Would a $50+ billion market be of interest?
What is this market? It is the combined markets of DRAM, SRAM and Flash i.e. the three most common digital memories in electronic devices.
Is it really a $50 billion market? OK, I may actually be underestimating the size of the market opportunity, but let’s just say that it is big. Now let’s say you could replace these three memories with a single memory. Enter the so-called “universal” memory. Universal memory is a mythical beast. There have not been any confirmed sightings, but it is hypothesized that it would provide the critical performance metrics of the three incumbent memories. It has the large write endurance of DRAM, the speed of SRAM and is non-volatile like Flash. The quest for a universal memory is a saga that has not yet reached an endpoint.
But why bring this up now? The other week Cypress issued a Press Release detailing a new 4Mb FRAM. FRAM was one of the original three technologies touted as a universal memory contender. MRAM and PCRAM were the other two. Today, FRAM and MRAM largely fill “niche” markets that require large write endurance. I do not want to belittle the accomplishments of these technologies as they are serving a particular market well. The third contender, PCRAM, was the most hyped, had the largest backing, was the last to enter production and was the first to fail. What happened? I do not have the space and have not provided the background needed to flesh out why none of these three technologies became a universal memory. However, a few considerations lead to a short answer.
a changing playing field and new systems
The short answer is that all three entrants in the 1st wave implemented new material systems and the playing field changed. First, the use of a new materials system inevitably requires considerable R&D, where there are always bumps in the road and trade-offs between desired capabilities and commercial production. Second, at the same time the three incumbents continued to progress with the large R&D dollars of the semiconductor industry. To this end Flash has broken through several brick walls that were supposed to bring about the end of its life. Of particular note is the storage capacity or density of current Flash memory. This capacity has increased by a factor of roughly 250 from the 1st wave to now. This is now a particularly high hurdle for any new memory to approach, let alone overcome. This change in storage capacity for the non-volatile component (Flash) is a key change to the playing field.
So where are we today? There is a second wave of contenders weaving their way through development. Will there be a universal memory amongst these? It is too early to say, but they too have experienced delays and bumps along the road. At the end of the day the potential market for a universal technology is too large to ignore. It is this potential market that keeps ned interested in the developments within the space. Should there be a universal memory emerge it will likely cause considerable upheaval to paradigms within the semiconductor industry and ned wants to be there.