Sabrewing: A lightweight architecture for combined floating-point and integer arithmetic

In spite of the fact that floating-point arithmetic is costly in terms of silicon area, the joint design of hardware for floating-point and integer arithmetic is seldom considered. While components like multipliers and adders can potentially be shared, floating-point and integer units in contemporary processors are practically disjoint. This work presents a new architecture which tightly integrates floating-point and integer arithmetic in a single datapath. It is mainly intended for use in low-power embedded digital signal processors and therefore the following design constraints were important: limited use of pipelining for the convenience of the compiler; maintaining compatibility with existing technology; minimal area and power consumption for applicability in embedded systems. The architecture is tailored to digital signal processing by combining floating-point fused multiply-add and integer multiply-accumulate. It could be deployed in a multi-core system-on-chip designed to support applications with and without dominance of floating-point calculations. The VHDL structural description of this architecture is available for download under BSD license. Besides being configurable at design time, it has been thoroughly checked for IEEE-754 compliance by means of a floating-point test suite originating from the IBM Research Labs. A proof-of-concept has also been implemented using STMicroelectronics 65nm technology. This prototype supports 32-bit signed two’s complement integers and 41-bit (8-bit exponent and 32-bit significand) floating-point numbers. Our evaluations show that over 67% energy and 19% area can be saved compared to a reference design in which floating-point and integer arithmetic are implemented separately. The area overhead caused by combining floating-point and integer is less than 5%. Implemented in ST’s general-purpose CMOS technology, the design can operate at a frequency of 1.35GHz, while 667MHz can be achieved in low-power CMOS. Considering that the entire datapath is partitioned in just three pipeline stages, and the fact that the design is intended for use in the low-power domain, these frequencies are adequate. They are in fact competitive with current technology low-power floating-point units. Post-layout estimates indicate that the required area of a low-power implementation can be as small as 0.04mm2 . Power consumption is on the order of several milliwatts. Strengthened by the fact that clock gating could reduce power consumption even further, we think that a shared floating-point and integer architecture is a good choice for signal processing in low-power embedded systems