Generic construction of monitors for floating point unit designs

This paper proposes a set of well defined steps to design functional verification monitors intended to verify Floating Point Units (FPU) described in HDL. The first step consists on defining the input and output domain coverage. Next, the corner cases are defined. Finally, an already verified reference model is used in order to test the correctness of the Device Under Verification (DUV). As a case study a monitor for an IEEE754-2008 compliant design is implemented. This monitor is built to be easily instantiated into verification frameworks such as OVM. Two different designs were verified reaching complete input coverage and successful compliant results.

Design and implementation of a floating point unit for rigel, a massively parallel accelerator

Scientific applications rely heavily on floating point data types. Floating point operations are complex and require complicated hardware that is both area and power intensive. The emergence of massively parallel architectures like Rigel creates new challenges and poses new questions with respect to floating point support. The massively parallel aspect of Rigel places great emphasis on area efficient, low power designs. At the same time, Rigel is a general purpose accelerator and must provide high performance for a wide class of applications. This thesis presents an analysis of various floating point unit (FPU) components with respect to Rigel, and attempts to present a candidate design of an FPU that balances performance, area, and power and is suitable for massively parallel architectures like Rigel.

A floating point CORDIC based SVD processor

An SVD processor system is presented in which each processing element is implemented using a simple CORDIC unit. The internal recursive loop within the CORDIC module is exploited, with pipelining being used to multiplex the two independent micro-rotations onto a single CORDIC processor. This leads to a high performance and efficient hardware architecture. In addition, a novel method for scale factor correction is presented which only need be applied once at the end of the computation. This also reduces the computation time. The net result is an SVD architecture based on a conventional CORDIC approach, which combines high performance with high silicon area efficiency.

A cost effective pipelined divider for double precision floating point number

he growth of high-performance application in computer graphics, signal processing and scientific computing is a key driver for high performance, fixed latency; pipelined floating point dividers. Solutions available in the literature use large lookup table for double precision floating point operations.In this paper, we propose a cost effective, fixed latency pipelined divider using modified Taylor-series expansion for double precision floating point operations. We reduce chip area by using a smaller lookup table. We show that the latency of the proposed divider is 49.4 times the latency of a full-adder. The proposed divider reduces chip area by about 81% than the pipelined divider in [9] which is based on modified Taylor-series.

Inexact floating-point adder for dynamic image processing

Power has become a key constraint in current nanoscale integrated circuit design due to the increasing demands for mobile computing and a low carbon economy. As an emerging technology, an inexact circuit design offers a promising approach to significantly reduce both dynamic and static power dissipation for error tolerant applications. Although fixed-point arithmetic circuits have been studied in terms of inexact computing, floating-point arithmetic circuits have not been fully considered although require more power. In this paper, the first inexact floating-point adder is designed and applied to high dynamic range (HDR) image processing. Inexact floating-point adders are proposed by approximately designing an exponent subtractor and mantissa adder. Related logic operations including normalization and rounding modules are also considered in terms of inexact computing. Two HDR images are processed using the proposed inexact floating-point adders to show the validity of the inexact design. HDR-VDP is used as a metric to measure the subjective results of the image addition. Significant improvements have been achieved in terms of area, delay and power consumption. Comparison results show that the proposed inexact floating-point adders can improve power consumption and the power-delay product by 29.98% and 39.60%, respectively.

Design and Analysis of Inexact Floating-Point Adders

Power has become a key constraint in nanoscale inte-grated circuit design due to the increasing demands for mobile computing and higher integration density. As an emerging compu-tational paradigm, an inexact circuit offers a promising approach to significantly reduce both dynamic and static power dissipation for error-tolerant applications. In this paper, an inexact floating-point adder is proposed by approximately designing an exponent sub-tractor and mantissa adder. Related operations such as normaliza-tion and rounding are also dealt with in terms of inexact computing. An upper bound error analysis for the average case is presented to guide the inexact design; it shows that the inexact floating-point adder design is dependent on the application data range. High dynamic range images are then processed using the proposed inexact floating-point adders to show the validity of the inexact design; comparison results show that the proposed inexact floating-point adders can improve the power consumption and power-delay product by 29.98% and 39.60%, respectively.

Trans-floating-point arithmetic removes nine quadrillion redundancies from 64-bit IEEE 754 floating-point arithmetic

IEEE 754 floating-point arithmetic is widely used in modern, general-purpose computers. It is based on real arithmetic and is made total by adding both a positive and a negative infinity, a negative zero, and many Not-a-Number (NaN) states. Transreal arithmetic is total. It also has a positive and a negative infinity but no negative zero, and it has a single, unordered number, nullity. Modifying the IEEE arithmetic so that it uses transreal arithmetic has a number of advantages. It removes one redundant binade from IEEE floating-point objects, doubling the numerical precision of the arithmetic. It removes eight redundant, relational,floating-point operations and removes the redundant total order operation. It replaces the non-reflexive, floating-point, equality operator with a reflexive equality operator and it indicates that some of the exceptions may be removed as redundant { subject to issues of backward compatibility and transient future compatibility as programmers migrate to the transreal paradigm.

