Sonia Martín López

Dynamic capacity-speed tradeoffs in SMT processor caches

Caches are designed to provide the best tradeoff between access speed and capacity for a set of target applications. Unfortunately, different applications, and even different phases within the same application, may require a different capacity-speed tradeoff. This problem is exacerbated in a Simultaneous Multi-Threaded (SMT) processor where the optimal cache design may vary drastically with the number of running threads and their characteristics. We propose to make this capacity-speed cache tradeoff dynamic within an SMT core. We extend a previously proposed globally asynchronous, locally synchronous (GALS) processor core with multi-threaded support, and implement dynamically resizable instruction and data caches. As the number of threads and their characteristics change, these adaptive caches automatically adjust from small sizes with fast access times to higher capacity configurations. While the former is more performance-optimal when the core runs a single thread, or a dual-thread workload with modest cache requirements, higher capacity caches work best with most multiple thread workloads. The use of a GALS microarchitecture permits the rest of the processor, namely the execution core, to run at full speed irrespective of the cache speeds. This approach yields an overall performance improvement of 24.7% over the best fixed-size caches for dual-thread workloads, and 19.2% for single-threaded applications.

Improving SMT performance: an application of genetic algorithms to configure resizable caches

Simultaneous Multithreading (SMT) is a technology aimed at improving the throughput of the processor core by applying Instruction Level Parallelism (ILP) and Thread Level Parallelism (TLP). Nevertheless a good control strategy is required when resources are shared among different threads, so that throughput is optimized. We study the application of evolutionary algorithms to improve the allocation of configurations on the cache hierarchy over a Simultaneous Multithreading (SMT) processor. In this way, resizable caches have demonstrated their efficiency by adapting their configuration according to workload settings, at runtime. More-over, some methodologies and a number of techniques, such as dynamic resource allocation, have previously been developed to optimize the cache hit behavior, trying to improve global SMT performance. In this paper we propose the use of a Genetic Algorithm (GA) to optimize dynamically reconfigurable cache designs. Given that different workloads feature different characteristics and needs, we apply a Genetic Algorithm (GA) for cache designing, in order to obtain a better dynamic configuration that increases the number of instructions per cycle (IPC). The obtained results show the feasibility of the approach and the potential of GAs for SMT optimization.

Drowsy cache partitioning for reduced static and dynamic energy in the cache hierarchy

Power consumption is a major concern in todays processor design. As technology shrinks, leakage power dominates the overall power consumption of the processor although it is expected that dynamic power gains relevance in future semiconductor technology. This is particularly relevant for the cache hierarchy, which contains an important percentage of the microprocessor transistors. In this work we propose the use of a phase adaptive cache design to reduce both leakage and dynamic power consumption with very little impact on the overall performance. We take advantage of the overwhelming preference of the memory accesses for the most recently used blocks, and the fact that these blocks are placed in a fast A partition of the cache. On the other hand, a B partition of the cache memory is placed in a drowsy mode in order to reduce the leakage power consumption of an important portion of the whole storage capacity. We test our design on a private second level cache reaching average dynamic energy savings almost 20% of the conventional cache designs dynamic energy, and leakage energy savings close to 45% of the conventional cache designs leakage energy. These results were achieved with minimal performance losses that stay within 2% of the original values.

Performance modeling of pipelined linear algebra architectures on FPGAs

The potential design space of FPGA accelerators is very large. The factors that define the performance of a particular implementation include the architecture design, number of pipelines, and memory bandwidth. In this paper we present a mathematical model that, based on these factors, predicts the computation time of pipelined FPGA accelerators. This model can be used to quickly explore the design space without any implementation or simulation. We evaluate the model, its usefulness, and ability to identify the bottlenecks and improve performance. Being the core of many compute-intensive applications, linear algebra computations are the main contributors to their total execution time. Hence, five relevant linear algebra computations are selected, analyzed, and the accuracy of the model is validated against implemented designs.

Las dificultades de financiación de las empresas de participación ante la crisis económica: la creación de una entidad financiera de crédito como alternativa

La crisis economica esta teniendo una importante repercusion en el tejido empresarial espanol, de tal forma que cada vez son menos el numero de nuevas empresas que se constituyen, y mas aquellas, que ya encontrandose operativas en el mercado, se ven abocadas al cierre, y a su disolucion. Las empresas que se han visto afectadas en mayor medida son la de menor dimension, quienes ven peligrar su viabilidad futura, debido a la falta de liquidez y acceso al credito, como consecuencia de las malas practicas del sector financiero internacional. En el caso de las empresas de participacion, caracterizadas por su reducida dimension, el hecho de ser empresario individual, o socio de una cooperativa de trabajo asociado o sociedad laboral, lleva implicito la condicion de financiador de la misma, actuando como banquero. Por lo que, ante el problema de la falta de liquidez, se plantea como alternativa el estudio de la creacion de una entidad financiera de credito por parte de estas empresas.

Linear algebra computations in heterogeneous systems

One of the main challenges of using heterogeneous systems results from the need to find the computation-to-hardware assignments that maximize the overall application performance. The important computational factors that must be taken into account include algorithmic complexity, exploitable parallelism, memory bandwidth requirements, and data size. To achieve high performance, a hardware platform is chosen to satisfy the needs of a computation with corresponding architectural features such as clock speed, number of parallel computational units, and memory bandwidth. In this paper five linear algebra computations that are commonly found in compute-intensive applications are selected and evaluated in terms of performance on CPU, GPU, and FPGA platforms across a wide range of data sizes. The results are used to provide guidelines to help select the best performing hardware platform based on the computational factors. Using a cutting edge signal processing application as a case study, we demonstrate the importance of making computation assignments for improved performance. Our experimental results show that a properly implemented heterogeneous system achieves a speedup of up to 39x and 3.8x compared to CPU-only and GPU-only systems respectively.

