skip to content

NanoManufacturing

Michael De Volder, Engineering Department - IfM
 
Subscribe to http://talks.cam.ac.uk/show/rss/5408 feed
This is a superlist of research seminars in Cambridge open to all interested researchers. Weekly extracts of this list (plus additional talks not yet on talks.cam) are emailed to a distribution list of over 200 Cambridge researchers by Research Services Division. To join the list click here https://lists.cam.ac.uk/mailman/listinfo/biophy-cure For more information see http://www.cure.group.cam.ac.uk or email drs45[at]rsd.cam.ac.uk
Updated: 3 days 21 hours ago

Wed 04 Jun 13:00: Title to be confirmed

Tue, 27/05/2025 - 16:37
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Wed 04 Jun 11:00: Exploring charge density waves in NbSe2 with machine learning

Tue, 27/05/2025 - 15:35
Exploring charge density waves in NbSe2 with machine learning

Niobium diselenide has garnered significant attention over the past few decades because of the coexis tence of superconductivity and charge density waves (CDWs), observable down to the monolayer limit. Introducing relative twist angles between monolayers, in the field of twistronics, offers a new variable to tune these systems, yet a fundamental question remains: do CDWs persist in moiré structures, and how are they altered compared to the pristine monolayer/bilayer? Traditional first-principles methods face limitations due to the computational resources required for long-wavelength moiré patterns; for instance, a 1-degree twist angle necessitates modeling over 10,000 atoms, making simulations impractical. This study employs first-principles data to develop machine learning interatomic potentials with the Allegro architecture, enabling scalable and accurate simulations. We investigate the formation and evolution of CDW order in monolayers and twisted bilayers, validating our results against density functional theory calculations with minimal errors in energy and forces. Beyond niobium diselenide, our goal is to establish a protocol for studying CDWs in two-dimensional systems. We outline strategies for producing training data and perform a detailed hyperparameter scan to identify key aspects for studying these systems [1].

  1. Norma Rivano et al. arXiv.2504.13675 2025

Add to your calendar or Include in your list

Tue 27 May 14:00: 007: End-to-End Encrypted Audio Calls via Blind Audio Mixing

Tue, 27/05/2025 - 13:52
007: End-to-End Encrypted Audio Calls via Blind Audio Mixing

End-to-end encryption (E2EE) for messaging has become an industry standard and is widely implemented in many applications. However, applying E2EE to audio calls, particularly group calls, remains a complex challenge. Unlike text messages, audio calls involve capturing audio streams from each participant, which must be combined into a single, coherent audio stream that all participants can hear. This is known as audio mixing. In a non-E2EE system, the audio is mixed by a central server, and the result is sent to each participant. In contrast, in an E2EE system, each audio stream must be encrypted locally and sent to every participant in the group call. This method presents major challenges with respect to network overhead, audio synchronization and limitation on applying audio enhancement techniques.

In this talk, we present a new approach using Fully Homomorphic Encryption (FHE), which enables end-to-end encryption for group voice calls. Concretely, we introduce blind audio mixing and an FHE -compatible compression technique.

Zoom link: https://www.google.com/url?q=https://cam-ac-uk.zoom.us/j/83912370794?pwd%3DKOjLaKTwbRWvlsSjiLSgpTqIkEs8xI.1

Add to your calendar or Include in your list

Wed 18 Jun 11:15: Title to be confirmed

Tue, 27/05/2025 - 13:39
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Thu 19 Jun 15:00: Quadrature By Rational Approximation

Tue, 27/05/2025 - 13:38
Quadrature By Rational Approximation

Many numerical algorithms rely on quadrature formulas such as Gauss quadrature, the trapezoidal rule, and their conformal transplantations to specialized domains. Each quadrature formula can be interpreted as a rational approximation to an analytic function with a branch cut. Reversing the logic, new quadrature formulas can be quickly derived even for specialized domains by numerical rational approximation via the AAA algorithm, avoiding the need for conformal maps or other analysis. The poles of the rational approximations delineate branch cuts, and the poles and residues are the quadrature nodes and weights. The talk will present ten examples: five known problems, plus a variant for each one. I hope it will change your understanding of quadrature formulas.

