Internship Opportunities at Altera

For information about the opportunities below, contact Benjamin Cope

  • Using high-level synthesis/C to target embedded vision applications to FPGAs
    • Computer vision applications are becoming more common in everyday life.  Taking these algorithms and implementing them on highly embedded systems is a rapidly growing area.   Cost effective, high performance technologies such as System on Chip (SoC) based FPGAs provide the ideal platform to meet the challenges of embedded vision systems of the future.The internship project will extend a real world, computer vision algorithm (such as an optical flow algorithm for object tracking), producing a parameterisable, optimized implementation targeting Altera SOC devices.  The internship will involve the use of recent developments in High Level Design tools and methodologies to explore their rapid prototyping potential.The project will explore a variety of implementation trade-offs to show how our customers can differentiate within the performance, cost, quality, power spectrum.  These may include:
      • Use of High level design tools and libraries targeting FPGA fabric
      • Development of vision library functions focused on the target application
      • Partitioning the algorithm across the dual ARM A9 and the FPGA fabric
      • Latency vs throughput vs resource usage
      • Algorithm tuning/refinement to fit strengths of implementation platform
      • Comparative analysis (performance and resource usage) of implemented functions versus commercially available IP megacores
      • Packaging and preparation of end design for industry tradeshow
        • Including self-driven market study to adapt design to be representative of applications in target domain (e.g. military, medical, industrial).

      Experience with the following would be highly desirable for this project;

      • Software programming in C/C++
      • OpenCV computer vision library
      • VHDL/Verilog

       

  • Study of FPGA application in future broadcast studio architectures
    • Description: With a move in the video market to UHD video formats (now commonplace in high-street electronics shops), the associated video bandwidth uplift means new studio architectures are needed. A move to Ethernet-based transport, away from traditional BNC-cable SDI interconnects, to meet new bandwidth demands is also enabling a path towards, and interest in, the use of virtualized common hardware platforms. The common hardware platforms typically first considered are COTS (commercial off the shelf) servers and the virtualized behaviour is that they can be repurposed to any (e.g. CODEC, mixing, format conversion) function. These servers replace bespoke hardware products, where each function required a unique piece of hardware.This internship will involve the study of the use of Altera FPGA devices in a world where all studio hardware is based on COTS servers, e.g. as plug-in accelerator cards. The activity will include exploring algorithmic performance (latency, power, and throughput) and analysing the bottlenecks for CPU versus FPGA implementation. Common studio functions will be used as a test bed for this study. A successful outcome of the internship would be a technical paper summarising the findings and for presentation to an academic conference and/or to Altera customers.The ideal candidates would have a passion for video engineering and experience with software programming in Linux/server environment.
  • Creation of Designs for Financial Algorithms
    • Altera’s UK-based Heterogeneous-computing Algorithmic Library (HAL) team – is responsible for building Altera’s algorithmic and mathematical library functions for both hardware and software. The team is looking for a highly motivated engineer to investigate and evaluate possible designs for a wide range of financial library components that in addition to being highly scalable are capable of being used directly from software and hardware. The ideal candidate will have a willingness to learn new financial concepts.Look for Library Functions for Financial Markets: Creation of Designs