OpenCL | EPFL Graph Search

OpenCL (Open Computing Language) est la combinaison d'une API et d'un langage de programmation dérivé du C, proposé comme un standard ouvert par le Khronos Group. OpenCL est conçu pour programmer des systèmes parallèles hétérogènes comprenant par exemple à la fois un CPU multi-cœur et un GPU. OpenCL propose donc un modèle de programmation se situant à l'intersection naissante entre le monde des CPU et des GPU, les premiers étant de plus en plus parallèles, les seconds étant de plus en plus programmables. Présentation OpenCL distingue le processeur hôte (processeur central faisant office de chef d'orchestre) des périphériques (CPU, GPU, ou autre) dont la mission est d'exécuter des noyaux de calcul intensifs. OpenCL distingue donc d'un côté l'application tournant sur le processeur hôte (et qui va appeler l'API OpenCL), et de l'autre côté les noyaux qui sont programmés en OpenCL-C (et dont la vocation est d'être exécuté sur les périphériques). OpenCL permet d'exprimer aiséme

Characterization of the paracrine stimulation potential of human CAP cells for cardiac repair

Xavier Adam

According to the last WHO numbers, cardiac diseases remain the first cause of death in the world. Cell therapy represents a promising approach to restore heart function. Cardiac derived adherent proliferation (CAP) cells were shown to support the heart regeneration by unexplained mechanisms when injected in a mouse myocarditis model. Not differentiating themselves in cardiomyocytes and expressing cellular markers for angiogenesis, they are now supposed to have a paracrine effect on heart resident cell populations. In the present work, this paracrine effect had to be characterized on a model cell line. A paracrine effect test system based the P19Cl6 teratocarcinoma cell line had to be established and used to carry out a first screening of paracrine interactions on CAP cells and mesenchymal stem cells (MSC) from bone marrow as reference. Initially, critical parameter of the test system had to be adjusted. Specific marker must be identified and their detection had to be proven possible in the designed test system. Suitable serum lots (k,g) have been identified on the base of growth rates and cardiac differentiation potential (beating activity, Gata-4, Myl2, Mef2c gene expression) using standard protocol (ɑMEM, 1% DMSO). Pretreatment of P19Cl6 with 10 [micro]M 5-azacytidine substantially raised the response to known potent cardiomyogenic inductors (DMSO, Oxytocin). Beating activity was remarkedly elevated, as well as Gata4 (5 to 300 fold), Myl2 (2 to 300 fold), and MEF2C (3 to 4 fold) expression induced compared to non-treated culture. Troponin I ELISA-(2 to 5 fold) and Connexin 43 imunnohistochemistry confirmed the reliable induction of cardiac differentiation in pretreated culture. Based on those results, the test system was conducted using 5-azacytidine pretreated P19Cl6 cells and explored their reactions to different direct (cell to cell in coculture) and indirect (secreted factors in preconditioned medium) interactions with CAP cells or MSC. Since no reference setup existed, several coculture compositions with fixed total cell number (10-90%) or fixed P19Cl6 cell number (10-1000%) were tested, as well as exposition to differently conditioned media (24h, 72h) in different dilutions (100%, 50%, 33%). Beating activity was almost never observed. However, CAP cell experiments (n=6) showed a significant increase in P19Cl6 GATA-4 expression compared to control P19Cl6 (3.3 fold). This expression was also increased in MSC analyses (n=4) but in small amount (1.8 fold). Screening of CAP and MSC interactions led to the identification of active/effective compositions and dilutions. In addition, aza pretreatment proved itself enhancing the sensitivity of the test system. To sum up, the results suggested that CAP cells as wells as MSC have a paracrine stimulation potential and gave starting points for deeper analyses.

2011

Enhancing Design Space Exploration by Extending CPU/GPU Specifications onto FPGAs

Andreas Peter Burg, João Maria Duarte Andrade, Gabriel Falcao Paiva Fernandes, Paolo Ienne, Georgios Karakonstantis, David Novo Bruna, Madhura Purnaprajna

The design cycle for complex special-purpose computing systems is extremely costly and time-consuming. It involves a multiparametric design space exploration for optimization, followed by design verification. Designers of special purpose VLSI implementations often need to explore parameters, such as optimal bitwidth and data representation, through time-consuming Monte Carlo simulations. A prominent example of this simulation-based exploration process is the design of decoders for error correcting systems, such as the Low-Density Parity-Check (LDPC) codes adopted by modern communication standards, which involves thousands of Monte Carlo runs for each design point. Currently, high-performance computing offers a wide set of acceleration options that range from multicore CPUs to Graphics Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs). The exploitation of diverse target architectures is typically associated with developing multiple code versions, often using distinct programming paradigms. In this context, we evaluate the concept of retargeting a single OpenCL program to multiple platforms, thereby significantly reducing design time. A single OpenCL-based parallel kernel is used without modifications or code tuning on multicore CPUs, GPUs, and FPGAs. We use SOpenCL (Silicon to OpenCL), a tool that automatically converts OpenCL kernels to RTL in order to introduce FPGAs as a potential platform to efficiently execute simulations coded in OpenCL. We use LDPC decoding simulations as a case study. Experimental results were obtained by testing a variety of regular and irregular LDPC codes that range from short/medium (e.g., 8,000 bit) to long length (e.g., 64,800 bit) DVB-S2 codes. We observe that, depending on the design parameters to be simulated, on the dimension and phase of the design, the GPU or FPGA may suit different purposes more conveniently, thus providing different acceleration factors over conventional multicore CPUs.

Assoc Computing Machinery2015

Shortening design time through multiplatform simulations with a portable OpenCL golden-model: the LDPC decoder case

Andreas Peter Burg, Paolo Ienne, Georgios Karakonstantis, David Novo Bruna, Madhura Purnaprajna

Hardware designers and engineers typically need to explore a multi-parametric design space in order to ﬁnd the best conﬁguration for their designs using simulations that can take weeks to months to complete. For example, designers of special purpose chips need to explore parameters such as the optimal bitwidth and data representation. This is the case for the development of complex algorithms such as Low-Density Parity-Check (LDPC) decoders used in modern communication systems. Currently, high-performance computing offers a wide set of acceleration options, that range from multicore CPUs to graphics processing units (GPUs) and FPGAs. Depending on the simulation requirements, the ideal architecture to use can vary. In this paper we propose a new design ﬂow based on OpenCL, a uniﬁed multiplatform programming model, which accelerates LDPC decoding simulations, thereby signiﬁcantly reducing architectural exploration and design time. OpenCL-based parallel kernels are used without modiﬁcations or code tuning on multicore CPUs, GPUs and FPGAs. We use SOpenCL (Silicon to OpenCL), a tool that automatically converts OpenCL kernels to RTL for mapping the simulations into FPGAs. To the best of our knowledge, this is the ﬁrst time that a single, unmodiﬁed OpenCL code is used to target those three different platforms. We show that, depending on the design parameters to be explored in the simulation, on the dimension and phase of the design, the GPU or the FPGA may suit different purposes more conveniently, providing different acceleration factors. For example, although simulations can typically execute more than 3× faster on FPGAs than on GPUs, the overhead of circuit synthesis often outweighs the beneﬁts of FPGA-accelerated execution.