A
Albert Sidelnik
Researcher at IBM
Publications - 36
Citations - 897
Albert Sidelnik is an academic researcher from IBM. The author has contributed to research in topics: Node (networking) & Massively parallel. The author has an hindex of 15, co-authored 36 publications receiving 864 citations. Previous affiliations of Albert Sidelnik include University of Illinois at Urbana–Champaign & Nvidia.
Papers
More filters
Patent
Providing policy-based operating system services in a hypervisor on a computing system
TL;DR: In this paper, the authors describe a policy-based operating system service in a hypervisor on a computing system, where the operating system includes a kernel proxy and a plurality of operating system services of a service type.
Proceedings ArticleDOI
Locality-Driven Dynamic GPU Cache Bypassing
TL;DR: This paper presents a design that integrates locality filtering based on reuse characteristics of GPU workloads into the decoupled tag store of the existing L1 D-cache through simple and cost-effective hardware extensions.
Patent
Providing policy-based operating system services in an operating system on a computing system
TL;DR: In this paper, the authors describe a policy-based operating system service in an operating system on a computing system that includes a kernel and a plurality of operating system services of a service type.
Proceedings ArticleDOI
Dynamic thread block launch: a lightweight execution mechanism to support irregular applications on GPUs
TL;DR: This paper proposes a new mechanism called Dynamic Thread Block Launch (DTBL) to extend the current bulk synchronous parallel model underlying the current GPU execution model by supporting dynamic spawning of lightweight thread blocks rather than kernels to execute dynamically occurring parallel work elements.
Patent
Mechanism for process migration on a massively parallel computer
TL;DR: In this article, the authors provide a mechanism for process migration on a massively parallel computer system. And they use MPI state data for a migrated compute node, such as MPI (or other communication library) state data across a full collection of compute nodes present in a given parallel system executing a parallel task.