Dual-task interference is a widely observed phenomenon where the decision-making becomes less efficient (slower or less accurate) [1]. This inefficiency is proposed to be a potential interference between the two decision-making processes, but the nature and extent of this interference have been debated [2]. For example, when two decisions are required simultaneously, dual-task interference might arise due to lower signals or higher noise available for each decision. Furthermore, the suppression of the signal may be partial or complete. However, the nature and extent of dual-task interference could not be quantified due to the lack of a quantitative model that can estimate the extent of the suppression of the signal and/or the amplification of the noise. We developed a novel model of dual-task interference by making an efficient two-dimensional drift-diffusion model that allows adjustable levels of signal and noise during simultaneous decision-making [3]. Our model has two stages: a simultaneous stage, where two streams of evidence are accumulated simultaneously for decision, and a non-simultaneous stage, where one decision has terminated and another is still underway (Figure 1A). Our method enables a separate and quantitative examination of the signal and noise levels in the simultaneous and non-simultaneous stages (Figure 1B). Because a brute-force calculation of the two-dimensional drift-diffusion process would suffer from prohibitive computational cost, we developed an efficient method that combines four one-dimensional drift-diffusion processes, one for each of the simultaneous and non-simultaneous stages of each of two decisions, that reduces the computation time ~100 times. Through simulation, we show that (1) as long as the signal-to-noise ratio remains unchanged, accuracy does not change with evidence strength, while the reaction time is higher when the signal is lower (Figure 1C); (2) when the signal is reduced with the noise unchanged, the accuracy decreases but the reaction time increases (Figure 1D); (3) when the noise is increased with the signal unchanged, the accuracy drops but the reaction time decreases (unlike what would be expected from speed-accuracy tradeoff; Figure 1E). Our model may allow an unprecedented level of granularity in understanding dual-task interference by allowing examination of the parametric effect of the signal and noise during dual tasks on the speed and accuracy of decisions.