Sun Microsystems

By Elena Loutskina, E. Varney. 2017

300.twolf 4 204 68.1 4 196 71.1 Hardware CPU: AMD Opteron (TM) 275 CPU MHz: 2200 FPU: Integrated CPU(s) enabled: 4 cores, 2 chips, 2 cores/chip CPU(s) orderable: 1,2 (order by # of chips) Parallel: No Primary Cache: 64KBI + 64KBD (on chip) per core Secondary Cache: 1024KB (I+D) (on chip) per core L3 Cache: N/A Other Cache: N/A Memory: 16GB (8x2GB, PC3200 CL3 DDR ECC Registered SDRAM) Disk Subsystem: SAS, 36GB, 10K RPM Other Hardware: None Software Operating System: Solaris 10 3/05 HW1 Compiler: Sun Studio 11 File System: ufs System State: Multi-user

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Fast ConvNets Using Group-Wise Brain Damage (2015, background): “The reduction allows to use highly optimized implementations of dense matrix multiplications (e.g. variants of BLAS [4] libraries) that have been developed over many years for all possible computing architectures.”
A Survey on Distributed Machine Learning (2019, methodology): “MLLib relies on the Scala linear algebra package Breeze [64], which in turn utilizes netlib-java [99] for optimization, a bridge for libraries like BLAS [20] and LAPACK [9] which are widely used in high-performance computing.”
COIL: Revisit Exact Lexical Match in Information Retrieval with Contextualized Inverted List (2021, methodology): “As shown in Figure 3, in this work we implement COIL’s by stacking vectors in each inverted list I t into a matrix M n t ×| I k | , so that similarity computation that traverses an inverted list and computes vector dot product can be done ef-ﬁciently as one matrix-vector product with optimized BLAS ”
Parallelizing Word2Vec in Shared and Distributed Memory (2016, methodology): “This is a level-1 BLAS operation [14] and is limited by memory bandwidth, thus not utilizing the increasing computational power of modern multi-core and many-core processors.”
TRIQS: A toolbox for research on interacting quantum systems (2015, methodology): “ lines (labelled by x i) and columns (labelled by y j) using the Sherman-Morrison and Woodbury formulas [17,18]. This generic algorithm is implemented in the TRIQS det_manip class, using BLAS Level 2 [19,20] internally. The class provides a simple API, in order to make these manipulations as straigh”
To Index or Not to Index: Optimizing Exact Maximum Inner Product Search (2017, background): “However, in this paper, we show that both of these indexes do not always outperform a hardware-efficient brute-force approach: a dense matrix multiply of the user and item matrices using optimized, cache-efficient Basic Linear Algebra Subprogram (BLAS) libraries [6], [15], [38].”
Inference of direct residue contacts in two-component signaling. (2010, background): “For all pairs, eji eij will also be needed, but as computer linear algebra systems (Blackford et al., 2002; Jones et al., 2001) provide matrix multiplication functions that accept a transposed matrix, it is unnecessary to store these values in memory.”
Model-driven Level 3 BLAS Performance Optimization on Loongson 3A Processor (2012, methodology): “The BLAS (Basic Linear Algebra Subprograms) interface is widely used in scientific computing and provides the building blocks for many numerical libraries [1].”
Machine Learning in Python: Main developments and technology trends in data science, machine learning, and artificial intelligence (2020, methodology): “This includes options such as the cuBLAS, cuSparse, and cuSolver libraries contained in the CUDA Toolkit.”