Albert Cohen

Albert Cohen

Auto-generated profile, based on publicly available information

Albert Cohen

Google, Paris, France

Research Focus & Metrics

Academic Career Length

25Years (2002-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 February 2021
Verified
Conference Paper
Closed access
CLChris Lattner
Chris Lattner
MAMehdi Amini
Mehdi Amini
UBUday Bondhugula
Uday Bondhugula
ACAlbert Cohen
Albert Cohen
ADAndy Davis
Andy Davis
OZOleksandr Zinenko
Oleksandr Zinenko
This work presents MLIR, a novel approach to building reusable and extensible compiler infrastructure. MLIR addresses software fragmentation, compilation for heterogeneous hardware, significantly reducing the cost of building domain specific compilers, and connecting existing compilers together. MLIR facilitates the design and implementation of code generators, translators and optimizers at different levels of abstraction and across application domains, hardware targets and execution environments. The contribution of this work includes (1) discussion of MLIR as a research artifact, built for extension and evolution, while identifying the challenges and opportunities posed by this novel design, semantics, optimization specification, system, and engineering. (2) evaluation of MLIR as a generalized infrastructure that reduces the cost of building compilers-describing diverse use-cases to show research and educational opportunities for future programming languages, compilers, execution environments, and computer architecture. The paper also presents the rationale for MLIR, its original design principles, structures and semantics.

1 June 2006 in International Journal of Parallel Programming
Verified
Journal Article
Closed access
PDF file
SGSylvain Girbal
Sylvain Girbal
NVNicolas Vasilache
Nicolas Vasilache
CBCédric Bastoul
Cédric Bastoul
ACAlbert Cohen
Albert Cohen
DPDavid Parello
David Parello
OTOlivier Temam
Olivier Temam

7 June 2008
Verified
Conference Paper
Open access
LPLouis-Noël Pouchet
Louis-Noël Pouchet
CBCédric Bastoul
Cédric Bastoul
ACAlbert Cohen
Albert Cohen
JCJohn Cavazos
John Cavazos
High-level loop optimizations are necessary to achieve good perfor- mance over a wide variety of processors. Their performance impact can be significant because they involve in-depth program transfor- mations that aim to sustain a balanced workload over the computa- tional, storage, and communication resources of the target architec- ture. Therefore, it is mandatory that the compiler accurately mod- els the target architecture as well as the effects of complex code restructuring. However, because optimizing compilers (1) use simplistic per- formance models that abstract away many of the complexities of modern architectures, (2) rely on inaccurate dependence analysis, and (3) lack frameworks to express complex interactions of trans- formation sequences, they typically uncover only a fraction of the peak performance available on many applications. We propose a complete iterative framework to address these issues. We rely on the polyhedral model to construct and traverse a large and expres- sive search space. This space encompasses only legal, distinct ver- sions resulting from the restructuring of any static control loop nest. We first propose a feedback-driven iterative heuristic tailored to the search space properties of the polyhedral model. Though, it quickly converges to good solutions for small kernels, larger bench- marks containing higher dimensional spaces are more challenging and our heuristic misses opportunities for significant performance improvement. Thus, we introduce the use of a genetic algorithm with specialized operators that leverage the polyhedral representa- tion of program dependences. We provide experimental evidence that the genetic algorithm effectively traverses huge optimization spaces, achieving good performance improvements on large loop nests with complex memory accesses.

11 October 2019 in ACM Transactions on Architecture and Code Optimization
Verified
Journal Article
Open access
PDF file
NVNicolas Vasilache
Nicolas Vasilache
OZOleksandr Zinenko
Oleksandr Zinenko
TTTheodoros Theodoridis
Theodoros Theodoridis
PGPriya Goyal
Priya Goyal
ZDZachary Devito
Zachary Devito
ACAlbert Cohen
Albert Cohen
Deep learning frameworks automate the deployment, distribution, synchronization, memory allocation, and hardware acceleration of models represented as graphs of computational operators. These operators wrap high-performance libraries such as cuDNN or NNPACK. When the computation does not match any predefined library call, custom operators must be implemented, often at high engineering cost and performance penalty, limiting the pace of innovation. To address this productivity gap, we propose and evaluate: (1) a domain-specific language with a tensor notation close to the mathematics of deep learning; (2) a Just-In-Time optimizing compiler based on the polyhedral framework; (3) carefully coordinated linear optimization and evolutionary algorithms to synthesize high-performance CUDA kernels; (4) the transparent integration of our flow into PyTorch and Caffe2, providing the fully automatic synthesis of high-performance GPU kernels from simple tensor algebra. The performance is comparable to, and often exceeds the performance of, highly tuned libraries.

1 January 2005
Verified
Book Chapter
Closed access
PDF file
GFGrigori Fursin
Grigori Fursin
ACAlbert Cohen
Albert Cohen
MOMichael O’boyle
Michael O’boyle
OTOlivier Temam
Olivier Temam

24 August 2014
Verified
Conference Paper
Open access
UBUday Bondhugula
Uday Bondhugula
VBVinayaka Bandishti
Vinayaka Bandishti
ACAlbert Cohen
Albert Cohen
GPGuillain Potron
Guillain Potron
NVNicolas Vasilache
Nicolas Vasilache
This paper deals with optimizing time-iterated computations on periodic data domains. These computations are prevalent in computational sciences, particularly in partial differential equation solvers. We propose a fully automatic technique suitable for implementation in a compiler or in a domain-specific code generator for such computations. Dependence patterns on periodic data domains prevent existing algorithms from finding tiling opportunities. Our approach augments a state-of-the-art parallelization and locality-enhancing algorithm from the polyhedral framework to allow time-tiling of stencil computations on periodic domains. Experimental results on the swim SPEC CPU2000fp benchmark show a speedup of 5× and 4.2× over the highest SPEC performance achieved by native compilers on Intel Xeon and AMD Opteron multicore SMP systems, respectively. On other representative stencil computations, our scheme provides performance similar to that achieved with no periodicity, and a very high speedup is obtained over the native compiler. We also report a mean speedup of about 1.5× over a domain-specific stencil compiler supporting limited cases of periodic boundary conditions. To the best of our knowledge, it has been infeasible to manually reproduce such optimizations on swim or any other periodic stencil, especially on a data grid of two-dimensions or higher.