Mark Benson

OpenCL is a programming framework for managing heterogeneous computing resources such as CPUs and GPUs. OpenCL can be used to accelerate computationally-intensive algorithms in a robust and scalable way. However, complicated problems often require complicated solutions, and OpenCL is a complicated (and powerful) solution. This article delineates the complexity of OpenCL in two ways: (1) by describing the architecture of OpenCL, and it's primary organizing principals (static model, dynamic… Show more

OpenCL is a programming framework for managing heterogeneous computing resources such as CPUs and GPUs. OpenCL can be used to accelerate computationally-intensive algorithms in a robust and scalable way. However, complicated problems often require complicated solutions, and OpenCL is a complicated (and powerful) solution. This article delineates the complexity of OpenCL in two ways: (1) by describing the architecture of OpenCL, and it's primary organizing principals (static model, dynamic model, memory model, and compilation model); and (2) by describing how to use OpenCL in real-world applications (creating, building, compiling, linking, executing, and debugging OpenCL programs). The article ends with a summary of next steps for OpenCL, and recommendations for further reading. Show less

In embedded systems there is a science to choosing the right processor for a given design. Usually processors are chosen based on objective factors such as peripheral support, physical packaging, memory, architecture family, and software tools available. This type of quantitative analysis lends itself well to spreadsheets and formulas that can help choose which processor is best. However, for small high-performance battery-powered devices that have heavy processing requirements, thermal… Show more

In embedded systems there is a science to choosing the right processor for a given design. Usually processors are chosen based on objective factors such as peripheral support, physical packaging, memory, architecture family, and software tools available. This type of quantitative analysis lends itself well to spreadsheets and formulas that can help choose which processor is best. However, for small high-performance battery-powered devices that have heavy processing requirements, thermal performance quickly becomes an urgent design issue to overcome and is often viewed as an elusive and opinion-fraught black-art. Modeling, managing, and characterizing the efficient dissipation of heat in these types of devices is complex. Fortunately, there are new processors coming to the embedded marketplace that have advanced features for managing thermal performance. This, combined with new advances in software modeling techniques and algorithms gives us a fresh playground to innovate new ways to manage thermals in a way that was not previously possible. Show less

As the state of the art in diagnostic devices evolves within the context of our flattening world economy, manufacturers are facing key defining challenges that affect not only business performance, but also users and patients: uncertain regulatory landscapes, increasingly-demanding user expectations, and tightening cost pressures. By analyzing and implementing design patterns from adjacent industries, we can influence diagnostic device design by opening the eyes of the design process to create… Show more

As the state of the art in diagnostic devices evolves within the context of our flattening world economy, manufacturers are facing key defining challenges that affect not only business performance, but also users and patients: uncertain regulatory landscapes, increasingly-demanding user expectations, and tightening cost pressures. By analyzing and implementing design patterns from adjacent industries, we can influence diagnostic device design by opening the eyes of the design process to create more nimble, innovative, and disruptive diagnostic products within an increasingly competitive climate. Show less

Slow regulatory cycles combined with the reality of fast-moving technology curves can cause new homecare telemedicine products to feel old to users that have the latest consumer technology in their pocket. This presentation explores problems and potential solutions to combating this effect. Problems include (A) fast-moving consumer technology keeps setting the bar for user interaction and design standards, (B) keeping current with latest technology often requires the creation of non-standard… Show more

Slow regulatory cycles combined with the reality of fast-moving technology curves can cause new homecare telemedicine products to feel old to users that have the latest consumer technology in their pocket. This presentation explores problems and potential solutions to combating this effect. Problems include (A) fast-moving consumer technology keeps setting the bar for user interaction and design standards, (B) keeping current with latest technology often requires the creation of non-standard proprietary solutions, and (C) the desire to exert control over the safety and efficacy of a medical product causes medical device companies to naturally trend towards vertical integration, and trend away from horizontal market relevance. This presentation presents possible solutions to each of these problems, but above all, argues that a careful and smart approach to interaction design is the key to wide-spread adoption and success. Show less

