It’s done! The painful and slow sinking of Atmel seems to be over now.
We knew for a long time that thing do not go well with Atmel due to their poor management. The Arduino wave kept them for a while above the water level, but in September 2015 they announced that are about to sign with the world unknown company Dialog Semiconductor deal for $4.8 billions mostly with shares exchange, but as the time pass the shares of Dialog Semiconductor went down and this deal was looking not so attractive as before, and Microchip offer for $3.56 billions in shares become more attractive!
Microchip in other hand continue to expand and never have been better – they are buying company after company and already have the IC portfolio of SMSC, MICREL, SST, Novocell etc etc and what is better, each time they buy company they improve the availability and make these chips more easier to buy and deal with. This with the nice application notes and support is the Microchip receipt for success.
Our experience with SMSC and MICREL was that these companies were working with just big customers and smaller companies couldn’t buy directly, while Microchip sales channels can satisfy both big and small customers. After Microchip bought these companies, they improved the availability and made these chips easier to buy for the small customers. Atmel is in the same state – probably this was the major reason to sink, although Atmel have similar products like Microchip and even better open source software support, they sales are terrible hard to deal with. Many components prices go unexpected up and down as Atmel production capabilities are humble, once some big customer place large order for one chip they stop making others and this make impossible to use them for serious projects. Once you put AVR in your product it is not unlikely these chips suddenly to go on allocation due to the poor management and planning Altmel has, something which (almost) never happen to Microchip.
I guess Microchip will not cut AVRs supply but it’s unlickely they will keep developing this line when they put so much efforts in the PICs, to keep duplicate development teams for similar products is not practically. More probably is they to invest in ARM line expansion as this is something they missed yet.
For years Microchip top management was like mule on bridge not wanting to step ahead🙂 They were refusing to buy ARM licensee and bet on MIPS and they were missing a lot of sale opportunities with this odd decision. Whatever they do with PIC32 it’s not so successful like the STM32s and LPCs and they miss sales for millions $$$. This is not because MIPS architecture is bad, quite opposite it’s well developed in networking devices, but MIPS Soc from Mediatek running Linux at 400Mhz cost $2 while Microchip sells MIPS PIC32 with no MMU running at 80Mhz for $5-6.
Now with Atmel they got lot of ARM licensees and with their efficient manufacturing and great sale channels this give them access to the ARM sale market and other companies got strong competitor.