Non Volatile Memory, NVM, NOR Flash, Cypress Spansion Merger, Ravichander Rao
I was studying the NOR flash market during early December, 2014 with an intention of understanding its current trends and future roadmap. As I finished gathering the data and understanding the underlying reasons, Spansion and Cypress merged. So the timing is right to present an analysis why these companies merged.
Cypress Semiconductor was making loss for the past two calendar years 2012 and 2013 because of the decline in the SRAM market, which is shrinking at CAGR of 13% since 2001 (~$500 million is the market share at 2014E).
On the other hand, Spansion, the market leader in the NOR Flash technology, was unable to keep up with the competition from NAND Flash. Consequently, its market share was shrinking and made loss in 2013, 2011 and 2010.
What made the two loss making companies merge together? Why do they believe that their losses will not compound after the merger?
Integrated Chip (IC) market share is expected to grow at a 25% CAGR between 2012 to 2017 in the mobile segment, which has been the driver for the semiconductor technology. The standard PC demand is expected to grow at 20%, but the growth driver for it is mainly the storage capacity, which is served by Hard Disk Drives or SSD (Solid State Devices / NAND memory). The rest are low growth and low revenue markets for the Flash memories.
As a percentage of the revenue, cellphone represents 52%. The rest are much lower in share.
The NOR flash was earlier used in the cellphone for memory applications. In there are a few emerging technologies available including NAND.
But with the advancement in the NAND technology (cost per GB dropping) and changes in the smart phone requirements - the increase the memory requirements, the cellphone market the landscape for the NOR flash changed dramatically. NAND could offer similar level of data transfer speed at a lower cost. This enabled the programmers / cellphone makers to switch to NAND. As the NOR flash lost to NAND Flash in the cellphone market, it deeply affected all the NOR flash manufacturers.
The SRAM was always a costlier solution, and was used for a booting application in the cellphone in conjunction with the NOR. The cost for operating a booting program was lowered down by SDRAM which mimics SRAM along with SDRAM and NAND combination.
In this segment, both Cypress (loss to touch screen and declining SRAM market share) and Spansion (loss to parallel NOR and NAND) were losing out the to competition. These products are the core markets for these two companies. So, it was an existential threat. NOR though is manufactured by Micron, another leading Memory device manufacturer, does not impact Micron because of their diversified memory portfolio - NOR, NAND, PCM, etc.
Further, NAND expects to take over NOR flash applications by offering eMMC solution in all of the growth segments of the IC market.
The final nail in the coffin for the NOR flash was cannibalization of Serial NOR flash, which is a costlier but reliable solution by Parallel NOR flash, which is comparatively less costlier solution, and pricing pressure from the Parallel NOR manufacturers in China (GigaDevice and XMC) and Taiwan (Winbond and Macronix).
Where does NOR Flash go from here?
Given the market condition and weakness of the NOR Flash, there are a few options left:
a) Collaborate with MLC NAND for Enterprise solution
b) Phase out SPI NOR
c) Penetrate markets and increase adoption in automotive, medical, IOT, big data, etc.
d) Partner with the programmers to switch to NOR from NAND
e) Increase differentiation / embark on campaigning to highlight the differentiation
Of the options a) to e), option c) is promising. However, this does not allow either NOR or SRAM to come back and compete with SDRAM and NAND in cellphone market. Hence, NOR and SRAM leaders needed to merge.
What to expect of this merger?
a) SPI NOR getting phased out.
b) New solutions such that SRAM & Parallel NOR are on a single chip. May be costlier but easier to handle as 3rd interface is eliminated.
c) More campaign during Flash Memory summits
d) More investments from Cypress in Big Data, Speech Recognition technology, Autonomous Vehicles, etc.
I was studying the NOR flash market during early December, 2014 with an intention of understanding its current trends and future roadmap. As I finished gathering the data and understanding the underlying reasons, Spansion and Cypress merged. So the timing is right to present an analysis why these companies merged.
Cypress Semiconductor was making loss for the past two calendar years 2012 and 2013 because of the decline in the SRAM market, which is shrinking at CAGR of 13% since 2001 (~$500 million is the market share at 2014E).
What made the two loss making companies merge together? Why do they believe that their losses will not compound after the merger?
Integrated Chip (IC) market share is expected to grow at a 25% CAGR between 2012 to 2017 in the mobile segment, which has been the driver for the semiconductor technology. The standard PC demand is expected to grow at 20%, but the growth driver for it is mainly the storage capacity, which is served by Hard Disk Drives or SSD (Solid State Devices / NAND memory). The rest are low growth and low revenue markets for the Flash memories.
As a percentage of the revenue, cellphone represents 52%. The rest are much lower in share.
The NOR flash was earlier used in the cellphone for memory applications. In there are a few emerging technologies available including NAND.
But with the advancement in the NAND technology (cost per GB dropping) and changes in the smart phone requirements - the increase the memory requirements, the cellphone market the landscape for the NOR flash changed dramatically. NAND could offer similar level of data transfer speed at a lower cost. This enabled the programmers / cellphone makers to switch to NAND. As the NOR flash lost to NAND Flash in the cellphone market, it deeply affected all the NOR flash manufacturers.
The SRAM was always a costlier solution, and was used for a booting application in the cellphone in conjunction with the NOR. The cost for operating a booting program was lowered down by SDRAM which mimics SRAM along with SDRAM and NAND combination.
In this segment, both Cypress (loss to touch screen and declining SRAM market share) and Spansion (loss to parallel NOR and NAND) were losing out the to competition. These products are the core markets for these two companies. So, it was an existential threat. NOR though is manufactured by Micron, another leading Memory device manufacturer, does not impact Micron because of their diversified memory portfolio - NOR, NAND, PCM, etc.
Further, NAND expects to take over NOR flash applications by offering eMMC solution in all of the growth segments of the IC market.
The final nail in the coffin for the NOR flash was cannibalization of Serial NOR flash, which is a costlier but reliable solution by Parallel NOR flash, which is comparatively less costlier solution, and pricing pressure from the Parallel NOR manufacturers in China (GigaDevice and XMC) and Taiwan (Winbond and Macronix).
Where does NOR Flash go from here?
Given the market condition and weakness of the NOR Flash, there are a few options left:
a) Collaborate with MLC NAND for Enterprise solution
b) Phase out SPI NOR
c) Penetrate markets and increase adoption in automotive, medical, IOT, big data, etc.
d) Partner with the programmers to switch to NOR from NAND
e) Increase differentiation / embark on campaigning to highlight the differentiation
Of the options a) to e), option c) is promising. However, this does not allow either NOR or SRAM to come back and compete with SDRAM and NAND in cellphone market. Hence, NOR and SRAM leaders needed to merge.
What to expect of this merger?
a) SPI NOR getting phased out.
b) New solutions such that SRAM & Parallel NOR are on a single chip. May be costlier but easier to handle as 3rd interface is eliminated.
c) More campaign during Flash Memory summits
d) More investments from Cypress in Big Data, Speech Recognition technology, Autonomous Vehicles, etc.
2 comments:
nice analysis! If you had to update your analysis for early 2017 what would be your conclusions? Thx Chitra
Hi Chitra,
Cypress is still making losses, while NAND manufacturers are making significant profits. This is tied to Cypress focusing on applications that are cost sensitive and markets having flat growth such as - home appliances, electronic connectors, etc. Growth drivers for the semiconductors today are data storage (cloud), premium end phones offering long battery life and autonomous vehicles. Things can be turned around, if NOR focuses on its strength - speed (XIM) driven applications.
Post a Comment