There was a flurry of news abut memristors maybe 6-9 months ago, then, for the last few months, now news at all. Then a few days ago, there was news from Unity Semiconductor about their memristor device. Now, a few days later, this news from HP/Hynix. I hope it's all true, but it's hard to know. What I don't understand is that Hynix and Toshiba (Samsung?) recently stated http://www.eetimes.com/electronics-news/4217803/Hynix-Toshiba-MRAM that STT-MRAM was the future and didn't even mention memristors. What's going on here?
A little perspective to early semiconductor memory history provides that there needs to be a need from a customer for a new technology. Is there a compelling one for the memristor?
From TI's early days chronicles: "In the early 1970s, a customer approached TI to develop a 1-K MOS memory, but the company elected not to pursue the contract because of insufficient design resources. Memory activities began in earnest when TI second-sourced Intel's 1K-DRAM. Later, TI began a conversion from metal gate to silicon gate designs. In 1972, a decision had to be made to predict the next generation of DRAM, either 2K or 4K. TI started a 4K program, which turned out to be the right choice. The 4-K DRAM became a significant product in the industry.
By 1974, TI had 4K DRAMs (both 18- and 22-pin versions) available."
Lee Harrison: Really high-end servers use SRAM and capacitors (and/or batteries). The basic physics was sound for PCM/PRAM as well, initially, until it turned out that the basic physics was actually quite misunderstood and wrong. HP and Hynix have neither the resources nor the management attention to produce a new chip within 18 months. Period. Samsung, who actually may have a better patent portfolio for the memristor, have learned their lesson after the PRAM fiasco. IBM is finished - just look at their millipede storage (those are the same people who are "working" on PCM/PRAM after that disaster).
So, no IBM is not rolling out any memristor cache.
"However, in 2008 researchers from HP published a paper in Nature that tied the hysterical I-V characteristics of two-terminal titanium oxide devices to the memristor prediction of Chua."
It's hysterical, all right.
Kudos to HP Labs and Stan for getting to a point where a product is on the--albeit distant--horizon. Clearly there are many pitfalls between now and 2013 they will have to avoid, but this is (literally) promising.