Thesis Directors: Prof. P. Scarlino, Dr P. Harvey-Collard
\r\nPhysics doctoral program
\r\nThesis Nr. 11800
\r\n
\r\nTo take part in the public defense, please contact directly the speaker
Can AI be part of the sustainability solution - without becoming part of the problem? Join us for an afternoon of expert talks, real-world use cases, and open discussion at the intersection of artificial intelligence and environmental responsibility.
\r\n
\r\nEPFL Innovation Park, Building C, Room Neptune
\r\n
\r\n22 April 2026
\r\n16:00 – 19:00
\r\n
\r\nMore information and registration: HERE
Are you developing a medical device or a digital health solution?
\r\n
\r\nThe Human Health by Design program supports early‑stage teams in designing, prototyping, and validating solutions grounded in real clinical needs.
\r\nThis 6‑month program (June–December 2026) combines field immersion, user‑centered design, and multidisciplinary expertise to accelerate the adoption of impactful, safe, and clinically relevant technologies.
\r\n
\r\nWhat the program offers\r\n
Open-source hardware can play a unique role for the semiconductor industry in the age of sustainable AI. It can enable design reuse, foster collaboration, and support workforce development. ESP (Embedded Scalable Platforms) is an open-source research platform for system-on-chip (SoC) design that combines a modular architecture with an agile design methodology. The ESP architecture simplifies the design and prototyping of heterogeneous chips with multiple RISC-V processor cores and dozens of loosely coupled accelerators, all interconnected by a scalable network-on-chip. The ESP methodology promotes system-level design while accommodating different specification languages and design flows.
\r\n
\r\nESP's capabilities have enabled a small team, primarily composed of graduate students, to realize two SoCs of increasing complexity, each within a few months. Conceived as a heterogeneous system integration platform and refined through years of teaching at Columbia University, ESP is well suited to advance collaborative engineering across the open-source hardware community.
\r\n
\r\nThis talk will be followed by a standing lunch in BC420 from 12:15 to 13:30.
We will present examples of finitely generated groups that are not quasi-isometric to any 2-generated group, and discuss variants of this question.
","image_description":"","creation_date":"2026-04-13T10:20:38","last_modification_date":"2026-04-13T10:20:38","link_label":"","link_url":"","canceled":"False","cancel_reason":"","place_and_room":"MA B1 11","url_place_and_room":"https://plan.epfl.ch/?room==MA%20B1%2011","url_online_room":"","spoken_languages":[],"speaker":"Yves de Cornulier","organizer":"Nicolas Monod","contact":"Nicolas Monod","is_internal":"True","theme":"","vulgarization":{"id":2,"fr_label":"Public averti","en_label":"Informed public"},"registration":{"id":3,"fr_label":"Entrée libre","en_label":"Free"},"keywords":"","file":null,"icalendar_url":"https://memento.epfl.ch/event/export/120355/","category":{"id":1,"code":"CONF","fr_label":"Conférences - Séminaires","en_label":"Conferences - Seminars","activated":true},"academic_calendar_category":null,"domains":[],"mementos":["https://memento.epfl.ch/api/v1/mementos/7/?format=json"]},{"id":71579,"title":"MEchanics GAthering –MEGA- Seminar: Bridging the Scales of Boiling: How Microscale Interactions Drive Macroscale Heat Transfer","slug":"mechanics-gathering-mega-seminar-bridging-the-scal","event_url":"https://memento.epfl.ch/event/mechanics-gathering-mega-seminar-bridging-the-scal","visual_url":null,"visual_large_url":null,"visual_maxsize_url":null,"lang":"en","start_date":"2026-04-16","end_date":"2026-04-16","start_time":"13:05:00","end_time":"14:00:00","description":"Abstract: Boiling water is something we observe every day in our kitchens, yet it is also one of the most powerful thermal management tools in modern engineering — from steam-driven power generation and cooling next-generation supercomputers to managing cryogenic fuels in spacecraft. Despite its apparent familiarity, the underlying physics of bubble nucleation remains poorly understood. The challenge lies at the microscopic level. Classical nucleation theories treat boiling as a collection of isolated, independent bubbles growing on a surface, failing to capture the complex reality of actual engineering surfaces — which consist of dense arrays of interacting nucleation sites. Using custom-engineered silicon chips with precisely spaced microcavities, we isolate and systematically control these microscopic \"neighborhoods.\" By uncovering how the spacing between neighboring sites governs bubble nucleation, coalescence, and stability, we gain a new understanding of how microscopic interactions ultimately drive macroscale thermal performance — paving the way for safer, more efficient cooling technologies.We present a novel necessary and sufficient principle for multiple testing methods controlling an expected loss. This principle asserts that every such multiple testing method is a special case of a general closed testing procedure based on e-values. It generalizes the Closure Principle, known to underlie all methods controlling familywise error and tail probabilities of false discovery proportions, to a large class of error rates -- in particular to the false discovery rate (FDR). By writing existing methods as special cases of this procedure, we can achieve uniform improvements, as we demonstrate for the e-Benjamini-Hochberg and the Benjamini-Yekutieli procedures, and the self-consistent method of Su (2018). We also show that methods derived using our novel e-Closure Principle generally control their error rate not just for one rejected set, but simultaneously over many, allowing post hoc flexibility for the researcher.
\r\nMoreover, we show that because all multiple testing methods for all error metrics are derived from the same procedure, researchers may even choose the error metric post hoc. Under certain conditions, this flexibility even extends to post hoc choice of the nominal error rate.
\r\n
\r\n
\r\n