This article presents a reconfigurable accelerator for REcurrent Neural networks with fine-grained cOlumn-Wise matrix–vector multiplicatioN (RENOWN). We propose novel latency-hiding architecture recurrent neural network (RNN) acceleration using column-wise multiplication (MVM) instead of the state-of-the-art row-wise operation. hardware (HW) can eliminate data dependencies to improve throughput...

The Navigation Office at ESOC has been closely involved in all aspects of the Precise Orbit Determination (POD) of Envisat throughout the mission lifetime. These activities started shortly after launch through membership of the Orbit Validation Team (OVT). During this time ESOC was involved in both calibration and validation of the DORIS doppler data, the SLR data and the different POD solution...

We present Submatrix-wise Vector Embedding Learner (Swivel), a method for generating lowdimensional feature embeddings from a feature co-occurrence matrix. Swivel performs approximate factorization of the point-wise mutual information matrix via stochastic gradient descent. It uses a piecewise loss with special handling for unobserved co-occurrences, and thus makes use of all the information in...

Two of the most promising new remote sensing technologies for increasing accuracy and efficiency of forest inventory are tree-wise or stand-wise measurements of airborne laser scanner data or 2(2D) and 3-dimensional measurements (3D) of digital aerial photographs. The aim of the study was to compare 3D measurements of digital aerial photographs and laser scanning to 2D measurements of digital a...

The classical, complete-information two-player games assume that the problem data (in particular the payoff matrix) is known exactly by both players. In a now famous result, Nash has shown that any such game has an equilibrium in mixed strategies. This result was later extended to a class of incomplete-information two-player games by Harsanyi, who assumed that the payoff matrix is not known exa...

This paper presents an e cient implementation of incomplete LU (ILU) factorizations that are derived from the Crout version of Gaussian elimination (GE). At step k of the elimination, the k-th row of U and the k-th column of L are computed using previously computed rows of U and columns of L. The data structure and implementation borrow from already known techniques used in developing both spar...