Transreal limits expose category errors in IEEE 754 floating-point arithmetic and in mathematics

The IEEE 754 standard for oating-point arithmetic is widely used in computing. It is based on real arithmetic and is made total by adding both a positive and a negative infinity, a negative zero, and many Not-a-Number (NaN) states. The IEEE infinities are said to have the behaviour of limits. Transreal arithmetic is total. It also has a positive and a negative infinity but no negative zero, and it has a single, unordered number, nullity. We elucidate the transreal tangent and extend real limits to transreal limits. Arguing from this firm foundation, we maintain that there are three category errors in the IEEE 754 standard. Firstly the claim that IEEE infinities are limits of real arithmetic confuses limiting processes with arithmetic. Secondly a defence of IEEE negative zero confuses the limit of a function with the value of a function. Thirdly the definition of IEEE NaNs confuses undefined with unordered. Furthermore we prove that the tangent function, with the infinities given by geometrical con- struction, has a period of an entire rotation, not half a rotation as is commonly understood. This illustrates a category error, confusing the limit with the value of a function, in an important area of applied mathe- matics { trigonometry. We brie y consider the wider implications of this category error. Another paper proposes transreal arithmetic as a basis for floating- point arithmetic; here we take the profound step of proposing transreal arithmetic as a replacement for real arithmetic to remove the possibility of certain category errors in mathematics. Thus we propose both theo- retical and practical advantages of transmathematics. In particular we argue that implementing transreal analysis in trans- floating-point arith- metic would extend the coverage, accuracy and reliability of almost all computer programs that exploit real analysis { essentially all programs in science and engineering and many in finance, medicine and other socially beneficial applications.

A Floating-point Extended Kalman Filter Implementation for Autonomous Mobile Robots

Localization and Mapping are two of the most important capabilities for autonomous mobile robots and have been receiving considerable attention from the scientific computing community over the last 10 years. One of the most efficient methods to address these problems is based on the use of the Extended Kalman Filter (EKF). The EKF simultaneously estimates a model of the environment (map) and the position of the robot based on odometric and exteroceptive sensor information. As this algorithm demands a considerable amount of computation, it is usually executed on high end PCs coupled to the robot. In this work we present an FPGA-based architecture for the EKF algorithm that is capable of processing two-dimensional maps containing up to 1.8 k features at real time (14 Hz), a three-fold improvement over a Pentium M 1.6 GHz, and a 13-fold improvement over an ARM920T 200 MHz. The proposed architecture also consumes only 1.3% of the Pentium and 12.3% of the ARM energy per feature.

Customizing floating-point units for FPGAs: Area-performance-standard trade-offs

The high integration density of current nanometer technologies allows the implementation of complex floating-point applications in a single FPGA. In this work the intrinsic complexity of floating-point operators is addressed targeting configurable devices and making design decisions providing the most suitable performance-standard compliance trade-offs. A set of floating-point libraries composed of adder/subtracter, multiplier, divisor, square root, exponential, logarithm and power function are presented. Each library has been designed taking into account special characteristics of current FPGAs, and with this purpose we have adapted the IEEE floating-point standard (software-oriented) to a custom FPGA-oriented format. Extended experimental results validate the design decisions made and prove the usefulness of reducing the format complexity

Floating-point exponentiation units for reconfigurable computing

The high performance and capacity of current FPGAs makes them suitable as acceleration co-processors. This article studies the implementation, for such accelerators, of the floating-point power function xy as defined by the C99 and IEEE 754-2008 standards, generalized here to arbitrary exponent and mantissa sizes. Last-bit accuracy at the smallest possible cost is obtained thanks to a careful study of the various subcomponents: a floating-point logarithm, a modified floating-point exponential, and a truncated floating-point multiplier. A parameterized architecture generator in the open-source FloPoCo project is presented in details and evaluated.

The statistical theory of relative errors in floating-point computation /

Thesis (M. S.)--University of Illinois at Urbana-Champaign.

Verification of floating point programs

In this thesis we present an approach to automated verification of floating point programs. Existing techniques for automated generation of correctness theorems are extended to produce proof obligations for accuracy guarantees and absence of floating point exceptions. A prototype automated real number theorem prover is presented, demonstrating a novel application of function interval arithmetic in the context of subdivision-based numerical theorem proving. The prototype is tested on correctness theorems for two simple yet nontrivial programs, proving exception freedom and tight accuracy guarantees automatically. The prover demonstrates a novel application of function interval arithmetic in the context of subdivision-based numerical theorem proving. The experiments show how function intervals can be used to combat the information loss problems that limit the applicability of traditional interval arithmetic in the context of hard real number theorem proving.