The Experts below are selected from a list of 19662 Experts worldwide ranked by ideXlab platform
Jonathan Sandberg - One of the best experts on this subject based on the ideXlab platform.
-
virtual memory mapped Network Interface for the shrimp multicomputer
International Symposium on Computer Architecture, 1994Co-Authors: Matthias A Blumrich, Richard D Alpert, Cezary Dubnicki, Edward W. Felten, Kai Li, Jonathan SandbergAbstract:The Network Interfaces of existing multicomputers require a significant amount of software overhead to provide protection and to implement message passing protocols. This paper describes the design of a low-latency, high-bandwidth, virtual memory-mapped Network Interface for the SHRIMP multicomputer project at Princeton University. Without sacrificing protection, the Network Interface achieves low latency by using virtual memory mapping and write-latency hiding techniques, and obtains high bandwidth by providing a user-level block data transfer mechanism. We have implemented several message passing primitives in an experimental environment, demonstrating that our approach can reduce the message passing overhead to a few user-level instructions.
Edward W. Felten - One of the best experts on this subject based on the ideXlab platform.
-
fast rpc on the shrimp virtual memory mapped Network Interface
Journal of Parallel and Distributed Computing, 1997Co-Authors: Angelos Bilas, Edward W. FeltenAbstract:The emergence of new Network Interface technology is enabling new approaches to the development of communications software. This paper evaluates the SHRIMP virtual memory mapped Network Interface by using it to build two fast implementations of remote procedure call (RPC). Our first implementations, called vRPC, is fully compatible with the SunRPC standard. We change the RPC runtime library, the operating system kernel is unchanged, and only a minimal change was needed in the stub generator to create a new protocol identifier. Despite these restrictions, our vRPC implementation is several times faster than existing SunRPC implementations. A round-trip null RPC with no arguments and results under vRPC takes about 33 ?s. Our second implementation, called ShrimpRPC, is not compatible with SunRPC but offers much better performance. ShrimpRPC specializes the stub generator and runtime library to take full advantage of SHRIMP's features. The result is a round-trip null RPC latency of 9.5 ?s, which is about 1 ?s above the hardware minimum.
-
virtual memory mapped Network Interface for the shrimp multicomputer
International Symposium on Computer Architecture, 1994Co-Authors: Matthias A Blumrich, Richard D Alpert, Cezary Dubnicki, Edward W. Felten, Kai Li, Jonathan SandbergAbstract:The Network Interfaces of existing multicomputers require a significant amount of software overhead to provide protection and to implement message passing protocols. This paper describes the design of a low-latency, high-bandwidth, virtual memory-mapped Network Interface for the SHRIMP multicomputer project at Princeton University. Without sacrificing protection, the Network Interface achieves low latency by using virtual memory mapping and write-latency hiding techniques, and obtains high bandwidth by providing a user-level block data transfer mechanism. We have implemented several message passing primitives in an experimental environment, demonstrating that our approach can reduce the message passing overhead to a few user-level instructions.
-
two virtual memory mapped Network Interface designs
High Performance Interconnects, 1994Co-Authors: Matthias A Blumrich, Edward W. Felten, Kai Li, C Dubnickil, Malena MesarinaAbstract:In existing multicomputers, software overhead dom inates the message-passing latency cost. Our research on the SHRIMP project at Princeton indicates that appropriate Network Interface support can significantly reduce this software overhead. We have designed two Network Interfaces for the SHRIMP multicomputer. Both support virtual memory mapped communication allowing user processes to communicate without doing expensive buffer management, and without using sys tem calls to cross the protection boundary separating user processes from the operating system kernel. This paper describes and compares the two Network in terfaces, and discusses performance tradeoffs between them.
Matthias A Blumrich - One of the best experts on this subject based on the ideXlab platform.
-
virtual memory mapped Network Interface for the shrimp multicomputer
International Symposium on Computer Architecture, 1994Co-Authors: Matthias A Blumrich, Richard D Alpert, Cezary Dubnicki, Edward W. Felten, Kai Li, Jonathan SandbergAbstract:The Network Interfaces of existing multicomputers require a significant amount of software overhead to provide protection and to implement message passing protocols. This paper describes the design of a low-latency, high-bandwidth, virtual memory-mapped Network Interface for the SHRIMP multicomputer project at Princeton University. Without sacrificing protection, the Network Interface achieves low latency by using virtual memory mapping and write-latency hiding techniques, and obtains high bandwidth by providing a user-level block data transfer mechanism. We have implemented several message passing primitives in an experimental environment, demonstrating that our approach can reduce the message passing overhead to a few user-level instructions.
