Phase change memory is a potential replacement for flash memory.  Intel and STMicroelectronics are rumored to be preparing to roll out the new memory technology as early as this year. Samsung has announced that their phase change memory will be ready in 2008. It is time to be acquainted with the benefits and drawbacks of the memory system that we are going to begin to see in our devices in the not too distant future.

Phase change memory is not a new concept. It was first explored in 1960s by a Lithuanian-American scientist by the name of Stanford R. Ovshinsky. Ovshinsky was a self-taught engineer, inventor, and physicist who invented amorphous semi-conductor materials. Gordon Moore, the co-founder of Intel wrote an article outlining the promise of the new technology, in the September 1970 issue of “Electronics.” There were a number of engineering difficulties to do with the technology, but it would seem that they have been overcome.

According to Wikipedia:

Phase-change memory (also known as PCM, PRAM, PCRAM, Ovonic Unified Memory and Chalcogenide RAM C-RAM) is a type of non-volatile computer memory. PRAM uses the unique behavior of chalcogenide glass, which can be “switched” between two states, crystalline and amorphous, with the application of heat. PRAM is one of a number of new memory technologies that are attempting to compete in the non-volatile role with the almost universal Flash memory, which has a number of practical problems these replacements hope to address.

If you thought that flash memory was fast, the new phase change memory is about 100,000 faster. The write speed has gone from 1ms to 10ns per byte. Not only is it faster, it is also more durable. Each time data is written to flash memory, the process will cause damage and degradation. Most flash devices will be able to withstand 10,000 – 100,000 writes per sector. Phase change memory on the other hand can withstand 100,000,000 writes per sector. It can also retain data without it becoming corrupted for much longer; estimates range from years to a decade, but those are conservative estimates.

Compared to the advantages of phase change memory, the disadvantages are minor. Perhaps the biggest disadvantage is that due to the high temperatures involved in the manufacturing process, it is impossible to solder pre-programmed phase change memory chips onto a board. The chips must be programmed after they have been soldered into place. Some might also consider that higher voltages are also required to write data to the phase change memory to be a disadvantage.

Phase change memory offers us faster and more durable memory. It could be ready for sale by the end of this year, though chances are that many manufacturers will wait a little before joining the cause. MP3 players and the like will probably continue to use flash memory for the time being. I  Assuming that Intel manage to put their phase change memory to market within the next few months and Samsung come in early next year, we can expect the technology to proliferate as the older flash chips get used up.  Until then, the build up is likely to be gradual.

Price: Unknown, but Samsung is going to begin with a 512 Megabyte unit.

Sources: Wikipedia, Samsung via Mobilewhack, CNet