With the increasing clock rate and transistor count of today's microprocessors, power dissipation is becoming a critical component of system design complexity. Thermal and power-delivery issues are becoming especially critical for high-performance computing systems. In this work, we investigate dynamic thermal management as a technique to control CPU power dissipation. With the increasing usage of clock gating techniques, the average power dissipation typically seen by common applications is becoming much less than the chip's rated maximum power dissipation. However system designers still must design thermal heat sinks to withstand the worse-case scenario. We define and investigate the major components of any dynamic thermal management scheme. Specifically we explore the tradeoffs between several mechanisms for responding to periods of thermal trauma and we consider the effects of hardware and software implementations. With approximate dynamic thermal management, the CPU can be designed for a much lower maximum power rating, with minimal performance impact for typical applications.

Dynamic thermal management for high-performance microprocessors

In this paper, we propose a set of rules for consistent estimation of the real performance and power features of the flip-flop and master-slave latch structures. A new simulation and optimization approach is presented, targeting both high-performance and power budget issues. The analysis approach reveals the sources of performance and power-consumption bottlenecks in different design styles. Certain misleading parameters have been properly modified and weighted to reflect the real properties of the compared structures. Furthermore, the results of the comparison of representative master-slave latches and flip-flops illustrate the advantages of our approach and the suitability of different design styles for high-performance and low-power applications.

/pdf/comparative-analysis-of-master-slave-latches-and-flip-flops-4ab70msy8h.pdf

Comparative analysis of master-slave latches and flip-flops for high-performance and low-power systems

In this paper we investigate possible ways to improve the energy efficiency of a general purpose microprocessor. We show that the energy of a processor depends on its performance, so we chose the energy-delay product to compare different processors. To improve the energy-delay product we explore methods of reducing energy consumption that do not lead to performance loss (i.e. wasted energy), and explore methods to reduce delay by exploiting instruction level parallelism. We found that careful design reduced the energy dissipation by almost 25%. Pipelining can give approximately a 2/spl times/ improvement in energy-delay product. Superscalar issue, however, does not improve the energy-delay product any further since the overhead required offsets the gains in performance. Further improvements will be hard to come by since a large fraction of the energy (50-80%) is dissipated in the clock network and the on-chip memories. Thus, the efficiency of processors will depend more on the technology being used and the algorithm chosen by the programmer than the micro-architecture.

Energy dissipation in general purpose microprocessors

Design and experimental evaluation of a new sense-amplifier-based flip-flop (SAFF) is presented. It was found that the main speed bottleneck of existing SAFF's is the cross-coupled set-reset (SR) latch in the output stage. The new flip-flop uses a new output stage latch topology that significantly reduces delay and improves driving capability. The performance of this flip-flop is verified by measurements on a test chip implemented in 0.18 /spl mu/m effective channel length CMOS. Demonstrated speed places it among the fastest flip-flops used in the state-of-the-art processors. Measurement techniques employed in this work as well as the measurement set-up are discussed in this paper.

/pdf/improved-sense-amplifier-based-flip-flop-design-and-1l8icizbke.pdf

Improved sense-amplifier-based flip-flop: design and measurements

In general-purpose microprocessors, recent trends have pushed towards 64 bit word widths, primarily to accommodate the large addressing needs of some programs. Many integer problems, however, rarely need the full 64 bit dynamic range these CPUs provide. In fact, another recent instruction set trend has been increased support for sub-word operations (that is, manipulating data in quantities less than the full word size). In particular, most major processor families have introduced "multimedia" instruction set extensions that operate in parallel on several sub-word quantities in the same ALU. This paper notes that across the SPECint95 benchmarks, over half of the integer operation executions require 16 bits or less. With this as motivation, our work proposes hardware mechanisms that dynamically recognize and capitalize on these "narrow-bitwidth" instances. Both optimizations require little additional hardware, and neither requires compiler support. The first, power-oriented, optimization reduces processor power consumption by using aggressive clock gating to turn off portions of integer arithmetic units that will be unnecessary for narrow bitwidth operations. This optimization results in an over 50% reduction in the integer unit's power consumption for the SPECint95 and MediaBench benchmark suites. The second optimization improves performance by merging together narrow integer operations and allowing them to share a single functional unit. Conceptually akin to a dynamic form of MMX, this optimization offers speedups of 4.3%-6.2% for SPECint95 and 8.0%-10.4% for MediaBench.

