Skip to main content

Director of the Barcelona Supercomputing Center Is an Exceptional HPC Leader

mateo valero

Mateo Valero is a professor of computer architecture at the Universitat Politècnica de Catalunya, now called BarcelonaTech, and Founding Director of the Barcelona Supercomputing Center. In his decades-long career focusing on high-performance computing (HPC) architectures, he has published approximately 700 papers, given more than 800 invited talks, and helped organize more than 300 international conferences. 

In recognition for his many years of dedication to HPC, Valero was awarded the HPCwire Readers’ Choice Award in 2020 for his “exceptional leadership in HPC” and for being “the driving force behind the renaissance of European HPC independence.”

Mateo Valero was also a panelist in the I Am HPC Plenary at SC23.

Fostering Global Collaboration

In this I Am HPC profile, Valero gives a look into his history with HPC, as well as where he sees the industry going.

mateo valero

Mateo Valero

Director, Barcelona Supercomputing Center

Q: What single event most made you realize you wanted a career in HPC/computing?

Valero: The realization that I wanted a career in HPC was sparked when, around 1977-78, I collaborated with my undergraduate students to construct some small-scale (two to four processors) multiprocessor systems featuring 4-8-bit Intel processors and shared memory.

Q: What do you consider your biggest contribution to the HPC/computing community?

Valero: My most significant contribution to the HPC/computing community has been the pioneering work my students and I conducted on Vector Architectures. During a time when the prevailing belief was that pure superscalar processors would render vector architectures obsolete, we defied conventional wisdom. We foresaw that high-performance vector processors would continue to thrive, and this prediction has become a reality today. We also made substantial contributions by bridging the worlds of ILP [instruction-level parallelism] and DLP [data-level parallelism], fostering a productive cross-pollination between these two fields. Our introduction of the concept of “fuzzy computation” evolved into the idea of “reduced data formats,” which has proven immensely valuable in terms of energy and space efficiency while also contributing to the advancement of current AI [artificial intelligence] hardware.

Q: In the past 35 years, what is the most significant overlooked breakthrough that has impacted the field in your eyes?

Valero: Vector architectures, notably those championed by Cray’s pioneering work and followed by other Japanese companies, have been a fundamental—and often an overlooked—breakthrough in the landscape of HPC and supercomputing. I consider them a hidden treasure in high-performance computing with their remarkable energy efficiency, clean semantics, and profound influence on contemporary architectures like GPUs [graphics processing units]. As the world seeks more efficient and powerful computing solutions, the impact of vector architectures continues to be felt, both in the quest for scientific discoveries and the design of innovative hardware.

Q: What would you like to see change about, within, or among the HPC/computing community?

Valero: I envision a transformation in the HPC/computing community, where we harness the power of our exceptional machines, supercomputers, and various instruments to foster global collaboration in addressing critical societal challenges. Issues such as climate change, sustainable energy solutions, and personalized medicine demand collective efforts from diverse institutions. One of the most challenging and promising research areas in this context is the development of a digital twin for the human body. Achieving breakthroughs in this field necessitates an extensive collaboration, where institutions come together and share their valuable resources. By promoting such cooperation, we can accelerate progress and make a significant impact on these pressing issues.

i am hpc

Stay Up to Date

Sign up to receive the SC newsletter in your inbox.

Information provided is treated in accordance with ACM & IEEE privacy policies.

Back To Top Button