that as DRAM density will increase further, and thus the components on chips get smaller, whereas operating voltages proceed to fall, DRAM chips shall be affected by such radiation extra incessantly, since lower-power particles will likely be ready to change a memory cell's state. On the other hand, smaller cells make smaller targets, and strikes to technologies comparable to SOI could make particular person cells much less inclined and so counteract, and even reverse, this development. Work printed between 2007 and 2009 confirmed extensively various error rates with over 7 orders of magnitude distinction, starting from 10−10 error/(bit·h), roughly one bit error per hour per gigabyte of memory, to 10−17 error/(bit·h), roughly one bit error per millennium per gigabyte of memory.

A big-scale examine primarily based on Google's very large number of servers was presented at the SIGMETRICS/Performance '09 convention. The precise error price discovered was several orders of magnitude larger than the previous small-scale or laboratory research, with between 25,000 (2.5×10−11 error/(bit·h)) and 70,000 (7.0×10−11 error/(bit·h), or 1 bit error per gigabyte of RAM per 1.8 hours) errors per billion device hours per megabit. Greater than 8% of DIMM memory modules had been affected by errors per yr. The consequence of a memory error is system-dependent. In systems without ECC, an error can lead both to a crash or to corruption of data; in massive-scale manufacturing websites, memory errors are one of the crucial-widespread hardware causes of machine crashes. Memory Wave Routine errors may cause safety vulnerabilities. A memory error can haven't any penalties if it changes a bit which neither causes observable malfunctioning nor impacts information utilized in calculations or saved. A 2010 simulation study confirmed that, for an internet browser, only a small fraction of memory errors prompted information corruption, although, as many memory errors are intermittent and correlated, the consequences of memory errors were better than can be anticipated for independent gentle errors.