Tertulia mensual Fulbright el 14 de junio de 2022, *19:00 hrs*: «Supercomputación y soberanía europea» por el Prof. Mateo Valero

Publicado el 7 junio 2022 | Archivado en Actividades Socios, Comunicacion 2.0, Ryder Club de España, Tertulias Fulbright | Salir del comentario

Tertulia Fulbright Martes 14 de junio de 2022,  19.00 hrs

Profesor Dr. D. Mateo Valero

“Supercomputación y soberanía europea

En medio de esta precocísimo tiempo veraniego que nos ha traido el mes de junio, nuestra tertulia mensual contará con un ponente de excepción, el Profesor Mateo Valero, presidente fundador del Centro de Supercomputación de Barcelona, que en unas tres décadas de existencia se ha convertido en uno de los centros europeos líderes en supercomputación y desarrollo de hardware en general. La cual es en estos momentos una de las áreas críticas en Europa, con el fin de garantizar su autonomía estratégica.

Nuestra tertulia será virtual, por medio de plataforma ZOOM. Para asistir debéis preinscribiros antes en el formulario que aparece al abrir este enlace (haciendo click sobre el enlace o copiándolo en el navegador):

https://us02web.zoom.us/meeting/register/tZIvcOGtqDgoE9O25393JNpnuyAwOviJBl_W

Os llegará entonces una notificación de Zoom al correo electrónico con el enlace definitivo personal que deberéis usar para entrar a la sesión. Deberéis tener descargada la aplicación gratuita desde www.zoom.us

Os deseamos a todos mucha salud, y os esperamos el martes 14 de junio.

Nuestro Ponente

Mateo Valero (http://www.bsc.es/cv-mateo) is the Founding Director of the Barcelona Supercomputing Center, where his research focuses on high performance computing architectures. He has published approximately 700 papers, has served in the organization of more than 300 InternationalConferences and has given more than 600 invited talks. Prof. Valero has been honored with numerous awards, among them the three most relevant awards in the Computer Architecture field: The Eckert-Mauchly Award 2007 by IEEE and ACM, the Seymour Cray Award 2015 by IEEE, and the Charles Babbage 2017 Award by IEEE. Among other awards, Prof. Valero has received The Harry Goode 2009 Award by IEEE, The Distinguish Service Award by ACM, and the Spanish National awards “Julio Rey Pastor” and “Leonardo Torres Quevedo”. Prof. Valero is a «Hall of the Fame» member of the ICT European Program (selected as one of the 25 most influents European researchers in IT during the period 1983-2008, Lyon, November 2008). In 2020 he was awarded the “HPCWire Reader’s Choice Awards” “for his exceptional leadership in HPC” and for “being a HPC pioneer since 1990 and the driving force behind the renaissance of European HPC independence”. He has been also honored with Condecoración de la Orden Mexicana del Águila Azteca 2018, the highest recognition granted by the Mexican Government. Prof. Valero holds Honorary Doctorates by 10 Universities, is a member of 9 academies, and a fellow of IEEE, ACM, and AAIA, the Asia-Pacific Artificial Intelligence Association.

In 1998 Mateo Valero was distinguished as “Favourite Son” of his home town, Alfamén (Zaragoza) and in 2006, his native town of Alfamén named its Public School after him. 

Abstract

Over that last 3 decades, we have witnessed a transition from closed software ecosystems being the foundation for HPC, enterprise, and business to open source software ecosystems based on Linux: from Arduino in the IoT space, to Android in the mobile space to Linux in HPC and cloud-based systems with various Open Source Software projects built on top. However, when examining hardware, current commercial off the shelf solutions are closed hardware ecosystems that only enable integration at the peripheral (PCIe) level. The combination of current technology trends, the slowing of Moore’s Law, and cost prohibitive silicon manufacturing inhibit significant power-performance gains by relying on traditional closed ecosystems, especially in HPC, with technology pushed to the extreme. This new regime forces systems to be much more specialized in order to achieve the power-performance profiles required for a supercomputer. In the past, HPC has led the way forward, defining the leading edge of technology. HPC can do this again with open hardware, as it has done in software with the adoption of Linux and open source in general. This is not only a technology imperative, but one born out of current geopolitics. Digital Technology (the generation and processing of data) is the basis for global trade, scientific discovery, and is ubiquitous in modern life. Thus, creation of digital technology in the form of processors, accelerators and the related digital infrastructure guarantees access to these building blocks of the digital economy regardless of the geopolitical environment. Given this technology and geopolitical backdrop, we describe how Europe can exploit its resources targeting research and development for technological independence.

In today’s technology environment, some of the rules have changed. This has produced a shift from abundant transistors to efficient use of transistors. Thus, to truly meet the power and performance requirements, we must specialize the hardware. At the same time, the software stack is evolving, becoming more abstract, enabling higher programmer productivity, but sacrificing hardware efficiency. Thus, application owners will need to co-design the full stack, all layers of hardware and software, in order to meet their performance and power (e.g., FLOPs/W) targets. This level of integration is not possible in a closed or even partially open ecosystem. The platform must be open to enable this tight integration. We see this openness today in the Linux OS, toolchain, runtimes, frameworks, and libraries, up to the application layer. This enables rapid development and extension of software systems. However, an open hardware infrastructure was lacking, making specialization nearly impossible, especially in a research context. Openness is required to tailor your hardware platform to the applications, thereby achieving the desired performance in the power constrained environment. There have been a couple of open source hardware platforms in the past, but Moore’s Law inhibited their adoption for many reasons: general purpose processor improvements, time to market, cost, software development, etc. Furthermore, unlike Linux, previous open source hardware was entangled in the companies that created them. Mirroring the same model as Linux, RISC-V has followed a similar development path and has enjoyed significant industrial and academic adoption. Like Linux before it, the RISC-V ecosystem is in the nascent period where it can become the de facto open hardware platform of the future. The RISC-V ecosystem has the same opportunity in hardware that what Linux created as a foundation for open source software. This enables the co-design of the RISC-V hardware and the entire software stack, creating a better overall solution than the closed hardware approach that is done today. RISC-V enables everyone to build what they want and need vs. buy something that partially meets their requirements. As European HPC recognized in the past with Linux, Europe has the opportunity to lead the charge, creating a full stack solution for everything from supercomputers to IoT devices, all based on an open ISA, providing interoperability and a freedom to create, build, and deploy superior technology based on European IP.

In this talk, first, we will provide background on HPC computing and the research we have conducted to shape the current state of the art in HPC. Using RISC-V as an instrument, we provide a vision for the future and a collection of current research and innovation projects, infrastructure, and the community that are building the foundation for the future. This is a new opportunity for Europe to lead the way to an HPC Future that is Wide Open! 

Comentarios

No hay mas respuestas