Recherche

Afficher tous les résultats pour

Accueil

Personne

Babak Falsafi

À propos
Confidentialité
Mentions légales

Graph Chatbot

Publications associées (89)

Distributed Logless Atomic Durability with Persistent Memory

Babak Falsafi, Alexandros Daglis, Siddharth Gupta

Datacenter operators have started deploying Persistent Memory (PM), leveraging its combination of fast access and persistence for significant performance gains. A key challenge for PM-aware software is to maintain high performance while achieving atomic du ...

2019

SMoTherSpectre: Exploiting Speculative Execution through Port Contention

Babak Falsafi, Atri Bhattacharyya, Mathias Josef Payer

Spectre, Meltdown, and related attacks have demonstrated that kernels, hypervisors, trusted execution environments, and browsers are prone to information disclosure through micro-architectural weaknesses. However, it remains unclear as to what extent other ...

ACM2019

Analog Neural Networks with Deep-submicron Nonlinear Synapses

Pascal Frossard, Babak Falsafi, Ahmet Caner Yüzügüler, Mario Paulo Drumond Lages De Oliveira, Firat Çelik

Deep neural network (DNN) inference tasks are computationally expensive. Digital DNN accelerators offer better density and energy efficiency than general-purpose processors but still not sufficient to be deployable on resource-constrained settings.Analog c ...

2019

LTRF: Enabling High-Capacity Register Files for GPUs via Hardware/Software Cooperative Register Prefetching

Babak Falsafi, Hamid Sarbazi-Azad, Mario Paulo Drumond Lages De Oliveira, Seyed Borna Ehsani

Graphics Processing Units (GPUs) employ large register files to accommodate all active threads and accelerate context switching. Unfortunately, register files are a scalability bottleneck for future GPUs due to long access latency, high power consumption, ...

2018

Atomic object reads for in-memory rack-scale computing

Babak Falsafi, Boris Robert Grot, Alexandros Daglis

A distributed memory system including a plurality of chips, a plurality of nodes that are distributed across the plurality of chips such that each node is comprised within a chip, each node includes a dedicated local memory and a processor core, and each l ...

2018

Training DNNs with Hybrid Block Floating Point

Babak Falsafi, Tao Lin, Mario Paulo Drumond Lages De Oliveira, Martin Jaggi

The wide adoption of DNNs has given birth to unrelenting computing requirements, forcing datacenter operators to adopt domain-specific accelerators to train them. These accelerators typically employ densely packed full-precision floating-point arithmetic t ...

NEURAL INFORMATION PROCESSING SYSTEMS (NIPS)2018

Design Guidelines for High-Performance SCM Hierarchies

Babak Falsafi, Javier Picorel Obando, Alexandros Daglis, Edouard Bugnion, Dmitrii Ustiugov, Mark Johnathon Sutherland, Dionysios Pnevmatikatos

With emerging storage-class memory (SCM) nearing commercialization, there is evidence that it will deliver the much-anticipated high density and access latencies within only a fewfactors of DRAM. Nevertheless, the latency-sensitive nature of memory-residen ...

ACM2018

Near-Memory Address Translation

Babak Falsafi, Dorde Jevdic, Javier Picorel Obando

Memory and logic integration on the same chip is becoming increasingly cost effective, creating the opportunity to offload data-intensive functionality to processing units placed inside memory chips. The introduction of memory-side processing units (MPUs) ...

Ieee2017

The Mondrian Data Engine

Babak Falsafi, Javier Picorel Obando, Boris Robert Grot, Alexandros Daglis, Nooshin Sadat Mirzadeh, Dmitrii Ustiugov, Mario Paulo Drumond Lages De Oliveira, Dionysios Pnevmatikatos

The increasing demand for extracting value out of ever-growing data poses an ongoing challenge to system designers, a task only made trickier by the end of Dennard scaling. As the performance density of traditional CPU-centric architectures stagnates, adva ...