Thanks to the laws of physics and the practical limits on rates of thermal heat dissipation, CPU clock cycles in integrated circuits are no longer increasing year-over-year at the same rate that memory densities are. For this reason, we have seen a proliferation of multi-core CPUs for commercial desktop PCs, and are now beginning to see this trend impact embedded systems as well. This presentation explains intersecting variables of performance, power, and complexity, and how these variables… Show more

Thanks to the laws of physics and the practical limits on rates of thermal heat dissipation, CPU clock cycles in integrated circuits are no longer increasing year-over-year at the same rate that memory densities are. For this reason, we have seen a proliferation of multi-core CPUs for commercial desktop PCs, and are now beginning to see this trend impact embedded systems as well. This presentation explains intersecting variables of performance, power, and complexity, and how these variables change dynamically in a multi-core embedded environment. The primary focus of this presentation is to delineate tools and techniques for managing homogeneous and heterogeneous embedded cores through intelligent architectural decomposition, particularly when low-power design is a primary constraint. Show less

Organizations in a variety of industries are adopting mobile wireless technologies at an explosive rate. This rapid deployment is dramatically increasing the demand on mobile networks. Wireless carriers claim that supporting the explosion in data services is affecting profitability. The likely outcome will be new data plans to better monetize data services, which could include charging by capacity, speed, time of day, or content type. This change creates uncertainty and potentially high costs… Show more

Organizations in a variety of industries are adopting mobile wireless technologies at an explosive rate. This rapid deployment is dramatically increasing the demand on mobile networks. Wireless carriers claim that supporting the explosion in data services is affecting profitability. The likely outcome will be new data plans to better monetize data services, which could include charging by capacity, speed, time of day, or content type. This change creates uncertainty and potentially high costs in the business models for wireless data applications. Show less

Cloud computing is an emerging trend. Consequently, more and more embedded devices are becoming connected to "The Cloud." Features that were previously included on the device are now being moved to The Cloud and provided as a service. This trend not only requires a new way of thinking about system design, but also enables a new level of algorithmic analysis that is moving us closer to unlocking the true promise of device clouds: data visualization.

As telehealth home care continues to evolve, the wireless and medical device industries face numerous challenges. This article explores technology, patient, and standards considerations, in addition to the necessary improvements to wireless technology – all factors critical to the ultimate success of in-home health care management.

Medium to large medical device companies often strive to become vertically integrated: they might own their own silicon fab, design their own hardware, engineer proprietary protocols, build their own testing and deployment equipment from scratch, and design their own custom manufacturing processes. Although there are good reasons for companies in regulated markets to become more vertically integrated (control over diverse processes, control of quality, and control of the supply chain), it can… Show more

Medium to large medical device companies often strive to become vertically integrated: they might own their own silicon fab, design their own hardware, engineer proprietary protocols, build their own testing and deployment equipment from scratch, and design their own custom manufacturing processes. Although there are good reasons for companies in regulated markets to become more vertically integrated (control over diverse processes, control of quality, and control of the supply chain), it can often cause a technology drift effect over time between the company’s core competencies and industry trends. This article discusses how outsourcing can be used strategically in key circumstances to gain valuable access to otherwise unavailable and diverse solutions from non-medical industries for medical device manufacturers thereby minimizing the technology drift effect. Show less

Due to the recent increased rate of Android adoption for mobile computing platforms, a natural question arises for embedded product developers and leaders regarding how to engage with the movement and leverage its success. By using the economic model of Cross-Elasticity of Demand (XED), we can come up with a framework that guides our thinking about how the changing price point for one product family (i.e. the adoption of Android), affects the demand for another (i.e. a given corporate software… Show more

Due to the recent increased rate of Android adoption for mobile computing platforms, a natural question arises for embedded product developers and leaders regarding how to engage with the movement and leverage its success. By using the economic model of Cross-Elasticity of Demand (XED), we can come up with a framework that guides our thinking about how the changing price point for one product family (i.e. the adoption of Android), affects the demand for another (i.e. a given corporate software product roadmap). This presentation attempts to introduce such a framework by analyzing (A) independent products in the marketplace, (B) substitute products in the marketplace, and (C) complementary products in the market place, with the goal of enabling intelligent and timely business decisions within a dynamic technical climate. Show less