Seyed Majid Zahedi

REF: resource elasticity fairness with sharing incentives for multiprocessors

With the democratization of cloud and datacenter computing, users increasingly share large hardware platforms. In this setting, architects encounter two challenges: sharing fairly and sharing multiple resources. Drawing on economic game-theory, we rethink fairness in computer architecture. A fair allocation must provide sharing incentives (SI), envy-freeness (EF), and Pareto efficiency (PE). We show that Cobb-Douglas utility functions are well suited to modeling user preferences for cache capacity and memory bandwidth. And we present an allocation mechanism that uses Cobb-Douglas preferences to determine each users fair share of the hardware. This mechanism provably guarantees SI, EF, and PE, as well as strategy-proofness in the large (SPL). And it does so with modest performance penalties, less than 10\% throughput loss, relative to an unfair mechanism.

Computational Sprinting: Architecture, Dynamics, and Strategies

Computational sprinting is a class of mechanisms that boost performance but dissipate additional power. We describe a sprinting architecture in which many, independent chip multiprocessors share a power supply and sprints are constrained by the chips’ thermal limits and the rack’s power limits. Moreover, we present the computational sprinting game, a multi-agent perspective on managing sprints. Strategic agents decide whether to sprint based on application phases and system conditions. The game produces an equilibrium that improves task throughput for data analytics workloads by 4--6× over prior greedy heuristics and performs within 90% of an upper bound on throughput from a globally optimized policy.

Reliable energy-aware application mapping and voltage-frequency island partitioning for GALS-based NoC

Reliable energy-aware application mapping, task scheduling, and voltage-frequency island partitioning so as to minimize the energy consumption while preserving the required bandwidth and latency is considered as a challenging problem in the designing of Multi-Processor System-on-Chip. To achieve modular design and low power consumption, Globally Asynchronous Locally Synchronous (GALS) design paradigm is a promising approach which fits very well with the voltage-frequency islands concept. In this paper, we formulate mapping problem of a real-time application with stochastic execution times onto multicore systems, scheduling tasks on processors, and assigning voltage-frequency levels to Processing Elements (PEs) as a Mixed Integer Linear Programming (MILP) in GALS-based Network-on-Chip. Furthermore, owing to the importance of reliability issue, we address the effects of transient faults in our proposed MILP formulation such that the reliability of the whole system incorporating several heterogeneous PEs is guaranteed to be better than a given threshold. Due to the NP-hardness of such a problem, a rounding by sampling-based heuristic algorithm is provided. Experimental results based on E3S benchmark suite and some real applications show the effectiveness of our proposed heuristic in achieving a near-optimal solution in a small fractional of time needed to find the optimal solution. Experimental results also show that, our formulation preserves the required reliability and increases the energy consumption by 70% in some cases.

Managing Heterogeneous Datacenters with Tokens

Ensuring fairness in a system with scarce, preferred resources requires time sharing. We consider a heterogeneous system with a few “big” and many “small” processors. We allocate heterogeneous processors using a novel token mechanism, which frames the allocation problem as a repeated game. At each round, users request big processors and spend a token if their request is granted. We analyze the game and optimize users’ strategies to produce an equilibrium. In equilibrium, allocations balance performance and fairness. Our mechanism outperforms classical, fair mechanisms by 1.7×, on average, in performance gains, and is competitive with a performance maximizing mechanism.

Fair and Efficient Social Choice in Dynamic Settings

We study a dynamic social choice problem in which an alternative is chosen at each round according to the reported valuations of a set of agents. In the interests of obtaining a solution that is both efficient and fair, we aim to maximize the long-term Nash welfare, which is the product of all agents’ utilities. We present and analyze two greedy algorithms for this problem, including the classic Proportional Fair (PF) algorithm. We analyze several versions of the algorithms and how they relate, and provide an axiomatization of PF. Finally, we evaluate the algorithms on data gathered from a computer systems application.