DR CLAUDE DIDERICH

The domain name sigmadream.com is not just a combination of words, it’s a reflection of my personal journey. It marries mathematical exactness, represented by the Greek letter sigma, with the emotional resonance of the word dream. This unique blend symbolises the goal of my website-to aggregate all key insights and communication facilities about me under one internet address. The use of sigma to express uncertainty in probability theory also underscores the importance of uncertainty in innovation and excellence. As Walt Disney once said, ‘If you can dream it, you can do it!’. It is this definition of a dream that is encapsulated in the domain name.

I am the founder and managing director of the boutique foresight, strategy, and innovation consulting firm innovate.d llc in Richterswil, Switzerland. I have over 25 years of experience in foresight, strategy development, business model innovation, and process design, with a special focus on service industries, particularly private banking and institutional asset management. I have acquired my expertise while in senior positions at Credit Suisse, Bank Julius Bär, UBS, Deloitte Consulting, and most recently as founder of innovate.d.

I hold a doctor ès sciences and a masters in computer science engineering from the Swiss Federal Institute of Technology in Lausanne, a certificate of advanced studies in strategy from the University of St. Gallen, and a specialisation certificate in design thinking and innovation from the University of Virginia. I am a member of the Association for Computing Machinery and the Strategic Management Society.

My approach to problem-solving is a harmonious blend of creativity and analytical rigour. I am passionate about customer-centric business model innovation and firmly believe that this balance is at the heart of successful problem-solving. By relying on design thinking, I love developing creative, human-centred solutions that are not only desirable but also feasible and viable.

Most firms excel at exploiting the present but feel challenged when confronted with exploring the future, let alone shaping it.

• Fostering critical thinking concerning long-term developments
• Exploring possible probable, as well as alternate futures to inform strategy development
• Moderating foresight workshops, developing diverse scenarios about the future and understanding their implications on the present

Numerous firms have learned the hard way that strategies, like yoghurts, do expire, especially if customers are left behind.

• Developing customer-centric strategies using design thinking
• Identifying sustainable competitive advantages using game theory
• Relating profitability to customer value creation

Innovation is about transforming novel ideas, like inventions, into products and services that satisfy customer needs and stimulate their willingness to pay.

• Designing new products and services that address specific customer needs, create value for all stakeholders, and can
• Distinguishing between innovating in demand-driven versus supply-driven markets
• Aligning innovation with environmental, societal, and regulatory requirements

Understanding technologies like artificial intelligence, blockchain, or the Internet of Things is mandatory to navigate today’s uncertain environment successfully.

• Applying artificial intelligence to solve real-world problems
• Understanding the value and limitations of state-of-the-art technologies
• Developing web-based software solutions, leveraging the latest technological advancements

Operational research deals with developing and applying analytical methods to improve decision-making.

• Modelling supply chain business needs in an analytical framework
• Reasoning about game theory and developing game-theoretic algorithms
• Solving combinatorial optimisation problems

In recent years, the demand for parallelising numerical applications has grown. Two significant challenges must be addressed when parallelising nested loops for distributed memory parallel computers: scheduling the loop iterations and mapping the computations and data elements onto the processors. The scheduling function must accommodate all data dependencies within the sequential loop nests. The mapping function must ensure that a specified degree of parallelism is achieved. Furthermore, the amount of communication overhead resulting from nonlocal data references should be minimised.

In this thesis, I examine the alignment problem, which is defined as the challenge of mapping computation and data onto the processors. I formalise the alignment problem as finding computation and data mapping functions that minimise the number of remote data accesses for a specified degree of parallelism by introducing a linear algebra framework. I demonstrate that this problem is NP-hard. I present an exact implicit enumeration algorithm. To enable the resolution of large alignment problems, I propose an efficient heuristic and apply it to various benchmarks.

I compare my approach to the alignment problem with related work. I prove that my framework employing a linear algebra paradigm generalises other approaches.

DOI: 10.5075/epfl-thesis-1810