Top 10 Forecasts for '10
I’ve collected several of my recent forecasts and added a few more to round the total forecasts out to compete my top 10 forecasts for 2010.
1. Numonyx recently announced that samples of 1G PCM would be available in 1Q10 with volume production to follow later in the year. This game-changing event is sure to trigger several responses regardless of how well Numonyx executes on their announced schedule. My first prediction is that the company will indeed meet their announced schedules.
2. Samsung previously held the bragging rights to the highest density PCM with their 512M, and samples were first announced in 2006. I expect that Samsung will treat Numonyx’ announcement very seriously and will make their own 1G samples available by mid-year.
3. I also expect that a creditable sample of an embedded non-volatile memory technology will be available in 2010. To be more specific, I expect that parts from Adesto will be validated by mid-year.
4. Samsung and Numonyx already have an agreement in place regarding the development of a common interface. I predict that the combination of both companies with pin and interface compatible 1G PCM parts combined with a creditable sample of a high-performance embedded memory technology will stimulate a new wave of investments in memory technologies.
5. Hynix and Samsung are partners in the development of STT, however I suspect that Hynix places a higher value on their Grandis relationship than they do on their Samsung partnership in this case. I anticipate that Numonyx’ PCM announcement will spur greater development efforts of STT and that Hynix will announce STT samples by the end of 2010.
6. The Hynix/Samsung/STT program sponsored by the Korean government was in response to other perceived market challenges. That lever works in both directions. I expect an aggressive program announcement for new memory technologies from Elpida in 2010.
7. Everspin’s MRAM still remains the highest revenue generator of all new memory technologies. I predict that Everspin’s revenue in 2010 will be double that of 2009 based on further penetration into industrial applications.
8. Ramtron’s FRAM technology has been unsuccessful in finding a high-volume and high-revenue application. I predict that Ramtron’s MaxArias Wireless Memory will earn several industry awards and will herald the arrival of a completely new class of hybrid-technology applications.
9. IBM’s two articles in July/September 2008 edition of the IBM Journal of Research and Development served as the introduction of enterprise-level storage-class memory and were also one of the most thorough evaluations at that time of emerging memory technologies. In those two articles, PCM was selected as the most likely candidate to be used in enterprise-class storage systems, while a second resistive memory technology was found to be similar in performance but less well understood. Due to the time that has passed and the amount of research that has continued, I predict that another resistive memory other than PCM will gain attention in 2010.
10. What do our friends at HP intend to do with Memristor? Is it both truly revolutionary and commercially viable? Can it compete as a memory technology or will it be used as a crossbar communications switch? If it is viable as a memory technology, does HP really intend to commercialize that technology and participate in the memory dogfight or spin it out and watch from the sidelines? I predict that at the end of 2010, I still won’t know the answer to those questions.
Best wishes to all in the New Year.
Currently no items