Distributed execution of transmural electrophysiological imaging with CPU, GPU, and FPGA

One of the main challenges of using cutting edge medical imaging applications in the clinical setting is the large amount of data processing required. Many of these applications are based on linear algebra computations operating on large data sizes and their execution may require days in a standard CPU. Distributed heterogeneous systems are capable of improving the performance of applications by using the right computation-to-hardware mapping. To achieve high performance, hardware platforms are chosen to satisfy the needs of each computation with corresponding architectural features such as clock speed, number of parallel computational units, and memory bandwidth. In this paper we evaluate the performance benefits of using different hardware platforms to accelerate the execution of a transmural electro physiological imaging algorithm, targeting a standard CPU with GPU and FPGA accelerators. Using this cutting edge medical imaging application as a case study, we demonstrate the importance of making intelligent computation assignments for improved performance. We show that, depending on the size of the data structures the application works with, the usage of an FPGA to run certain computations can make a big difference: a heterogeneous system with all three hardware platforms (CPU+GPU+FPGA) can cut the execution time by half, compared to the best result using one single accelerator (CPU+GPU). In addition, our experimental results show that combining CPU, GPU, and FPGA platforms in a single system achieves a speedup of up to 62×, 2×, and 1605× compared to systems with a single CPU, GPU, or FPGA platform respectively.

Criterios de eficiencia y buenas prácticas en los procesos de concentración de las organizaciones de participación

En el estudio de la Responsabilidad Social Empresarial en las sociedades cooperativas hay que identificar sus ambitos de aplicacion, asi como los grupos de interes, su particular relacion con la empresa y el tratamiento de la eficiencia como principio empresarial y como combinacion de esfuerzos hacia la Responsabilidad Social. De esta forma a lo largo del presente trabajo se plantea el estudio de la Responsabilidad Social Empresarial para el caso de las sociedades cooperativas, por una parte desde la esfera interna o nucleo que de la propia definicion de Responsabilidad Social atenderia a los trabajadores, los clientes y los proveedores, esto es a los socios; y por otra, desde una esfera mas amplia que atenderia a los competidores, la comunidad local, el medio ambiente y la sociedad en general. Para ello, se analizan como requerimientos fundamentales en el ambito interno, la eficiencia en la participacion de los socios en los flujos empresariales, y en el ambito externo, la concentracion empresarial de las sociedades cooperativas, asi como las variables que se ven modificadas en los procesos de concentracion empresarial en aras de conseguir valores que ensalcen el buen gobierno y la Responsabilidad Social Empresarial. Por ultimo, se tratar de determinar los efectos que los procesos de concentracion tienen en la determinacion del valor de la Responsabilidad Social en las sociedades cooperativas.

El emprendimiento y el empleo a través de la empresa individual: contraste de fuentes estadísticas

El entorno micro y macroeconomico actual supone un importante reto para las microempresas y los trabajadores autonomos en los que la obligada gestion eficiente sera el filtro discriminante en la seleccion de los que han de sobrevivir. La economia esta en un proceso de cambio importantisimo que requiere de empresarios. En la busqueda de la innovacion y la eficiencia, este colectivo empresarial no puede aislarse y debe buscar la implicacion de los diferentes agentes economicos y sociales – asociaciones, universidades, camaras de comercio, la Administracion Publica – y hoy mas que nunca el Sistema Financiero que debe ser corresponsable de que el trabajo autonomo supere la crisis economica sin restringirles la liquidez para desarrollar su actividad teniendo en cuenta sus dificultades adicionales. En orden a facilitar el estudio del colectivo y asi poder analizar sus necesidades este trabajo pretende cuantificarlo contrastando las distintas fuentes estadisticas disponibles y, establece una comparacion entre las mismas, para poder llegar a una distribucion de los trabajadores autonomos emprendedores lo mas precisa posible.

Las empresas de participación ante el reto de la gestión del cambio demográfico

El cambio demografico que se esta viviendo a nivel mundial, y de forma particular en Europa, como consecuencia del envejecimiento de la poblacion ante el incremento de la esperanza de vida y la drastica reduccion de las tasas de fertilidad, ha hecho saltar las alarmas ante la necesidad de llevar a cabo una gestion adecuada que no ponga en peligro la viabilidad financiera de los sistemas de proteccion social. Los Estados Miembros han de acometer las reformas necesarias que conduzcan a la modernizacion de sus sistemas de proteccion social garantizando tanto pensiones adecuadas y viables como una asistencia sanitaria y una asistencia de larga duracion de calidad, accesible y duradera. Para lograr estos objetivos existe el acuerdo generalizado de fomentar politicas de empleo que incentiven el envejecimiento activo y la prolongacion de la vida profesional para frenar la salida prematura del mercado laboral de los trabajadores mayores de 45 anos. Entre las medidas a adoptar para el mantenimiento de los trabajadores en las empresas estan la adaptacion de los contenidos de los puestos de trabajo, el aprovechamiento del conocimiento interno y la formacion continua de los trabajadores. En los casos en los que ya se haya producido la expulsion del mercado laboral, seran las empresas de participacion las que pueden representar una via de salida ante la situacion de desempleo. Pero para que los desempleados de mayor edad decidan emprender su propia iniciativa empresarial necesitan formacion, asesoramiento y ayudas.