Add to your calendar or Include in your list

Fri 12 Dec 13:00: Seminars in Cancer Please note this seminar is on a Friday

Tue, 27/05/2025 - 13:15
Seminars in Cancer

Abstract not available

Please note this seminar is on a Friday

Add to your calendar or Include in your list

Thu 20 Nov 13:00: Seminars in Cancer

Tue, 27/05/2025 - 13:14
Seminars in Cancer

Abstract not available

Add to your calendar or Include in your list

Fri 30 May 16:00: Searching for Ultra-light dark matter with atomic & nuclear clocks and Interferometers

Tue, 27/05/2025 - 12:29
Searching for Ultra-light dark matter with atomic & nuclear clocks and Interferometers

Fundamental physics has come to somewhat of a crossroads. With the as-yet absence of new particles at high energy colliders it has become increasingly attractive to consider the possibility that the new physics we seek has remained hidden not by an inaccessible energy barrier but via incredibly weak couplings to the Standard Model. Emerging quantum sensing technologies have recently unlocked a number of tantalising avenues to probe new physics at this feebly interacting frontier. In this talk I will outline how atomic & nuclear clocks and interferometers offer a means to detect theories of ultra-light dark matter which cause fundamental constants to oscillate in time. I will then propose a new method for detecting scalar ultra-light dark matter – nuclear interferometry – and show that this may provide access to unchartered theoretical territory.

Add to your calendar or Include in your list

Tue 03 Jun 11:00: More ARMs than arms: From Sunk to Silicon Supremacy

Tue, 27/05/2025 - 10:28
More ARMs than arms: From Sunk to Silicon Supremacy

Every second, over a thousand ARM microprocessors are manufactured; in total, more than 320 billion have been shipped – far exceeding the estimated 120 billion humans who have ever lived. Given that most people have two arms and now carry around 40 ARMs, this is a no-contest “arms race” won by silicon. A pivotal factor in that success was an instruction set called Thumb, “the really useful bit on the end of your ARM .” Conceived 30 years ago on a train to a Japanese ski resort – following a disastrous meeting with Nintendo – Thumb was born of both necessity and audacity. Three weeks later, it was hastily presented to Nokia in a last-ditch attempt to convince them that a chip which wouldn’t exist for another 12 months was exactly the one they needed for their next generation mobile phone. The design rejected ARM ’s heritage as a CPU for computers and instead targeted the power- and cost-sensitive embedded space – a gamble that ultimately unlocked the high-volume markets ARM needed to survive. This talk explores Thumb’s origins, its technical design, and critical role in ARM ’s commercial breakthrough, along with its enduring legacy in today’s ubiquitous, low-power, digital world.

Biography:

Dave Jaggar joined ARM in 1991 and spent nine years transforming the architecture that would become the foundation of modern embedded computing. He authored the first ARM Architecture Reference Manual, formalizing the architecture and introducing Thumb, a second instruction set that enabled ARM ’s widespread adoption in low-power, high-volume devices. Jaggar also pioneered on-chip debug support, restructured the architecture to support full operating systems, and created a new floating-point instruction set. As the founding director of the ARM Austin Design Center, he helped expand the company’s global technical footprint. He holds 29 US patents and is co-recipient of the 2019 IEEE /RSE James Clerk Maxwell Medal for groundbreaking contributions to computer architecture.

Add to your calendar or Include in your list

Wed 03 Dec 14:30: Title to be confirmed

Tue, 27/05/2025 - 10:02
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Wed 19 Nov 14:30: Title to be confirmed

Tue, 27/05/2025 - 09:50
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Wed 28 May 13:30: Correlations of the Möbius function

Tue, 27/05/2025 - 08:12
Correlations of the Möbius function

I will survey recent progress on Chowla’s conjecture on the correlations of the Möbius function. I will discuss some of the methods that have been used to approach this conjecture, and how these methods have turned out to be useful also for some other problems in additive combinatorics and ergodic theory.

Add to your calendar or Include in your list

Thu 19 Jun 13:00: Hope in Hard Places: Bridging the Cancer Care Gap in Resource Limited Settings: Lessons and Innovations from the Uganda Cancer Institute

Tue, 27/05/2025 - 08:08
Hope in Hard Places: Bridging the Cancer Care Gap in Resource Limited Settings: Lessons and Innovations from the Uganda Cancer Institute

Initial support by the British Empire Cancer Campaign in Uganda led to the description of Burkitt Lymphoma and subsequently to the establishment of the Uganda Cancer Institute (UCI). The Uganda Cancer Institute was established in 1967 as a result of a collaboration between Makerere University, the Ministry of Health and the US National Cancer Institute. It was established as a treatment centre for the then recently discovered Burkitt Lymphoma, and was expanded in 1969 to cater for all cancer. The Institute participated in the seminal initial studies on combination chemotherapy. However, years of political turmoil led to a steady decline in care, research and training. Over the past ten years, the UCI has been building capacity address the cancer care gap and here we describe some of the steps taken towards this effort. The Institute has expanded clinical care capacity, increase human resource capacity and is currently building a cancer research and innovation facility. The Institute is undertaking high quality research and here we describe Our model could also serve other developing countries in building capacity for cancer care and research to address the growing burden of cancer in LMI Cs.

