Source author record

Dirk Koch

Dirk Koch appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Catalog footprint

What is connected

4works
4topics
4close collaborators

Actions

Connect this record

Log in to claim

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

4 published item(s)

preprint2020arXiv

FOS: A Modular FPGA Operating System for Dynamic Workloads

With FPGAs now being deployed in the cloud and at the edge, there is a need for scalable design methods which can incorporate the heterogeneity present in the hardware and software components of FPGA systems. Moreover, these FPGA systems need to be maintainable and adaptable to changing workloads while improving accessibility for the application developers. However, current FPGA systems fail to achieve modularity and support for multi-tenancy due to dependencies between system components and lack of standardised abstraction layers. To solve this, we introduce a modular FPGA operating system -- FOS, which adopts a modular FPGA development flow to allow each system component to be changed and be agnostic to the heterogeneity of EDA tool versions, hardware and software layers. Further, to dynamically maximise the utilisation transparently from the users, FOS employs resource-elastic scheduling to arbitrate the FPGA resources in both time and spatial domain for any type of accelerators. Our evaluation on different FPGA boards shows that FOS can provide performance improvements in both single-tenant and multi-tenant environments while substantially reducing the development time and, at the same time, improving flexibility.

preprint2011arXiv

No-Break Dynamic Defragmentation of Reconfigurable Devices

We propose a new method for defragmenting the module layout of a reconfigurable device, enabled by a novel approach for dealing with communication needs between relocated modules and with inhomogeneities found in commonly used FPGAs. Our method is based on dynamic relocation of module positions during runtime, with only very little reconfiguration overhead; the objective is to maximize the length of contiguous free space that is available for new modules. We describe a number of algorithmic aspects of good defragmentation, and present an optimization method based on tabu search. Experimental results indicate that we can improve the quality of module layout by roughly 50 % over static layout. Among other benefits, this improvement avoids unnecessary rejections of modules