ARCO - Microarquitectura i Compiladors
La principal finalitat del grup és investigar en el camp del disseny arquitectònic dels futurs processadores tenint en compte especialment els nous aspectes tecnològics que el condicionaran, tant per processadors d'alt rendiment com pels destinats a l'electrònica de consum.
El grup continua estant format per la mateixa base de professors doctors que ja vam obtenir l'ajut en les convocatòries SGR tant l'any 2005 com el 2009. Està format per professors del Dpt. d'Arquitectura de Computadors de la UPC juntament amb dos professors de la URV. A aquesta base s'hi ha afegit respecte la darrera convocatòria 4 nous membres doctors amb contracte post-doctoral. La variabilitat cal buscar-la en els membre pre-doctors, on es donen de baixa aquells que ja han acabat la tesi i es donen d'alta els nous estudiants de doctorat.
Volem destacar que el grup ha produït al voltant de 500 articles en congressos i revistes d'alt impacte, més de 100 patents i ha dirigit mes de 25 tesis.
The research activities of the ARCO group focus on computer architecture, compilers and parallel processing, with special emphasis on microarchitecture and code generation techniques for energy-efficient and reliable computing systems. One of its main current focuses is on intelligent computing systems, where the goal is to devise novel architectures for computer vision, language processing and graphics processors.
The research activities of the ARCO group focus on computer architecture, compilers and parallel processing, with special emphasis on microarchitecture and code generation techniques for energy-efficient and reliable computing systems. One of its main current focuses is on intelligent computing systems, where the goal is to devise novel architectures for computer vision, language processing and graphics processors.
Col·leccions
-
Articles de revista [46]
-
Reports de recerca [11]
Enviaments recents
-
Visibility rendering order: Improving energy efficiency on mobile GPUs through frame coherence
(2018-08-20)
Article
Accés obertDuring real-time graphics rendering, objects are processed by the GPU in the order they are submitted by the CPU, and occluded surfaces are often processed even though they will end up not being part of the final image, ... -
A novel register renaming technique for out-of-order processors
(Institute of Electrical and Electronics Engineers (IEEE), 2018)
Text en actes de congrés
Accés restringit per política de l'editorialModern superscalar processors support a large number of in-flight instructions, which requires sizeable register files. Conventional register renaming techniques allocate a new storage location, i.e. physical register, for ... -
Strategies to enhance the 3T1D-DRAM cell variability robustness beyond 22 nm
(2014-10-01)
Article
Accés obert3T1D cell has been stated as a valid alternative to be implemented on L1 memory cache to substitute 6T, highly affected by device variability as technology dimensions are reduced. In this work, we have shown that 22 nm ... -
MeRLiN: Exploiting dynamic instruction behavior for fast and accurate microarchitecture level reliability assessment
(Association for Computing Machinery (ACM), 2017)
Text en actes de congrés
Accés obertEarly reliability assessment of hardware structures using microarchitecture level simulators can effectively guide major error protection decisions in microprocessor design. Statistical fault injection on microarchitectural ... -
HW/SW co-designed processors: Challenges, design choices and a simulation infrastructure for evaluation
(Institute of Electrical and Electronics Engineers (IEEE), 2017)
Text en actes de congrés
Accés obertImproving single thread performance is a key challenge in modern microprocessors especially because the traditional approach of increasing clock frequency and deep pipelining cannot be pushed further due to power constraints. ... -
Removing checks in dynamically typed languages through efficient profiling
(Institute of Electrical and Electronics Engineers (IEEE), 2017)
Text en actes de congrés
Accés obertDynamically typed languages increase programmer's productivity at the expense of some runtime overheads to manage the types of variables, since they are not declared at compile time and can change at runtime. One of the ... -
Suitability of FinFET introduction into eDRAM cells for operate at sub-threshold level
(Institute of Electrical and Electronics Engineers (IEEE), 2017)
Comunicació de congrés
Accés restringit per política de l'editorialThis paper explores the feasibility, in terms of performance and reliability, of gain-cell embedded DRAM (eDRAM) to be operative at sub-threshold range, when they are implemented with 10 nm FinFET devices. The use of ... -
Low-power automatic speech recognition through a mobile GPU and a Viterbi accelerator
(2017-04-12)
Article
Accés obertAutomatic speech recognition (ASR) has become a core technology for mobile devices. Delivering real-time and accurate ASR has a huge computational cost, which is challenging to achieve in tightly energy-constrained platforms ... -
UNFOLD: a memory-efficient speech recognizer using on-the-fly WFST composition
(Association for Computing Machinery (ACM), 2017)
Text en actes de congrés
Accés restringit per política de l'editorialAccurate, real-time Automatic Speech Recognition (ASR) requires huge memory storage and computational power. The main bottleneck in state-of-the-art ASR systems is the Viterbi search on a Weighted Finite State Transducer ... -
An ultra low-power hardware accelerator for acoustic scoring in speech recognition
(Institute of Electrical and Electronics Engineers (IEEE), 2017)
Text en actes de congrés
Accés restringit per política de l'editorialAccurate, real-time Automatic Speech Recognition (ASR) comes at a high energy cost, so accuracy has often to be sacrificed in order to fit the strict power constraints of mobile systems. However, accuracy is extremely ... -
Keeping control transfer instructions out of the pipeline in architectures without condition codes
(1987-05)
Report de recerca
Accés obertThe execution of branch instructions involves a loss of performance in pipelined processors. In this paper we present a mechanism for executing this kind of instruction with a zero delay. This mechanism has been proposed ... -
Branch classification to control instruction fetch in simultaneous multithreaded architectures
(Institute of Electrical and Electronics Engineers (IEEE), 2002)
Text en actes de congrés
Accés obertIn simultaneous multithreaded architectures many separate threads are running concurrently, sharing processor resources, thereby realizing a high utilization rate of the available hardware. However, this also implies that ...
RSS 1.0