Embracing Change in Embedded Design
Posted 08/05/2022 by Lattice Semiconductor
At this year’s Embedded World 2022, Lattice presented a keynote address by Steve Douglass, head of R&D, the importance of flexibility and adaptability in embedded system design – both at the software layer and the hardware layer – to keep pace with the major technology trends of today and tomorrow. This blog provides a high-level recap of the presentation.
We are living in a time in our industry with some of the most profound changes in technology that are fundamentally reshaping our world. The global 5G rollout only sets the stage for more low-power, low-latency connected devices, with virtually every device and system being connected to the internet. Not only are we connecting billions of devices to the internet, they’re also being equipped with many more sensors – cameras, microphones, radar, lidar, accelerometers – making them more aware of the world around them. Advances in computer vision and edge computing are improving the capabilities of those devices, turning them into an ambient and persistent layer across our environment. And these devices will be generating massive amounts of data equipped with AI algorithms, which leads to an explosion of intelligence at the Edge.
The convergence of these trends is driving an increased need to make embedded designs more flexible and adaptable so they can keep pace with accelerating and constantly evolving technology requirements. Key challenges system designers must be prepared to face without undergoing full on, bottoms-up redesign and deployment include increasing compute requirements, system feature expansion, escalating security threats, and faster time to market.
Accelerating Compute Requirements
One of the challenges with driving adaptability is the rapidly growing need for compute at the Edge, especially from the rapid advancements in AI. This trend is echoed in embedded computing devices and is driving the need for heterogeneous computing with dedicated hardware accelerators for the most daunting computational tasks.
FPGAs are, by nature, incredibly flexible, making them an ideal choice for embedded designs. They have a programmable logic core that can be configured to implement virtually any logic function and their programmable I/O can support many different standards and protocols. This functionality can be updated over time, even after systems have been deployed in the field, providing unparalleled upgrade and supply chain flexibility, and enabling faster time to market.
Expanding System Features
Adaptable design techniques can make embedded systems more flexible and capable, integrating AI and evolving it over time.
Expanding intelligence of in-cabin car systems or smart doorbells, embedded vision applications requiring advanced Computer Vision algorithms including autonomous mobile robots, industrial machine vision cameras are just a few great examples of where adaptable design methods powered by FPGAs can be used for its optimal power efficiency and flexibility. FPGAs are an ideal choice as designers can modify the accelerators and future-proof their design against changing AI requirements over time.
Increasing Security Threats
A downside to a fully connected world is the increased ability of bad actors to leverage that connectedness in unsavory ways. As embedded designs get more complex, it expands the attack surface and the vectors for attack. This makes both software and hardware attack protections more necessary than ever.
Hardware Root of Trust (HRoT) and Platform Firmware Resiliency (PFR) are key to creating a highly trusted device. If an FPGA is used for the HRoT, the crypto algorithms can be updated over time and can be used to monitor and protect the software of the compute elements in the system for true PFR. It takes both the software and hardware adaptability to stay ahead of growing security threats to our embedded designs.
Faster Time to Market
In addition to the challenges addressed above, system designers under more pressure than ever to get their solutions to market quickly. Indeed, in tech, being first to market can mean the difference between success and failure.
Lattice is focused on providing solution stacks that help our customers implement their adaptable embedded hardware and software designs more quickly and efficiently. Our solution stacks provide turnkey application-specific solutions that combine reference platforms and designs, demos, IP building blocks, FPGA design tools, and custom design services that help engineers get the designs built and delivered to market more quickly.
Today, the Lattice solution stacks portfolio includes solutions for market applications including AI with Lattice sensAI™, embedded vision with Lattice mVision™, factory automation with Lattice Automate™ , security with Lattice Sentry™, and 5G ORAN deployment with Lattice ORAN™. And we have more on the horizon.
As the world becomes ever more connected with the Internet of Everything, as Edge systems become more intelligent and - in some cases - fully autonomous, the inherently flexible nature of FPGAs becomes increasingly essential. As systems engineers, adopting an adaptable engineering mindset will help ensure systems and solutions are equipped to keep pace with the evolving challenges that interconnectedness presents.
At Lattice, we’re focused on providing ways for our customers to address the ever-changing needs of applications around AI, security, embedded vision, control functions, and more with our small, low power FPGA devices and technologies.
Learn how you can outpace market expectations with Lattice FPGAs and solutions by reaching out to us today!