Conveners
Lightening talks
- Maria Dadarlat
-
Dr Deep Chatterjee (Massachusetts Institute of Technology)07/10/2022, 12:35
As gravitational-wave (GW) detectors become more sensitive and probe ever more distant reaches, the number of detected binary neutron star mergers will increase. However, detecting more events farther away with GWs does not guarantee corresponding increase in the number of electromagnetic counterparts of these events. Current and upcoming wide-field surveys that participate in GW follow-up...
Go to contribution page -
Will Benoit07/10/2022, 12:38
The successful electromagnetic observation of the neutron star merger GW170817 led to explosive growth in the field of multi-messenger astronomy. With that growth has come new challenges and opportunities. The computational needs of gravitational-wave astronomy have risen alongside the sensitivity of the global network of gravitational-wave detectors, and will continue to rise as more...
Go to contribution page -
Sukanya Krishna (Univ. of California San Diego (US))07/10/2022, 12:41
At the LHC, the FPGA-based real-time data filter system that rapidly decides which collision events to record, known as the level-1 trigger, requires small models because of the low latency budget and other computing resource constraints. To enhance the sensitivity to unknown new physics, we want to put generic anomaly detection algorithms into the trigger. Past research suggests that graph...
Go to contribution page -
Megan Hope Lipton07/10/2022, 12:44
Rodents rely on proprioceptive information from the periphery to guide and coordinate precise forelimb and hindlimb movements, a process called sensorimotor integration. The mouse primary somatosensory (S1) and primary motor (M1) cortices are known to be necessary for adapting motor commands to new sensory environments, and recent work suggests neurons in the forelimb area of S1 encode...
Go to contribution page -
Shikun Liu07/10/2022, 12:47
The high instantaneous luminosity of the CERN Large Hadron Collider leads to multiple proton-proton interactions in the same or nearby bunch crossings (pileup). Advanced pileup mitigation algorithms are designed to remove this noise from pileup particles and improve the performance of crucial physics observables. This study implements a semi-supervised graph neural network for particle-level...
Go to contribution page -
Xiaohan Liu07/10/2022, 12:50
A specific type of Electroencephalography (EEG) signal, sleep spindle, is believed to contribute to neuronal plasticity and memory consolidation. In this project, we proposed a system that is based on ultra-low latency and power FPGA to detect and interact with the sleep spindles to help neuroscientists to understand the mechanism behind the theory. The proposed system will have a programmed...
Go to contribution page -
Simon Rothman (Massachusetts Inst. of Technology (US))07/10/2022, 12:53
Many high-energy physics analyses rely on various machine learning models for both event reconstruction and signal/background discrimination. One of the major sources of systematic uncertainty in these analyses is due to residual mismodelling in the detailed simulation samples used to train these algorithms. In this work, we will discuss a novel approach to correcting for these systematic...
Go to contribution page -
Rohan Shenoy (Univ. of California San Diego (US))07/10/2022, 12:56
The High Granularity Calorimeter (HGCAL) is part of the High Luminosity upgrade of the CMS detector at the Large Hadron Collider (HL-LHC). For the trigger primitive generation of the 6 million channels in this detector, data compression at the front end may be accomplished by using deep-learning techniques using an on-ASICs network. The Endcap Trigger Concentrator (ECON-T) ASIC foresees an...
Go to contribution page -
Andrew Skivington (University of California-San Diego, Duarte Lab)07/10/2022, 12:59
The Large Hadron Collider has recently started Run 3, which means protons are being collided at an astonishing energy of 13.6 TeV. Within the LHC is the CMS detector. Moreover, the Level one trigger is an integral part of the CMS detector and in many ways is considered the first responder of the High Level Trigger system. Its job is to make important initial cuts to the massive amounts of...
Go to contribution page -
Hanchen Ye07/10/2022, 13:02
Efficient coarse-grained dataflow and well-organized data movement are essential for large-scale HLS (High-level Synthesis) designs to achieve promising performance and energy efficiency. Traditional HLS tools typically rely on designers to empirically specify the desired dataflow and data movement scheme and iteratively tune the design choices until reaching a satisfactory result. Although...
Go to contribution page -
Jialiang Zhang07/10/2022, 13:05
The fast-growing complexity of new applications and new use scenarios poses serious challenges for computing systems. Heterogeneous systems consist of different types of processors and accelerators, and provide unique combined benefits of hardware acceleration from each individual component. CPU-FPGA heterogeneous systems provide both programmable logic and general-purpose processors, and they...
Go to contribution page -
Zhixing "Ethan" Jiang (University of Washington (US)), Elham E Khoda (University of Washington (US))07/10/2022, 13:08
The transformer become widely use for the Natural language processing (NLP) task. Besides the NLP tasks, the transformer could be used to analyze any other time series signal processing data, such as images prediction, sensor detection, or glitch detection.
Go to contribution page
However, most of studies on transformer were implemented on GPU. The implementation of transformer on the FPGA was quite undiscovered. In...