Massoud Pedram

Bio: Massoud Pedram is an academic researcher from University of Southern California. The author has contributed to research in topics: Energy consumption & CMOS. The author has an hindex of 77, co-authored 780 publications receiving 23047 citations. Previous affiliations of Massoud Pedram include University of California, Berkeley & Syracuse University.

Balancing On-Chip Network Latency in Multi-application Mapping for Chip-Multiprocessors

Abstract: As the number of cores continues to grow in chip multiprocessors (CMPs), application-to-core mapping algorithms that leverage the non-uniform on-chip resource access time have been receiving increasing attention. However, existing mapping methods for reducing overall packet latency cannot meet the requirement of balanced on-chip latency when multiple applications are present. In this paper, we address the looming issue of balancing minimized on-chip packet latency with performance-awareness in the multi-application mapping of CMPs. Specifically, the proposed mapping problem is formulated, its NP-completeness is proven, and an efficient heuristic-based algorithm for solving the problem is presented. Simulation results show that the proposed algorithm is able to reduce the maximum average packet latency by 10.42% and the standard deviation of packet latency by 99.65% among concurrently running applications and, at the same time, incur little degradation in the overall performance.

Two high-performance and low-power serial communication interfaces for on-chip interconnects

Abstract: This paper presents two novel methods for on-chip serial communication in which the clocks of the transmitter and the receiver are generated with two separate ring oscillators. These oscillators are identical, although they can have a small frequency difference. In the first method, a strobe line that toggles exactly once with every frame of n-bit data is used to activate the oscillators. Local counters are used to count the number of bits in the data frame and to stop the local oscillators when the frame has been processed. In the second method, a single physical line is used to transmit both data and (in-band) control information, further reducing the power dissipation. The data transmission is controlled by the output of a starter flip-flop that indicates the empty/full state of an input buffer, whereas the data reception is controlled by the decoding of a "1" start bit and a "0" end bit, both of which are added to the n-bit data word to form a frame. Circuit simulation results demonstrate that both communication methods have high bandwidth and low power dissipation.

Joint Charge and Thermal Management for Batteries in Portable Systems With Hybrid Power Sources

Abstract: This paper introduces a joint charge and thermal management problem for batteries in a battery–supercapacitor hybrid power source of a portable system, which has been equipped with a forced convection cooling technique, such as a fan. A key consideration in such a system is that the battery aging depends strongly on the battery temperature, which is in turn a function of the workload running on the device and the control policy for the fan. More precisely, this paper presents a hierarchical algorithm for maximizing the battery lifespan under given workload conditions. The algorithm relies on a combination of reinforcement learning and dynamic programming techniques. Simulation results show that the proposed algorithm achieves up to $2 {\times } $ improvements in battery lifespan, resulting in completion of up to 80% additional workload before the battery expires.

Gate-all-around FET based 6T SRAM design using a device-circuit co-optimization framework

Abstract: Gate-all-around nanowire transistor is deemed as one of the most promising solutions that enables continued CMOS scaling. Compared with FinFET, it further suppresses short-channel effects by providing superior electrostatic control over the channel. Due to the unique device structure, gate-all-around nanowire transistor also allows more efficient layout design by exploiting 3-dimensional stacking configurations. In this paper, we investigate the 6T SRAM cell design for gate-all-around nanowire transistors using a device-circuit co-optimization framework. At the device level, TCAD simulation and current source modeling method are applied to extract the model. Layout designs with horizontal, lateral, vertical stacking device structures are explored. At the circuit level, read and write assist techniques are studied to relieve the negative impact of low on-currents on SRAM stabilities incurred by nanowire channels. Operating at 300 mV, assist techniques can increase the read static noise margin and the write static noise margin of 6T SRAM up to 82% and 92%, respectively.

FPGA synthesis for minimum area, delay and power

Abstract: In this paper, we address the problems of minimizing the area, delay and power during synthesis of field programmable gate arrays (FPGAs). We use Boolean decomposition techniques to minimize the number of configurable logic blocks (CLBs), the depth of the network and the power dissipations. We use OBDDs to represent functions so that our methods can be implemented more effectively. Our mapping algorithm is based on function decomposition which was pioneered by Ashenhurst [1959].

I and i

Abstract: There is, I think, something ethereal about i —the square root of minus one. I remember first hearing about it at school. It seemed an odd beast at that time—an intruder hovering on the edge of reality. Usually familiarity dulls this sense of the bizarre, but in the case of i it was the reverse: over the years the sense of its surreal nature intensified. It seemed that it was impossible to write mathematics that described the real world in …

Machine learning

Abstract: Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Pattern Recognition and Machine Learning

Abstract: Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.