2017

Fat Caches For Scale-Out Servers

Babak Falsafi, Stavros Volos, Dorde Jevdic, Boris Robert Grot

The authors propose a high-capacity cache architecture that leverages emerging high-bandwidth memory modules. High-capacity caches capture the secondary data working sets of scale-out workloads while uncovering significant spatiotemporal locality across da ...

Institute of Electrical and Electronics Engineers2017

Unified prefetching into instruction cache and branch target buffer

Babak Falsafi, Ilknur Cansu Kaynak, Boris Robert Grot

A system and method of coupling a Branch Target Buffer (BTB) content of a BTB with an instruction cache content of an instruction cache. The method includes: tagging a plurality of target buffer entries that belong to branches within a same instruction blo ...

2017

FPGAs versus GPUs in Data centers

Babak Falsafi

This article presents position statements and a question-and-answer session by panelists at the Fourth Workshop on Computer Architecture Research Directions. The subject of the debate was the use of field-programmable gate arrays versus GPUs in datacenters ...

Institute of Electrical and Electronics Engineers2017

Unlocking Energy

Rachid Guerraoui, Babak Falsafi, Javier Picorel Obando, Vasileios Trigonakis

Locks are a natural place for improving the energy efficiency of software systems. First, concurrent systems are mainstream and when their threads synchronize, they typically do it with locks. Second, locks are well-defined abstractions, hence changing the ...

Usenix Assoc2016

The Case for RackOut: Scalable Data Serving Using Rack-Scale Systems

Babak Falsafi, Stanko Novakovic, Boris Robert Grot, Alexandros Daglis, Edouard Bugnion

To provide low latency and high throughput guarantees, most large key-value stores keep the data in the memory of many servers. Despite the natural parallelism across lookups, the load imbalance, introduced by heavy skew in the popularity distribution of k ...

2016

Towards Near-Threshold Server Processors

David Atienza Alonso, Luca Benini, Pablo Garcia del Valle, Babak Falsafi, Javier Picorel Obando, Ali Pahlevan, Arash Pourhabibi Zarandi, Marina Zapater Sancho

The popularity of cloud computing has led to a dramatic increase in the number of data centers in the world. The ever-increasing computational demands along with the slowdown in technology scaling has ushered an era of power-limited servers. Techniques suc ...

IEEE/ACM2016

Near-Memory Data Services

Babak Falsafi

This column introduces seven invited position papers about the challenges and opportunities in near-data processing. "Near-Memory Data Services" by Babak Falsafi, "Automata Processing: The Memory Is the Processor!" by Mircea Stan, Kevin Skadron, "Overcomin ...

Institute of Electrical and Electronics Engineers2016

An Analysis of Load Imbalance in Scale-out Data Serving

Babak Falsafi, Stanko Novakovic, Boris Robert Grot, Alexandros Daglis, Edouard Bugnion

Despite the natural parallelism across lookups, performance of distributed key-value stores is often limited due to load imbalance induced by heavy skew in the popularity distribution of the dataset. To avoid violating service level objectives expressed in ...

2016

SABRes: Atomic Object Reads for In-Memory Rack-Scale Computing

Babak Falsafi, Stanko Novakovic, Boris Robert Grot, Alexandros Daglis, Edouard Bugnion, Dmitrii Ustiugov

Modern in-memory services rely on large distributed object stores to achieve the high scalability essential to service thousands of requests concurrently. The independent and unpredictable nature of incoming requests results in random accesses to the objec ...

Ieee2016

Scale-out non-uniform memory access

Babak Falsafi, Stanko Novakovic, Boris Robert Grot, Alexandros Daglis, Edouard Bugnion

A computing system that uses a Scale-Out NUMA (“soNUMA”) architecture, programming model, and/or communication protocol provides for low-latency, distributed in-memory processing. Using soNUMA, a programming model is layered directly on top of a NUMA memor ...

2015

Asynchronous memory access chaining

Babak Falsafi, Yusuf Onur Koçberber, Boris Robert Grot

In-memory databases rely on pointer-intensive data structures to quickly locate data in memory. A single lookup operation in such data structures often exhibits long-latency memory stalls due to dependent pointer dereferences. Hiding the memory latency by ...

2015

Page 2 sur 5