Fbufs: a high-bandwidth cross-domain transfer facility
SOSP '93 Proceedings of the fourteenth ACM symposium on Operating systems principles
Experiences with a high-speed network adaptor: a software perspective
SIGCOMM '94 Proceedings of the conference on Communications architectures, protocols and applications
U-Net: a user-level network interface for parallel and distributed computing
SOSP '95 Proceedings of the fifteenth ACM symposium on Operating systems principles
The SpectrumWare approach to wireless signal processing
Wireless Networks
Effects of buffering semantics on I/O performance
OSDI '96 Proceedings of the second USENIX symposium on Operating systems design and implementation
IO-LITE: a copy-free UNIX I/O system
IO-LITE: a copy-free UNIX I/O system
Virtual Sample Processing: Extending the Reach of Multimedia
Multimedia Tools and Applications
The APIC Approach to High Performance Network Interface Design: Protected DMA and Other Techniques
INFOCOM '97 Proceedings of the INFOCOM '97. Sixteenth Annual Joint Conference of the IEEE Computer and Communications Societies. Driving the Information Revolution
Container shipping: a uniform interface for fast, efficient, high-bandwidth i/o
Container shipping: a uniform interface for fast, efficient, high-bandwidth i/o
An experimental study of network performance impact of increased latency in software defined radios
Proceedings of the second ACM international workshop on Wireless network testbeds, experimental evaluation and characterization
The impact of software radio on wireless networking
ACM SIGMOBILE Mobile Computing and Communications Review
Hi-index | 0.00 |
Commodity PCs are on the verge of being capable of performing a variety of signal processing tasks that previously required special purpose hardware. Advances in the speed and width of their processors and internal buses allow these machines to manipulate data at rates that would allow their users to interact with a diverse range of sampled media, such as the raw RF spectrum and ultrasound. However, today's PCs lack an I/O system capable of delivering the appropriate bandwidth to these signal processing applications. These applications demand high continuous throughput I/O that smooths the inter-sample jitter introduced by interrupts, I/O bus latency, scheduling latency, etc. This paper presents a system that provides this functionality. Our system is composed of two tightly integrated parts: a PCI device that provides high raw I/O bus throughput and operating system enhancements to manage the device and provide low overhead transfers across the boundary between kernel and user space. The performance is excellent, providing up to 512 Mbits/sec of continuous throughput for an application. A description of both parts of the system is given, along with performance measurements, and a brief description of an application.