John Morris

Seeking solutions in configurable computing

Configurable computing offers the potential of producing powerful new computing systems. Will current research overcome the dearth of commercial applicability to make such systems a reality? Unfortunately, no system to date has yet proven attractive or competitive enough to establish a commercial presence. We believe that ample opportunity exists for work in a broad range of areas. In particular, the configurable computing community should focus on refining the emerging architectures, producing more effective software/hardware APIs, better tools for application development that incorporate the models of hardware reconfiguration, and effective benchmarking strategies.

A software component verification tool

Component based software engineering depends on reliable, robust components, since it may omit a unit test phase wholly or partially from the development cycle. The paper describes a tool that allows a component developer to design and run verification tests. In developing components for our library, we found it necessary to provide multiple mechanisms for identifying and capturing tests to overcome the limitations of any single mechanism. Once specified, test specifications and test results are stored in XML documents, providing a standard, portable form of storing, retrieving and updating test histories. One module of our component test bench, the test pattern verifier, has been designed to be general, lightweight and portable, so that it can be packaged with a component and its test specifications. This allows a component user to verify a components compliance with specifications in a target environment.

A generalised spreadsheet verification methodology

Although spreadsheets have been around for over thirty years, we are only just realising their importance. Most companies use spreadsheets in their decision-making processes, but rarely employ any form of testing. This paper shows how an all-uses test adequacy technique can be integrated into Microsofts Excel. The modular technique adopted makes the implementation spreadsheet package independent. It also includes a user interface, to assist developers specify test cases and a technique for recording test cases and session information. In particular it presents a systematic technique for constructing test cases. As a key problem with spreadsheet development is the inexperience of developers, this paper describes an easy to use tool that will improve the standard of spreadsheets developed.

Setting a Framework for Trusted Component Trading

Software engineers constructing systems from third-party components need to be able to assess the quality of those components in order to have any confidence in the final system. We argue that the most effective and economic way to do this is for developers to supply test specifications along with the components that they offer to the market. We describe a simple, portable test specification based on XML documents. When a test specification is packaged with a purchased component, system builders are efficiently able to not only verify a component developer’s claim for quality but also use the test specification to elucidate fine details of a component’s capabilities that may be missing or not manifest in a natural language specification. A system builder selecting a component for integration into a new system needs to consider more than its ability to perform the required functions: the degree of testing can be an important indication of its reliability. Thus we propose an index, describing the degree of testing that a component has undergone, which can be used to rate components and aid selection decisions. Combined with a test specification, a purchaser can efficiently determine whether a component meets claims for its quality.

Determining component reliability using a testing index

Component-Based Software Engineering has the potential to provide reliable systems based on tested components quickly and economically, but these systems will only be as reliable as the components from which they are constructed. We propose a 6-point scale which can be used to rate the degree to which a component has been tested. This scale can be used by developers to assess the risk of using a third party component. Since a variety of test strategies are used, it is necessary to correlate testing strategies with our scale. In this paper, we examine the testing strategies specified in British Standard 7925-2 and show how they relate to the reliability levels that we propose. Since well-behaved use of resources is also a key factor in overall system reliability, we propose that an R tag be added to the rated level when resource usage has been verified to be within reasonable bounds.

Performance of the Achilles Router

The Achilles Router provides low latency, high bandwidth connections between processors, enabling a network of low-cost processors to perform as a high-performance parallel processor. It is also economical, being constructed from low cost Field Programmable Gate Arrays (FPGAs). These programmable devices allow it to be re-programmed for use in a variety of network topologies; they also permit ‘tuning’ the router for optimum performance in different applications. A key factor in its simplicity and performance is the 3-D structure: this allows us to build a full cross-bar switch with a wide, high-bandwidth datapath. The simple cross-bar circuit also has very low latency: we measured latencies of ~ 800ns in the hardware and 2.5μs when software overheads were included. We measured the basic performance of an inter-processor link using Achilles, and then, using a range of benchmarks with different characteristics, showed that Achilles clearly outperforms Fast Ethernet.

Maintaining a suite of binocular facility instruments at the Large Binocular Telescope

Facility Instruments at the Large Binocular Telescope (LBT) include the Large Binocular Camera (LBC), a pair of wide-field imagers at the prime focus, the LUCIFER (or LUCI) near-infrared imager and spectrograph pair, and the Multi-Object Double Spectrograph (MODS), a pair of long-slit spectrographs. The disciplines involved in instrument support are reviewed, as well as scheduling of support personnel. A computerized system for instrument maintenance scheduling and spare parts inventory is described. Instrument problems are tracked via an online reporting system, and statistics on types of instrument problems are discussed, as well as applicability of the system to troubleshooting.

High-bandwidth and low latency from a three-dimensional reconfigurable interconnect

A high bandwidth, low latency processor interconnect has been constructed using off-the-shelf FPGAs and a small amount of additional logic. The Achilles system comprises a router in which the main circuit elements (the FPGAs) have been placed on ten small PCBs and connected together to make a router `stack and a PCI interface. On the first prototype, channel bandwidth has been measured at 28 Mbytes/second and latency was measured to be 800 ns. The newer low-voltage FPGAs will be used in the next Achilles variant, allowing the bandwidth capability will increase to close to the theoretical maximum for PCI bus. The router stack and PCI interface also has application as a general purpose reconfigurable processor which may be arbitrarily extended by adding further stacks to the system.

A simulator for high speed digital communications

Since parallel processors are generally constrained by the available interprocessor data transfer capability, system designers generally try to push interconnection systems to their limits in bandwidth. Practical and economic systems are constrained by many physical and packaging considerations such as a need to use commercially available connectors. We describe here VisiSolve-a simulator that we have built to predict the behaviour of interconnect systems that can readily be assembled from off-the-shelf components. It uses a finite element approach and predicts the dynamic electric field in the cells of the mesh. The irregular geometries of the individual parts of such components require us to adapt the mesh used in simulations in regions where the needs of a practical connector-small size, low insertion force and automatic assembly-have dictated the shape and path of the conductors. We have adopted a method which uses the constitutive error-the discrepancy between electric fields calculated directly and from /spl nabla//spl times/H when H was calculated directly/spl times/as an indicator that refinement is needed.