Random lantency in a PCM device is 17 times faster than Flash memory and over 3000 times faster than a hard drive. The biggest drawback in PCM is the price, which is about 10 times higher than DRAM at this point. The pricing, however, will come down over time and as fabricating processes become improved. PCM chips use the same material, chalcogenide, that is used inside to store data in rewritable optical disks. But instead of using a laser to change the properties of the material and thus create the zeros and ones that make up data, the chips use electricity that flows through a resistor. The resistor heats up and does the job of the laser, changing the materials’ properties to represent a zero or a one.
In Feb 2009, Numonyx had taped out a 1Gbit PCM die using 45nm process technology. Numonyx has a 32nm process apparently planned for 2010. Doller also said that IBM researchers had demonstrated workable PCM dies with a 5nm process.
Others promising new memory technology are CMOx, a multilevel scRAM (storage-class RAM) chip; STT RAM (spin-transfer torque RAM); TAS-MRAM (thermally assisted switching magnetic RAM); and Hewlett-Packard’s memristor.
CMOx—Startup Unity Semiconductor is pushing CMOx, a multilayer flash chip that promises four times the density and five to 10 times the write speed of today’s high-end NAND flash.
Darrell Rinerson, co-founder and CEO of Unity, said the company has developed a 64GB CMOx (metal oxide) chip and describes it as a “passive rewritable cross-point memory array” with no transistors in the memory cell. CMOx is next-generation nonvolatile memory based on a proprietary switching effect that occurs in certain metal-oxide combinations. The 64GB chips are scheduled for pilot production in late 2010, with volume production set for 2011.
TAS-MRAM—This is being developed by startup Crocus Technology. As one of the most promising “spintronics” applications, MRAM combines the advantages of high writing and reading speed, limitless endurance and nonvolatility. The integration of MRAM in FPGA (field-programmable gate array) allows the logic circuit to rapidly configure the algorithm, the routing and logic functions, and easily realize the dynamical reconfiguration and multicontext configuration. It is nonvolatile, faster than SRAM (static RAM), potentially cheap, and features low power consumption and a high integration level.
Field-Induced MRAM (Toggle MRAM)—Historically, Field-Induced MRAM is hard to scale and has stability and retention problems. But Crocus, which is also working on this, aims to solve these problems using a thermally activated magnetic latch called Thermally Assisted Switching. This allows each flash cell to retain memory value. It also apparently scales well.
STT-RAM—Grandis, another chip startup, is championing this one. Grandis claims its proprietary Spin-Transfer Torque RAM technology has all the characteristics of an ideal “universal memory” and represents a breakthrough over first-generation, field-switched MRAM technology. Also known as SpinRAM, STT-RAM’s synthesis of nonvolatility, fast read and write speed, unlimited endurance, and extendibility beyond the 45-nanometer semiconductor node provides significant advantages over conventional memory technologies and allows system designers to develop new products with high performance, low power consumption and low cost, according to Grandis.
HP’s Memristor—HP Senior Fellow and Director of Quantum Science Research Stan Williams, speaking at the conference, described the technology this way: “This is sort of the missing element of the processor puzzle. It takes its place alongside the resistor, capacitor and inductor in the chip. An ideal memristor is a passive two-terminal electronic device that is built to express only the property of memristance (just as a resistor expresses resistance and an inductor expresses inductance).” In summary, let’s just say a memristor makes an SSD act as if it’s on steroids