HP and Hynix tout ReRAM as Flash memory replacement
Hewlett Packard (HP) and Hynix are teaming up to promote the use of ReRAM in future electronic devices. The next-generation chip - which is based on memristor technology - could eventually replace flash memory used in today's mobile phones and MP3 players.
Indeed, ReRAM is a natural choice for mobile devices due to low power consumption and non-volatile memory that is capable of fulfilling the traditional roles of flash, DRAM and even hard drives.
The versatile Memristor-based modules are also faster than current solid-state storage technologies, can perform complex logic tasks and retain information even when power is off.
"Memristors represent a fourth basic passive circuit element," said Dr. Stanley Williams, HP Senior Fellow and IQSL founding Director.
"They existed only in theory until 2006 - when researchers in HP Labs' Information and Quantum Systems Laboratory (IQSL) first intentionally demonstrated their existence."
According to Williams, the agreement between the two industry heavyweights will allow the companies to jointly develop and advance both memristor technology as well as Resistive Random Access Memory (ReRAM) modules.
"People have been attempting to make resistive memory for a long time. But because they didn't understand that the devices they had were memristors, they weren't making good progress.
"Once you understand the mathematical framework for memristors, you can design circuits that perform the way they are intended to perform."
And that understanding, along with deep theoretical knowledge and practical experimental expertise, are part of what HP is sharing with Hynix, explained Williams.
"It's not just the memristor...There's architecture, circuit design, error correction coding - we're bringing the complete package.”
Williams also noted that memristor ReRAM chips can be fabbed using existing semiconductor manufacturing processes.
"[As such], the HP team is confident that such chips will be easily swappable for flash memory – so products won't need to be redesigned before they can enjoy the benefits in speed and and power use that memristor chips promise to provide.
“[Now], we began this work in 2006 and we’re aspiring to have a product ready by 2013. [Yes], that's only seven years. So we're compressing a normal R&D timetable by a factor of two."