Name
Abstract | ||
|---|---|---|
In this paper, we propose ACCLOUD (ACcelerated CLOUD) as a a novel architecture for cloud data centers. ACCLOUD features FPGA cards which work with the cloud servers or stand alone for hardware acceleration in the data plane. To this end, FPGA reconfigurable regions are virtualized and offered to the user together with other cloud resources including CPU, memory or disk. The cloud control is carried out within OpenStack framework incorporating the hardware resources. We propose a novel resource management approach for ACCOUD which fulfills the user requests by considering the tradeoffs between assignment of soft computing resources and hardware accelerators. |
Year | Venue | Keywords |
|---|---|---|
2018 | Signal Processing and Communications Applications Conference | Cloud computing,Data center,Hardware accelerator,FPGA,Resource management |
Field | DocType | ISSN |
Resource management,Forwarding plane,Computer vision,Computer science,Server,Field-programmable gate array,Hardware acceleration,Acceleration,Artificial intelligence,Soft computing,Operating system,Cloud computing | Conference | 2165-0608 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Yazar, Alper | 1 | 0 | 0.68 |
| Ahmet Erol | 2 | 0 | 0.34 |
| Ece Guran Schmidt | 3 | 146 | 16.27 |