How Much Randomness Makes a Tool Randomized?

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21240%2F11%3A00180827" target="_blank" >RIV/68407700:21240/11:00180827 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

How Much Randomness Makes a Tool Randomized?

Popis výsledku v původním jazyce

Most of presently used academic logic synthesis tools, including SIS and ABC, are fully deterministic. Up to the knowledge of the authors, this holds for all available commercial tools as well. This means that no random decisions are made; the algorithmsfully rely on deterministic heuristics. In this paper we present several hints of insufficiency of such an approach and show examples of perspective randomized logic synthesis algorithms. Judging from our experiments, these algorithms have a higher potential of performing better than the deterministic ones. Further we study how much randomness is actually needed for the algorithms to perform well. We show that some algorithms require only a small amount of randomness, while still taking full advantageof their randomized nature. On the other hand, some algorithms require a very high level of randomness to perform well. We propose reasons for this behavior and show a way of computing the necessary measure of randomness required.

Název v anglickém jazyce

How Much Randomness Makes a Tool Randomized?

Popis výsledku anglicky

Most of presently used academic logic synthesis tools, including SIS and ABC, are fully deterministic. Up to the knowledge of the authors, this holds for all available commercial tools as well. This means that no random decisions are made; the algorithmsfully rely on deterministic heuristics. In this paper we present several hints of insufficiency of such an approach and show examples of perspective randomized logic synthesis algorithms. Judging from our experiments, these algorithms have a higher potential of performing better than the deterministic ones. Further we study how much randomness is actually needed for the algorithms to perform well. We show that some algorithms require only a small amount of randomness, while still taking full advantageof their randomized nature. On the other hand, some algorithms require a very high level of randomness to perform well. We propose reasons for this behavior and show a way of computing the necessary measure of randomness required.

Druh

O - Ostatní výsledky

CEP obor

JC - Počítačový hardware a software

OECD FORD obor

—

Projekt

<a href="/cs/project/GA102%2F09%2F1668" target="_blank" >GA102/09/1668: Zvyšování spolehlivosti a provozuschopnosti v obvodech SoC</a><br>

Návaznosti

Z - Vyzkumny zamer (s odkazem do CEZ)

Rok uplatnění

2011

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů