%0 Journal Article
%T Thermal-Aware Test Schedule and TAM Co-Optimization for Three-Dimensional IC
%A Chi-Jih Shih
%A Chih-Yao Hsu
%A Chun-Yi Kuo
%A James Li
%A Jiann-Chyi Rau
%A Krishnendu Chakrabarty
%J Active and Passive Electronic Components
%D 2012
%I Hindawi Publishing Corporation
%R 10.1155/2012/763572
%X Testing is regarded as one of the most difficult challenges for three-dimensional integrated circuits (3D ICs). In this paper, we want to optimize the cost of TAM (test access mechanism) and the test time for 3D IC. We used both greedy and simulated annealing algorithms to solve this optimization problem. We compare the results of two assumptions: soft-die mode and hard-die mode. The former assumes that the DfT of dies cannot be changed, while the latter assumes that the DfT of dies can be adjusted. The results show that thermal-aware cooptimization is essential to decide the optimal TAM and test schedule. Blindly adding TAM cannot reduce the total test cost due to temperature constraints. Another conclusion is that soft-die mode is more effective than hard-die mode to reduce the total test cost for 3D IC. 1. Introduction Three-dimensional integrated circuits (3D ICs) provide a promising solution to process scaling and heterogeneous system integration [1¨C3]. In spite of many advantages, 3D ICs still have many challenges ahead. Among them, high temperature issue is probably the most critical one, because vertical heat dissipation paths in 3D ICs are longer than those in 2D IC [4¨C7]. Thus, high temperatures cause serious yield loss problem when testing 3D ICs. Many papers have proposed algorithm of test schedule optimization for 2D IC [8, 9], including thermal-aware test scheduling [10¨C13]. In [10], two optimization algorithms are proposed which try to spread heat more evenly in a chip via layout information and a progressive weight function. A rectangular 2D bin packing can solve the test scheduling problem by considering dynamic thermal profiles [11]. A thermal-safe test scheduling method used resource conflict graph for optimization [12]. After a test schedule is obtained, a 2D thermal resistance model is applied to check whether the thermal constraint is met. This technique, however, does not consider the TAM constraint. The thermal-resistance model which used superposition principle has been introduced for 2D IC test scheduling optimization [13]. Many techniques used integer linear programming (ILP) to find an optimal solution. However, when thermal constraints are considered, there could be an exponential growth in the problem size because of the need for evaluating all possible combinations. A die-level test scheduling method for 3D IC was proposed in the previous work [14]. In their work, they addressed the issue of test scheduling to minimize overall test time for stack testing as well as postbond testing without temperature consideration.
%U http://www.hindawi.com/journals/apec/2012/763572/