Zobrazeno 1 - 10
of 149
pro vyhledávání: '"Kornilios, P."'
Solid-state drives (SSDs) are extensively used to deploy persistent data stores, as they provide low latency random access, high write throughput, high data density, and low cost. Tree-based data structures are widely used to build persistent data st
Externí odkaz:
http://arxiv.org/abs/2006.04658
Autor:
Kourtis, Kornilios, Dazzi, Martino, Ioannou, Nikolas, Grosser, Tobias, Sebastian, Abu, Eleftheriou, Evangelos
Computational memory (CM) is a promising approach for accelerating inference on neural networks (NN) by using enhanced memories that, in addition to storing data, allow computations on them. One of the main challenges of this approach is defining a h
Externí odkaz:
http://arxiv.org/abs/2003.04293
With rapid improvements in NVM storage devices, the performance bottleneck is gradually shifting to the network, thus giving rise to the notion of "data movement wall". To reduce the amount of data movement over the network, researchers have proposed
Externí odkaz:
http://arxiv.org/abs/2002.11528
Autor:
Gerber, Simon, Zellweger, Gerd, Achermann, Reto, Hoffmann, Moritz, Kourtis, Kornilios, Roscoe, Timothy, Milojicic, Dejan
In this paper, we rethink how an OS supports virtual memory. Classical VM is an opaque abstraction of RAM, backed by demand paging. However, most systems today (from phones to data-centers) do not page, and indeed may require the performance benefits
Externí odkaz:
http://arxiv.org/abs/1911.08367
Autor:
Kaufmann, Michael, Kourtis, Kornilios, Mendler-Dünner, Celestine, Schüpbach, Adrian, Parnell, Thomas
Distributed machine learning training is one of the most common and important workloads running on data centers today, but it is rarely executed alone. Instead, to reduce costs, computing resources are consolidated and shared by different application
Externí odkaz:
http://arxiv.org/abs/1909.04885
Akademický článek
Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.
K zobrazení výsledku je třeba se přihlásit.
Autor:
Adam Javorčík, Ilias Strachinis, Evanthia Thanou, Panagiotis Kornilios, Aziz Avcı, Nazan Üzüm, Kurtuluş Olgun, Çetin Ilgaz, Yusuf Kumlutaş, Petros Lymberakis, Zoltán T. Nagy, Daniel Jablonski
Publikováno v:
Diversity, Vol 16, Iss 3, p 184 (2024)
In this study, we investigate the phylogeographic patterns of Dolichophis species in the Aegean region, aiming to elucidate their genetic diversity and putative historical colonisation routes through mitochondrial and nuclear DNA data. Our findings r
Externí odkaz:
https://doaj.org/article/4893fa401c31457697c84c364ebbea08
In this paper we experimentally analyze the convergence behavior of CoCoA and show, that the number of workers required to achieve the highest convergence rate at any point in time, changes over the course of the training. Based on this observation,
Externí odkaz:
http://arxiv.org/abs/1811.02322
In this paper we analyze, evaluate, and improve the performance of training generalized linear models on modern CPUs. We start with a state-of-the-art asynchronous parallel training algorithm, identify system-level performance bottlenecks, and apply
Externí odkaz:
http://arxiv.org/abs/1811.01564
To satisfy increasing storage demands in both capacity and performance, industry has turned to multiple storage technologies, including Flash SSDs and SMR disks. These devices employ a translation layer that conceals the idiosyncrasies of their mediu
Externí odkaz:
http://arxiv.org/abs/1801.05637