In various global markets - DRAMeXchange is reporting that first quarter buying activity may be enough to bring the DRAM industry out of its projected slump. According to reports published today at DigiTimes, both Kingston and Samsung's believe the slump will continue beyond Q1 2008. However, DRAMeXchange believes the industry may rebound in Q1 2008 thanks to "annual facility maintenance" whereby large corporations upgrade their equipment.
According to the DRAMeXchange press release, while the 512 Mbit DRAM industry weakness persists at at about -1.2% this quarter. The 1 Gb segment was hit a little harder with a 7.1% decline. These declines are expected to persist through the remainder of the year, with many OEMs facing potential inventory pressures, which will keep them from buying. The only light at the end of the tunnel, even from DRAMeXchange's perspective, is the seasonal system upgrades. Without those, it appears DRAM prices will continue to decline, inventories will continue to increase, and DRAM makers will face greater and greater profit losses.
It seems very clear that the DRAM industry is in a type of depression. It's almost as if there is some great breakthrough that's needed to reshape that industry. Having the current division between DRAM and flash, for example, would no longer be required if non-volatile memory solutions like FeRAM or ReRAM could be developed and made commercially viable.
The dream of the instant-on computer, the one which has everything exactly as it was left when it was shut down, all windows open, files open, everything the same from a software point of view, with only the re-nogitation of Internet connections, wireless connections, etc., being required, would be made possible with such a solution.
It seems very clear that with the low profit margins of the DRAM industry, the continuous slumps, the gluts, etc., that such a non-volatile, high-speed memory solution, one which could replace hard disk storage and main memory with a single product, would be the holy grail. If I were a memory maker, every last ounce of profit I had would be going into that type of R&D. In fact, I think any self-respecting company would forego every ounce of profit in pursuit of that ultimate form of memory because the first one to mass produce it would simultaneously undo the DRAM, flash and hard drive markets, all at the same time. They'd have a patent on the future of media storage in all its forms. I think that's worth a little sacrificing between now and then.