Parity Checking

One standard IBM set for the industry is that the memory chips in a bank of nine each handle 1 bit of data: 8 bits per character plus 1 extra bit called the parity bit. The parity bit enables memory-control circuitry to keep tabs on the other 8 bits—a built-in cross-check for the integrity of each byte in the system.

If the circuitry detects an error, the computer stops and displays a message informing you of the malfunction. If you are running a GUI operating system, such as Windows or OS/2, a parity error generally manifests itself as a locked system. When you reboot, the BIOS should detect the error and display the appropriate error message.

SIMMs and DIMMs are available both with and without parity bits. Originally, all PC systems used parity-checked memory to ensure accuracy. Starting in 1994, a disturbing trend developed in the PC-compatible marketplace. Most vendors began shipping systems without parity checking or any other means of detecting or correcting errors!

These systems can use cheaper nonparity SIMMs, which saves about 10%–15% on memory costs for a system. Parity memory results in increased initial system cost, primarily because of the additional memory bits involved.

Parity can't correct system errors, but because parity can detect errors, it can make the user aware of memory errors when they happen. This has two basic benefits:

  • Parity guards against the consequences of faulty calculations based on incorrect data.

  • Parity pinpoints the source of errors, which helps with problem resolution, thus improving system serviceability.

PC systems can easily be designed to function using either parity or nonparity memory. The cost of implementing parity as an option on a motherboard is virtually nothing; the only cost is in actually purchasing the parity SIMMs or DIMMs.

This enables a system manufacturer to offer its system purchasers the choice of parity if the purchasers feel the additional cost is justified for their particular applications.

Unfortunately, several of the big names began selling systems without parity to reduce their prices, and they did not make it well known that the lower cost meant parity memory was no longer included as standard.

This began happening mostly in 1994 and 1995, and it has continued until recently, with few people understanding the full implications. After one or two major vendors did this, most of the others were forced to follow to remain price-competitive.

Because nobody wanted to announce this information, it remained sort of a dirty little secret within the industry. Originally, when this happened you could still specify parity memory when you ordered a system, even though the default configurations no longer included it.

There was a 10%–15% surcharge on the memory, but those who wanted reliable, trustworthy systems could at least get them, provided they knew to ask, of course. Then a major bomb hit the industry, in the form of the Intel Triton 430FX Pentium chipset, which was the first major chipset on the market that did not support parity checking at all!

It also became the most popular chipset of its time and was found in practically all Pentium motherboards sold in the 1995 timeframe. This set a disturbing trend for the next few years. All but one of Intel's Pentium processor chipsets after the 430FX did not support parity-checked memory; the only one that did was the 430HX Triton II.

Since then, Intel and other chipset manufacturers have put support for parity and ECC memory in most of their chipsets (especially so in their higher-end models). The low-end chipsets, however, typically do lack support for either parity or ECC. If more reliability is important to you, make sure the systems you purchase have this support.