Lattice Blog


Latest Consumer News from Lattice

Consumer News from Lattice
Posted 05/13/2020 by Lattice Semiconductor

Posted in

CrossLink Reference Designs Boost Design Options

Need to develop extremely small form factor embedded vision solutions? New reference designs for CrossLink promise to expand your design options while delivering greater ease of use. An N to 1 side-by-side sensor aggregation reference design allows designers to employ low power SoCs with a limited number of MIPI interfaces instead of larger, more power- hungry SoCs with a higher number of MIPI interfaces. For newer SoCs that support MIPI virtual channel, developers can use CrossLink’s N to 1 virtual channel aggregation reference design. Together, these two reference designs will offer designers the option to use any MIPI-enabled SoCs to implement MIPI sensor aggregation while addressing small form factor design challenges.

CrossLink-NX: Unmatched Design Versatility

How versatile is the CrossLink-NX? From gaming handheld, VR headsets to mobile phones, CrossLink-NX can be used in a wide range of consumer applications. Handheld gaming system architecture often leverages the mobile processor ecosystem. Sometimes, displays call for the system to store an entire video frame in external memory for processing. This task must be accomplished without out any discernable latency. CrossLink-NX provides a unique combination of fast external memory interface and large internal RAM in a small and low power form factor to deliver a low latency visual experience. With a 1066 Mbps DDR3 memory interface and a large internal RAM, Crosslink-NX meets that need. Similarly, designers building AR/VR headsets often need to aggregate the output of 5 to 10 image sensors simultaneously. With the CrossLink-NX designers can merge multiple sensor outputs into a CSI-2 receiver, process that data in an ISP, and aggregate the output into a DSI interface for display.

Another way designers are using the CrossLink-NX in consumer applications is to offload computational tasks in high resolution mobile cameras. Adding high resolution camera to phones using mid to low end mobile processors can be challenging. These processors require a visible delay to render a high-solution photo due to the lack of processing power. Using the CrossLink-NX’s significant number of LUTs and DSPs blocks as well as its low power 2.5 Gbps MIPI interface, designers can offload much of the processing in the ISP and let the users view high resolution images instantly.

CrossLink-NX Examples

Lattice mVision Solutions Stack Accelerates Embedded Vision Development

Thinking of building an embedded vision system application? Check out Lattice’s mVision Solutions Stack. Offering designers everything they need to accelerate implementation of an embedded vision application, the stack brings together all the tools required to develop and build very low power FPGA-based embedded vision applications for IoT, consumer robotics, and drones. The new technology stack starts by offering modular hardware platforms through Lattice’s Embedded Vision Development Kit. It then adds IP blocks, a variety of FPGA design tools, reference designs and demos, and a network of third-party custom design services. To support today’s low power edge applications, the stack is optimized for power consumption as low as 150 mW. It also features package sizes as small as 2.5 x 2.5 mm.

Giving New Life to Legacy Designs

Building embedded vision systems is a great place to be right now. Who knew that the number of apps using embedded vision would grow exponentially as developers in many fields discover new applications for the technology. Give credit to the MIPI Alliance and its low-cost sensor and display protocols for helping clear the path to new use cases for low cost, compact vision systems. But also note that the rising availability of low density, low power FPGAs in bridging applications, such as Lattice’s CrossLink FPGAs, now allow designers to extend the life of systems using proven legacy sensors and displays or upgrade existing systems and code by moving to a MIPI-enabled processor. To learn more about these trends check out Lattice’s new white paper and blog entry on the topic.