/pdf/dynamically-exploiting-narrow-width-operands-to-improve-32eumhcvb6.pdf

Dynamically exploiting narrow width operands to improve processor power and performance

A 28 mW/MHz at 80 MHz structured-custom RISC microprocessor design is described. This 32-b implementation of the PowerPC architecture is fabricated in a 3.3 V, 0.5 /spl mu/m, 4-level metal CMOS technology, resulting in 1.6 million transistors in a 7.4 mm by 11.5 mm chip size. Dual 8-kilobyte instruction and data caches coupled to a high performance 32/64-b system bus and separate execution units (float, integer, loadstore, and system units) result in peak instruction rates of three instructions per clock cycle. Low-power design techniques are used throughout the entire design, including dynamically powered down execution units. Typical power dissipation is kept under 2.2 W at 80 MHz. Three distinct levels of software-programmable, static, low-power operation-for system power management are offered, resulting in standby power dissipation from 2 mW to 350 mW. CPU to bus clock ratios of 1/spl times/, 2/spl times/, 3/spl times/, and 4/spl times/ are implemented to allow control of system power while maintaining processor performance. As a result, workstation level performance is packed into a low-power, low-cost design ideal for notebooks and desktop computers. >

A 2.2 W, 80 MHz superscalar RISC microprocessor

This superscalar microprocessor is a 32b implementation of the PowerPC Architecture. With an estimated performance/power ratio of 25SPECint92/W at 80 MHz, this RISC style chip offers workstation-level performance packed into a low-power consumption, low-cost design ideal for notebooks and desktop computers. >

http://ieeexplore.ieee.org/iel2/1111/8026/00344665.pdf

A 3.0 W 75SPECint92 85SPECfp92 superscalar RISC microprocessor

This superscalar microprocessor is a 32b implementation of the PowerPC Architecture(TM) specification based on a micro-architecture designed for high performance and low power Two instructions per cycle can be dispatched in this superscalar design The processor includes dual 32kB 8-way instruction and data caches, a floating-point unit, two integer units, a branch unit, a load/store unit, and a system unit An L2 tag and cache controller with a dedicated L2 bus interface are added to provide a low-cost L2 cache solution using commodity SRAMs for the data

A 250 MHz 5 W RISC microprocessor with on-chip L2 cache controller

This 32 b superscalar processor, having 18 mW/MHz projected power consumption at 66 MHz, is designed for desktop companions and high-end embedded multimedia applications with graphics-intensive requirements such as high-performance video games. This processor, the latest member of the PowerPC microprocessor family, can also be used in other low-power computing applications. The processor is fabricated in a 3.3 V, 0.5 m, 4-level metal CMOS resulting in 1 M transistors in a 7.07/spl times/7.07 mm/sup 2/ chip. Dual 4 kB instruction and data caches coupled to a high-performance 64 b multiplexed bus and separate execution units (float, integer, branch, and load-store) result in 2 instructions per clock cycle peak rate. Low-power design includes dynamically-powered-down execution units. Standby power is <2 mW. CPU to bus clock ratios of 2/spl times/ and 3/spl times/ allow control of system power while maintaining processor performance.

Tai Ngo

Papers

A 2.2 W, 80 MHz superscalar RISC microprocessor

A 3.0 W 75SPECint92 85SPECfp92 superscalar RISC microprocessor

A 250 MHz 5 W RISC microprocessor with on-chip L2 cache controller

A 1.2 W 66 MHz superscalar RISC microprocessor for set-tops, video games, and PDAs