Ganesh Moorthy is the executive vice president at Microchip Technology, which manufactures microcontroller, memory, and analog semiconductors. The company’s products include the PICmicro and dsPIC/PIC24 microcontrollers; serial-EEPROM, Keeloq, RF, thermal, power, and battery-management analog devices; and linear, interface, and mixed-signal devices. Moorthy recently spoke with EDN about supporting embedded developers.

Microchip has broadened its processor portfolio to span from 8- to 32-bit processors, including microcontrollers and digital-signal controllers. What is Microchip’s strategy for supporting this range of architectures?

Our strategy is to offer a broad range of microcontrollers across the 8-, 16-, and 32-bit architecture ranges. We continue to hear feedback from our customers that this [range] is what they want, and we continue to innovate for each of these architecture points. The reason for this [offering] is that, in the embedded space, the needs of a customer are quite fragmented, and it is important to have a number of products that can all solve any one problem to enable the customer to pick and choose as they go through a design and identify their real system requirements—the final product that meets the exact system requirements.

Our own experience says that, in as much as 70 to 80% of designs, the product the customer goes to production with is different from the product they begin the design development on. The reason for this [difference] is that, during the time frame that they are conducting the design, the customers [have the time] to better refine their system requirements. They are able to appropriately adjust on the fly to new needs identified by their sales and marketing teams to make their end product more competitive and to be able to do all of this in some seamless fashion. So, they do not have to throw away all of their engineering effort each time they need to make a change. The ability to shift to the next appropriate product—whether for cost, size, features, or whatever is driving the change—is a very powerful need. Having a broad portfolio of products spanning 8, 16, and 32 bits is intended to offer that breadth of choice, no matter how their needs are changing.

How does Microchip address balancing general-purpose and application-specific processing?

Almost all of our products could be considered general-purpose products. Now, we do take many of them and give them an application twist, so to speak. For some of the products, we will add peripherals that are more fine-tuned to running specific applications, such as for motor control or touch sensing, or add some communication protocol, such as CAN [controller-area networking], USB, or Ethernet. Fundamentally, the products are broad-based, and then we take one or more dimensions where we give it a little more personality for a particular application without impeding its use in other applications.

The goal is to make it a low-risk effort to enter a design and get to production without having to rework the software and the best total cost implementation for the application. The PIC platform is a singular common platform that supports development across the entire line of 8-, 16-, and 32-bit products with an integrated development environment, assemblers, other software, and hardware debugging for emulation. The 32-bit controllers are upgrades from the 16-bit controllers.

Besides growing computational-processing capabilities, what other capabilities do you see being integrated with embedded processors? How are these features helping to drive innovation?

In the area of analog, we have a significant amount of analog integrated onboard, such as oscillators to avoid needing external oscillators, brown-out-resets, as well as ADCs and DACs. Then, we have added more traditional peripherals, multiple UARTs, multiple I²Cs [inter-integrated circuits], multiple SPIs [serial-peripheral interfaces], as well as hardware modules to help handle software protocols, such as LIN [local-interconnect network], that can run off a standard UART, but, if the UART has additional hardware features, it is easier to implement the protocol. We’ve added the capability of adding a touch peripheral to our products without the need for external components.

There is a range of communication peripherals, USB and USB On-The-Go communication, as well as Ethernet MAC [media-access-control] and PHY [physical] layers integrated onboard. The peripheral-pin select allows customers to remap pins to whatever I/O they want, so they can fit [the design] into smaller packages and use more intelligent routing on their boards.

Our focus is on the innovation that customers want to apply and the flexibility they want to have in what they do. We try to enable innovation. At any given point, customers are thinking about what innovation means to them, and, at one point, it may mean getting to lower cost or smaller size or being able to implement something, like touch sense, that they couldn’t implement before. We service 63,000 customers worldwide, and it would be impossible to try and pick whose need is going to be the place where all the growth is going to come from. We largely try to make the products and the development tools easy to use and low cost. We offer lots of free material on our Web site so that our customers can try things out and technical support to help them overcome any problems. In the end, we want to enable our customer’s innovation.

Software represents a significant portion of your product as tools, drivers, libraries, application notes, and reference designs; where is this software going, and where are you focusing your energy?

The more complex our products get and the more complex the end solutions our customers are designing into are, the more important the software side of the equation becomes. The customers’ ability to complete their designs is in many ways a function of whether we have sufficient amounts of reference designs, reference code, application notes, and development tools that support all of this development. It is an area we invest a lot in, but it is not the area we get paid on; it is all done to enable the sale of the microcontrollers, but they are no less important. When we bring our products to market, we complete all of these pieces in parallel with the silicon development to ensure that, when our product comes to market, it comes to market as a complete product, which is not just the silicon but with all the other accompaniments that enable the customer to use that silicon effectively. So, what is around the silicon is a major focus of what we do and where we invest.—Interview conducted and edited by Robert Cravotta