Mediaspace scheduled maintenance: Aug 25, 2026 07:00 - 12:00 AM. During this time, videos will be temporarily unavailable. Check status updates.

Concept

Single-precision floating-point format

Related publications (28)

Trading off Energy versus Accuracy in Modern Computing Systems : From Digital Circuit Design to Programming Techniques

Jérémy Lucien Maurice Schlachter

The slowdown of Moore's law, which has been the driving force of the electronics industry over the last 5 decades, is causing serious problem to Integrated Circuits (ICs) improvements. Technology scaling is becoming more and more complex and fabrication co ...

EPFL2018

Genius loci e memorie urbane. Il progetto di Ragnar Östberg per l’area di Eldkvarn

Chiara Monterumisi

The paper explores the city of Stockholm, apparently far from the main tracks of other significant European examples due to its peculiar morphological features. The study investigates the Stockholm Town Hall (Stadshuset) and the nearby never-to-be realized ...

2018

Training DNNs with Hybrid Block Floating Point

Babak Falsafi, Martin Jaggi, Tao Lin, Mario Paulo Drumond Lages De Oliveira

The wide adoption of DNNs has given birth to unrelenting computing requirements, forcing datacenter operators to adopt domain-specific accelerators to train them. These accelerators typically employ densely packed full-precision floating-point arithmetic t ...

NEURAL INFORMATION PROCESSING SYSTEMS (NIPS)2018

Adversarially Robust Optimization with Gaussian Processes

Volkan Cevher, Jonathan Mark Scarlett, Ilija Bogunovic

In this paper, we consider the problem of Gaussian process (GP) optimization with an added robustness requirement: The returned point may be perturbed by an adversary, and we require the function value to remain as high as possible even after this perturba ...

2018

Adversarially Robust Optimization with Gaussian Processes

Volkan Cevher, Jonathan Mark Scarlett, Ilija Bogunovic

NEURAL INFORMATION PROCESSING SYSTEMS (NIPS)2018

Evidence of a structure in (K)over-bar(0)Lambda(+)(c) consistent with a charged Xi(c)(2930)(+), and updated measurement of (B)over-bar(0) -> (K)over-bar(0)Lambda(+)(c) (Lambda)over-bar(c)(-) at Belle

Mingkui Wang, Olivier Schneider, Chao Wang, Yiming Li, Sun Hee Kim, Tara Nanut, Jun Yong Kim, Roberto Gargiani, Chang-Hyun Park, Seyed Sina Mirrazavi Salehian, Arnaud Michel Gelb, Yong Chan Choi, Gaétan Charles Antoine Giriat, Linlin Cao, Wondwossen Wubie Eshetu, Jiayun Shen

We report evidence for the charged charmed-strange baryon Xi(c)(2930)(+) with a signal significance of 3.9 sigma with systematic errors included. The charged Xi(c)(2930)(+) is found in its decay to K-S(0)Lambda(+)(c) in the substructure of (B) over bar (0) ...

SPRINGER2018

Mixed-precision architecture based on computational memory for training deep neural networks

Irem Boybat Kara, Abu Sebastian

Deep neural networks (DNN) have revolutionized the field of machine learning by providing unprecedented human-like performance in solving many real-world problems such as image or speech recognition. Training of large DNNs, however, is a computationally in ...

IEEE2018

Logistic regression over encrypted data from fully homomorphic encryption

Zhicong Huang

Background: One of the tasks in the 2017 iDASH secure genome analysis competition was to enable training of logistic regression models over encrypted genomic data. More precisely, given a list of approximately 1500 patient records, each with 18 binary feat ...

2018