-
two virtual memory mapped Network Interface designs
High Performance Interconnects, 1994Co-Authors: Matthias A Blumrich, Edward W. Felten, Kai Li, C Dubnickil, Malena MesarinaAbstract:In existing multicomputers, software overhead dom inates the message-passing latency cost. Our research on the SHRIMP project at Princeton indicates that appropriate Network Interface support can significantly reduce this software overhead. We have designed two Network Interfaces for the SHRIMP multicomputer. Both support virtual memory mapped communication allowing user processes to communicate without doing expensive buffer management, and without using sys tem calls to cross the protection boundary separating user processes from the operating system kernel. This paper describes and compares the two Network in terfaces, and discusses performance tradeoffs between them.
O. Nakamura - One of the best experts on this subject based on the ideXlab platform.
-
Operating system support for Network control: a virtual Network Interface approach for end-host OSs
IEEE 2002 Tenth IEEE International Workshop on Quality of Service (Cat. No.02EX564), 2002Co-Authors: T. Okumura, M. Minami, Daniel Mosse, O. NakamuraAbstract:Because of user demands for better quality of service, Network-aware applications have been of increasing necessity. To enable more control, the end-host operating system (OS) is the entity responsible for providing appropriate service level and API to user applications. However, most of the work in this area remains domain-specific and without a generalizable scheme for providing Network control as an OS service. We propose an OS service, namely the virtualization of Network Interface, that lies between Network Interface and userland. The virtual Network Interface is hierarchically attachable to various OS-supported entity, such as threads, processes, and sockets. We argue that the mechanism provides flexible control, as well as the system protection that is required for operating system services. For a proof of the concept, we show an implementation on a PC-Unix, using the procfs file system abstraction. We also carried out a systematic evaluation. The system exhibited the expected control behavior, while keeping the performance small.
Kai Li - One of the best experts on this subject based on the ideXlab platform.
-
PCRCW - Network Interface Support for User-Level Buffer Management
Parallel Computer Routing and Communication, 1994Co-Authors: Cezary Dubnicki, Kai Li, Malena MesarinaAbstract:The Network Interfaces of existing multicomputers and workstations require a significant amount of software overhead to provide protection and buffer management in order to implement message-passing protocols. This paper advocates a physical memory mapping method in a Network Interface design that supports user-level buffer management. The method requires only a minimal addition to the traditional DMA-based Network Interface design and eliminates the need for memory buffer management in the operating system kernel. As a result, the software overhead on message passing is reduced by up to 78% and both receive system call and receive interrupt can be avoided.
-
virtual memory mapped Network Interface for the shrimp multicomputer
International Symposium on Computer Architecture, 1994Co-Authors: Matthias A Blumrich, Richard D Alpert, Cezary Dubnicki, Edward W. Felten, Kai Li, Jonathan SandbergAbstract:The Network Interfaces of existing multicomputers require a significant amount of software overhead to provide protection and to implement message passing protocols. This paper describes the design of a low-latency, high-bandwidth, virtual memory-mapped Network Interface for the SHRIMP multicomputer project at Princeton University. Without sacrificing protection, the Network Interface achieves low latency by using virtual memory mapping and write-latency hiding techniques, and obtains high bandwidth by providing a user-level block data transfer mechanism. We have implemented several message passing primitives in an experimental environment, demonstrating that our approach can reduce the message passing overhead to a few user-level instructions.
-
two virtual memory mapped Network Interface designs
High Performance Interconnects, 1994Co-Authors: Matthias A Blumrich, Edward W. Felten, Kai Li, C Dubnickil, Malena MesarinaAbstract:In existing multicomputers, software overhead dom inates the message-passing latency cost. Our research on the SHRIMP project at Princeton indicates that appropriate Network Interface support can significantly reduce this software overhead. We have designed two Network Interfaces for the SHRIMP multicomputer. Both support virtual memory mapped communication allowing user processes to communicate without doing expensive buffer management, and without using sys tem calls to cross the protection boundary separating user processes from the operating system kernel. This paper describes and compares the two Network in terfaces, and discusses performance tradeoffs between them.
