News & Analysis
Numonyx-Samsung PCM deal: What experts are saying
Mark LaPedus
6/24/2009 1:07 PM EDT
Common specifications between the two companies will be completed this year, with both companies expecting to have compliant devices available in 2010. The move will possibly bring PCM in the cell phone.
Numonyx and Samsung are the two leading companies researching PCM, a next-generation memory. So will PCM finally take off after years of R&D? Here's the reaction from analysts, memory gurus and competitive vendors about the Numonyx-Samsung deal:
Jim Cantore, president of consulting firm JLC Associates
''The (Samsung-Numonyx deal) could be the key for driving the legitimacy in phase-change memory. (PCM) is optimized to replace NOR. (But before PCM becomes a reality), you really need to see the densities of PCM in the 16- and 32-gigabit range. They have a ways to go.''
-----------------------------------------
Mark Gogolewski, chief technology officer of IP provider Denali Software Inc.
''PCM appears to be on deck. PCM is moving from 'R' to 'D.' It looks like a NOR replacement.''
------------------------------------------
Alan Niebel, CEO of WebFeet Research
''(The Samsung-Numonyx deal) is good for PCM. It's a beginning for PCM in the cell phone. It doesn't do much for storage-class memories in the server.''
------------------------------------------
Kurt Pfluger, chief executive of RRAM maker 4DS Inc.
''I don't wish to comment on PCM, but reader comments are insightful.'' (He did not elaborate, but one reader called PCM a sham.
--------------------------------------------
Darrell Rinerson, chairman, president and CEO of Unity Semiconductor Corp.
''PCM is not scalable. It can't be scaled to 20-nm because of the high current.''
---------------------------------------------
Farhad Tabrizi, president and CEO of Grandis Inc.
''Lack of common standards and specifications can hinder the adoption of exciting new memory technologies. Samsung and Numonyx are following the right approach by working together to develop package specifications and to ensure pin-for-pin compatibility. It will speed the adoption of PCM, which has the potential to offer better performance than NAND and NOR flash in the marketplace.
Grandis is taking the same approach with STT-RAM, which has not just the non-volatility of flash and PCM but also the high speed and endurance of DRAM and SRAM. Since STT-RAM can be implemented in both standalone and embedded solutions, we are working with semiconductor houses, OEMs and industry groups to define standards and specifications for STT-RAM across multiple market segments. Our licensees are expecting to begin sampling STT-RAM chips during the second half of 2010.
PCM is known to have very low write speed and very low yield. As a result, we believe PCM will have hard time to compete with STT-RAM as the next generation disruptive 'universal memory.'
|
|
|
|
|
Other John
6/26/2009 4:36 AM EDT
You gotta be kidding. Numonyx laid off the entire research lab in California and along with it with a thousand man-years expertise in PCM. Intel closed the FAB where a decade of PCM research had created wonderfully advanced technologies that have just been flushed away. Where do they think new technology comes from, heaven? Who is surprised that this product will be late if ever? When will management accept blame for gross mis-management?
Sign in to Reply
Other John
6/26/2009 4:37 AM EDT
You gotta be kidding. They laid off the entire research lab in California along with a thousand man-years expertise in PCM. Where do they think new technology comes from, heaven? Who is surprised that this product will be late if ever?
Sign in to Reply