The efficient use of resources in cloud computing environments is one of many topics that have attracted many scientists in recent years. Today there is a wide variety of computing resources available in cloud computing infrastructures, such as multi-core CPUs, GPUs and FPGAs. In particular, accelerator-based architectures using FPGAs can provide significant business added value in Cloud environments. FPGA devices offer better performance due to their high processing capacity with predictable latency and low power consumption. However, managing these resources in a Cloud environment is not trivial, and is the focus of this work. In this paper, we present an architecture intended to integrate FPGAs as accelerators in a private cloud. The proposal takes into account quality of service (QoS) requirements. Thus, QoS indicators are related to user requirements. The proposed architecture seeks the right combination between CPUs and FPGAs in a Cloud environment.
|Translated title of the contribution||Design Of A Heterogeneous Architecture For Efficient Management Of Fpgas Resources In A Private Cloud|
|Original language||Spanish (Ecuador)|
|Title of host publication||Diseño De Una Arquitectura Heterogénea Para La Gestión Eficiente De Recursos Fpgas En Un Cloud Privado|
|State||Published - 26 Jun 2019|