blob: 91fcb6e1f94169f655e32fef99d9874da16ece45 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
|
\subtitle{Selected Projects}{\faSuitcase}
\begin{projects}
\project{Since 2026} {Morphosphere} {Distributed Data Architecture for Interactive 3D/4D Tomography Repositories}
\action{\since{2025}} {} {Co-authored proposal; PI / work-package lead for data reduction and organization}
\action{\since{2026}} {} {Project targets a data lifecycle spanning data-generation facilities, HPC centres \& users}
\action{\since{2026}} {} {Exploring compression, caching, and abstractions to enable interactive data access}
\action{\since{2026}} {} {Co-designing DAG tomography framework; evaluating NVIDIA Graphs/Holoscan}
% Canonical, Electronics, Physicts
\project{Since 2018} {KaaS} {KATRIN-as-a-Service (KaaS) platform for distributed data acquisition systems}
\action{\since{2018}} {4} {Initiated platform, secured funding, built team of 4, and led production rollout}
\action{\since{2022}} {} {Coordinated integration across Canonical, KIT HPC, electronics \& experiment teams}
\action{\since{2018}} {4} {KaaS integrates data from fast detectors and thousands of monitoring sensors}
\action{\since{2022}} {1} {Led development of containerized AF\_XDP data-acquisition framework}
\action{\ivl{2018}{2019}} {} {Cut costs by adapting legacy software for distributed ops and integrating HPC resources}
% \action{\since{2020}} {3} {Modeled data access patterns and guided research on organization of time-series data}
% \action{\since{2015}} {1} {Supervised development of online monitoring framework for the KaaS ecosystem}
% \action{\ivl{2019}{2020}} {3} {Coordinated migration of the KATRIN services to the KaaS platform}
% \action{\since{2020}} {2} {Lead research on cloud technologies for heterogeneous control systems}
\project{\ivl{2017}{2022}} {Toma} {Uncovering \& Exploiting GPU Architectural Features}
\action{\ivl{2017}{2018}} {} {Identified performance imbalances and hidden parallelism in NVIDIA and AMD GPUs}
\action{\ivl{2017}{2018}} {} {Developed method exploiting intra-SM parallelism on NVIDIA (up to 30\% speed-up)}
\action{\ivl{2017}{2018}} {} {Accelerated tomography 4–6× through architecture-aware resource balancing}
\action{\ivl{2019}{2020}} {} {Published paper "\urlref{Reviewing GPU architectures to build efficient back projection...}{https://doi.org/10.1007/s11554-019-00883-w}"}
\project{\ivl{2016}{2017}} {CMS} {GPU-based L1 Trigger for CERN CMS using GPUDirect (6~$\mu$s latency budget)}
\project{\ivl{2011}{2017}} {UFO} {Coordinated software work-package and led algorithm parallelization in a project "Ultrafast tomography with online monitoring and image-based control"}
% \project{\ivl{2011}{2016}} {ADEI/ASTOR} {Web-platforms for visualization of volumetric (tomographic) and time-series data}
% \action{\ivl{2011}{2014}} {4} {Developed Platform for exploration and analysis of time-series in terascale archives}
% \project{\ivl{2009}{2013}} {GPU} {Parallel algorithms and software optimization (GPGPU, SIMD): 10-100x speed-up}
% \project{\ivl{2003}{2006}} {PHD} {PhD in Computer Engineering on communication protocols for distributed DAQ}
% \project{\ivl{1999}{2008}} {DAQ} {Distributed Data Acquistion Systems for particle detector networks}
\end{projects}
|
