From a Startup Perspective. Using startup methodology to try and create a feasible business model for FPGAs
Examensarbete för masterexamen
Management and economics of innovation (MPMEI), MSc
This thesis has been part of a project between students from "High Powered Com-puter Systems" and "Management and Economics of Innovation at Chalmers Uni-versity of Technology" where before the project started the members of the project came up with a product vision that was going to be evaluated with the use of startup methodology. The vision was based on FPGA technology utilizing the capability to reconfigure itself between two di˙erent modes. Due to the need of a pivot, an evaluation of the technology was also made. The research was performed using unstructured interviews with experts in the tech-nology and players in its relevant markets in combination with the Customer Devel-opment framework by Steve Blank and Bob Dorf. The study resulted in insights on the properties of the FPGA technology such as price, flexibility and performance. Along with this is technical information in the field of machine learning such as training at the edge, training on the FPGA and applications of machine learning in an autonomous vehicles. In addition, the results highlighted existing use-cases of FPGAs in industries and discussions on possible future developments. The interview also resulted in perspectives on how hardware is chosen with respect to factors such as power consumption, volume and other al-ternative hardware. Lastly, The authors found the initial idea less feasible because of problems with both edge learning and training on the FPGA. There is a possibility to further develop training on the FPGA. In the pivoting process, the authors faced a diÿ-culty developing a suitable alternative business case that involve using the switching mechanism due to the nature of the technology and its applications. In conclusion, FPGA technology has many advantages and disadvantages. Its flexi-bility is a great strength but when considering price factors the industry move toward other options such as ASIC, especially in higher volumes. These factors make the FPGA suitable for a market that is low volume or requires higher flexibility. Space applications is one possible potential market as it fits well with these criteria. There-fore, it is worth further investigation in future research.
startup, entrepreneurship, FPGA, business model