Samsung, and Micron Technologies Inc. through its acquisition of Numonyx NV, are the two companies that have got close to offering non-volatile phase change memory for commercial use. Micron is offering serial and parallel versions of a 128-Mbit PCM. However, there are almost no reports of phase-change memories in the field.
"Alongside STT-MRAM and ReRAM currently under joint development with Toshiba and HP respectively, PCRAM will enrich our portfolio of next-generation memory technologies. SK Hynix will continue to endeavor to seek possible partnerships that will elevate our competence in the ever evolving semiconductor industry," said Hyun Jong Song, senior vice president and head of the future strategy division at SK Hynix, in a statement.
"Phase-change memory technology has the potential to enable a new class of low-cost, high-performance memory technologies for consumer devices, cloud computing, data storage and other enterprise applications," said T.C. Chen, IBM Fellow and vice president of science and technology for IBM Research, in a statement issued by SK Hynix. "Working with SK Hynix will speed the development and production of PCRAM devices based on our breakthrough multi-bit, phase-change memory technology," he added.
"PCRAM may be able to reshape the landscape of the memory industry by introducing storage-class memory (SCM), a promising next generation memory class, designed to boost performance and reduce power consumption for enterprise servers. PCRAM will bridge a gap between the current DRAM and solid-state drives (SSDs) as it takes the role of a buffer memory," SK Hynix said in the statement.
The memory opportunity for servers, which includes DRAM and SSD, will expand from $8 billion in 2012 to $16 billion in 2016, and the SCM opportunity at that time, driven by demand for PCRAM from server makers, will be worth $1.4 billion and continue to grow for years thereafter, SK Hynix said referencing market researcher Gartner Inc. as its source.
One thing to note is that while Hynix's have numerous patents related to phase change memory they have only one ReRAM/memristor patent (for an application in RFID). Thus all else being equal there is more economic incentive for Hynix to develop phase change memory. Below is a link to comparative US patent data for companies invested in memory resistors vs. phase change memory.
The way SK Hynix approaches these memories, by sharing risk with big names like Toshiba for STT-MRAM, HP for memristor and IBM for phase change memory, indicates they are not really into using these in the near future. So it's funny to watch the big-name partner get all excited and make optimistic projections.