Designers can never have enough MCUs, according to Lucio Di Jasio

4 min read

Towards the end of 2015, Atmel employees were contemplating life as part of Dialog Semiconductor, but a counter offer from Microchip was successful and the deal – the company’s seventeenth and biggest so far – completed in April 2016.

The acquisition saw a significant expansion to Microchip’s already broad MCU portfolio. But where many expected rationalisation, the reverse has happened; Microchip has started to add more ‘Atmel’ parts to its line up. The message is ‘you can’t have enough MCUs’.

Lucio di Jasio, business development manager with Microchip, said: “We now have 1250 8bit MCUs on offer; the range covers any application you can think of in the 8bit market.”

The benefit of the merger, he noted, is that Microchip can now offer the ‘two most liked’ 8bit MCU architectures. “Combining these two architectures,” he contended, “results in a powerhouse. And we have a lot of new customers.”

He also suggested the two companies have more in common than meets the eye when it comes to MCUs. “Our histories have been running in parallel,” he observed. “OTP, then flash, then 16bit, then 32bit. The two companies are like twins separated at birth.”

And so, in Di Jasio’s opinion, Microchip will take good care of Atmel and the AVR architecture in particular. “We haven’t spent more than $3billion to throw products away,” he said. However, he also suggested that the Atmel legacy might need a bit of closer attention. “Atmel had not focused on the AVR portfolio for some time.”

Looking to address that, Microchip has just launched the ATtiny817 range, developed – according to Di Jasio – in ‘record time’ since the acquisition. “This shows our commitment to the 8bit market and to AVR.”

Di Jasio’s contention that the two companies’ 8bit portfolios have developed in parallel is evidenced by a common feature in their MCUs. Microchip introduced core independent peripherals (CIPs) around 2010, while Atmel developed the Event System a little earlier. Both approaches have the same goal; reducing the requirement for peripherals to work via the device’s core. “Atmel had the same idea,” said Di Jasio, “and experimented with it in the XMega range. The Event System is similar to CIPs, but presents in a different way. It allows peripherals to send triggers to each other, remote from the CPU. We’ve introduced the first ATtinys with XMega features.”

Despite having 1250 8bit MCUs on offer, Microchip continues to expand the range; the latest addition being the PIC18FK40 family. “Customers can now choose between the PIC16, PIC18 and AVR ranges,” Di Jasio offered. “We’re developing parts with the latest technology, more RAM, more flash, more analogue and more comparators.”

One area where Microchip saw a particular gap in coverage was in rapid development tools. “There wasn’t anything available for AVR users,” Di Jasio claimed. “So we have expanded the Atmel START online tool – previously targeted only at ARM based MCUs – to include AVR, allowing users to create skeletons of projects, with base drivers. This is an important addition to the tool chain,” he added.

Meanwhile, Microchip will continue to support Atmel’s Studio 7 integrated development platform, as well as its own MPLAB integrated development environment. But whether the tools will eventually merge is yet to be decided. “It’s up to customers to tell us whether they want this to happen,” said Di Jasio. “We will be listening to what they have to say, but we have already had feedback on this.”

Looking at the 32bit arena, Di Jasio notes the portfolio has also expanded significantly. “The Atmel 32bit parts allow us to address areas that haven’t been covered by the MIPs based PIC32 range.

“The Atmel devices not only suit ultra low power applications,” he continued, “but also have MPUs, so they can run Linux, for example. That functionality would have taken years to develop, so we get access to new markets now. However, there are differences between the MIPS and ARM architectures that suit them to different applications. MIPS parts are good in audio, for example, while the benefit of the ARM architecture is that it brings access to the ecosystem, libraries and third party tools. More choice is always better.”

An interesting discussion point is whether the Atmel Cortex-M0 parts will compete with Microchip’s PIC based devices. Microchip’s answer is that not all applications need 32bit MCUs and that 8bit can provide a better solution in many cases.

Interestingly, Di Jasio contends that, while there is a transition in the industry from the use of 8bit to 32bit MCUs, there are designers moving the other way. “They wanted to try out 32bit MCUs,” he said, “but it didn’t work for them. It’s not a massive trend, but we do see it happening.”

But what’s the future for the 8bit portfolio – whether it’s Microchip or Atmel? Can such a broad range of devices be sustained? Di Jasio’s answer is that nothing is going to be discontinued. “You can expect to see our philosophy of customer driven obsolescence continue; as long as there are customers wanting to use a specific device, we will continue to provide it for them.”

Pointing to research which suggests that 30% of R&D budgets are consumed by redesigns due to obsolete products, Di Jasio said the Microchip approach offers customers the opportunity to decide when to redesign products, rather than being forced into it. “It’s up to customers to decide on the risk, it’s up to them when and how they want to innovate. We enable that.”

In fact, he claims Microchip is still seeing interest in parts designed 20 years ago. “People still ask questions like ‘how long will they remain in production?’ and ‘can I still place an order’? As long as they buy them, we’ll make them.”

And he says interest in the 8bit MCU remains strong. “Volume and revenue from the 8bit MCU sector continue to increase,” he said, “and we’re talking about billions. If you, as a manufacturer, can show you’re continuing to invest in a sector, then you will get a return on that investment – and Microchip is investing more today in the AVR portfolio than Atmel did over the last three years,” he concluded.