Add to your calendar or Include in your list

Wed 04 Jun 16:00: Title to be confirmed

Mon, 26/05/2025 - 18:22
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Tue 27 May 14:30: Modularity of certain trianguline Galois representations

Mon, 26/05/2025 - 13:46
Modularity of certain trianguline Galois representations

An unpublished result of Emerton states that every trianguline representation of the absolute Galois group of Q, satisfying certain conditions, arises as a twist of the Galois representation attached to an overconvergent p-adic cuspidal eigenform of finite slope. I will outline a new approach to prove this result by patching trianguline varieties and eigenvarieties for modular forms on GL2 to establish an “R=T” theorem in the setting of rigid analytic spaces. There are several nice consequences to such a theorem, including a new approach to deduce the classicality of overconvergent eigenforms of small slope, as well as applications to the Fontaine-Mazur conjecture.

Add to your calendar or Include in your list

Tue 10 Jun 14:00: Latent Concepts in Large Language Models

Mon, 26/05/2025 - 08:31
Latent Concepts in Large Language Models

Large Language Models (LLMs) have achieved remarkable fluency and versatility—but understanding how they represent meaning internally remains a challenge. In this talk, we explore the emerging science of latent concepts in LLMs: the semantic abstractions implicitly encoded in their internal activations.

We examine how concepts—such as truthfulness, formality, or sentiment—can be represented as low-dimensional structures, discovered through training dynamics, and understood through the lens of linear algebra and associative memory. We discuss the implications for interpretability, robustness, and control, including how concepts can be steered at test time to adjust model behavior without retraining. Specifically, we explore empirical and theoretical evidence supporting the linear representation hypothesis, where such concepts correspond to vectors or affine subspaces, emerging naturally from training dynamics and next-token prediction objectives. We further show that LLMs behave as associative memory systems, retrieving outputs based on latent similarity rather than logical inference. This behavior underlies phenomena such as context hijacking, where semantically misleading prompts can bias the model’s response.

We introduce formal latent concept models that unify these ideas, describe conditions under which concepts are identifiable, and propose learning algorithms for extracting interpretable, controllable representations. We argue that such latent concept modeling offers a principled framework for bridging representation learning with interpretability and model alignment, and offers a promising path toward safer, more controllable, and more trustworthy AI.

Add to your calendar or Include in your list

Fri 13 Jun 15:00: Title to be confirmed

Sun, 25/05/2025 - 23:55
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Fri 20 Jun 15:00: Title to be confirmed

Sun, 25/05/2025 - 23:55
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Fri 17 Oct 15:00: Title to be confirmed

Sun, 25/05/2025 - 23:55
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Fri 30 May 15:00: Statistical Finite Elements via Interacting Particle Langevin Dynamics

Sun, 25/05/2025 - 23:55
Statistical Finite Elements via Interacting Particle Langevin Dynamics

In this work, we develop a class of interacting particle Langevin algorithms to solve inverse problems for partial differential equations (PDEs). We leverage the statistical finite elements formulation to obtain a finite-dimensional statistical model, where the parameter is that of the forward map and the latent variable is the discretised solution of the PDE , assumed to be partially observed. We then adapt a recently proposed expectation-maximisation like scheme, the interacting particle Langevin algorithm (IPLA), for this problem and obtain a joint estimation procedure for the parameters and the latent variables. The estimation of an unknown source term is demonstrated for linear and nonlinear Poisson PDEs, as well as the diffusivity parameter for the linear Poisson PDE . We provide computational complexity estimates for forcing estimation in the linear case, including comprehensive numerical experiments and preconditioning strategies that significantly improve the performance.

Add to your calendar or Include in your list

Latest news

We are hiring!

4 January 2021

We are seeking to hire a research assistant to work on carbon nanotube based microdevices. More information is available here: www.jobs.cam.ac.uk/job/28202/

We are Hiring!

4 January 2021

We are seeking to hire a postdoc researcher to work on the structuring of Li-ion battery electrodes. More information is available here: www.jobs.cam.ac.uk/job/28197/