Split-Precharge Differential Noise-Immune Threshold Logic Gate (SPD-NTL) Suryanarayana Tatapudi, and Valeriu Beiu 1 School of Electrical Engineer and Computer Science, Washington State University Pullman, Washington 99164-2752, USA {statapud, vbeiu}@eecs.wsu.edu Abstract. After a short review of the state-of-the-art, a new low-power differ- ential threshold logic gate is introduced: split-precharge differential noise- immune threshold logic (SPD-NTL). It is based on combining the split-level precharge differential logic, with a technique for enhancing the noise immunity of threshold logic gates: noise suppression logic. Another idea included in the design of the SPD-NTL gates is the use of two threshold logic banks imple- menting f and f_bar, and working together with the noise suppression logic blocks for enhanced performances. Simulations in 0.25 µm CMOS @ 2.5 V show the functionality of the gate up to 2 GHz. An advanced layout based on high matching centroid techniques is currently under development. 1 Introduction Research on neural networks (NNs) started sixty years ago. The seminal year for the development of the “science of mind” was 1943 when the article A Logical Calculus of the Ideas Immanent in Nervous Activity by McCulloch and Pitts was published [29]. For modeling a neuron, they introduced the threshold logic (TL) gate (TLG): ) , , ( 1 n x x f K ( ) θ sgn 1 − = ∑ = n i i i x w (1) where w i is the synaptic weight associated to x i , θ is the threshold, and n is the fan-in. The general belief that a neuron is a TLG can be questionable. That is why, the TL model has been tested on a spike train generated by the Hodgkin-Huxley model with a stochastic input [22]. The result was that the TL model correctly predicts nearly 90% of the spikes, justifying the description of a neuron as a TLG. The tremendous impetuous of VLSI technology has made neurocomputer design a lively research topic. There are many theoretical complexity results showing that TL circuits (TLCs) are more powerful than classical Boolean circuits [9]. Beside, TLGs 1 V. Beiu is partly sponsored by the Air Force Research Laboratory under agreement number F29601-02- 2-0299. The U.S. Government is authorized to reproduce and distribute reprints for Governmental pur- poses notwithstanding any copyright notation thereon. The views and conclusions contained herein are those of the author and should not be interpreted as necessarily representing the official policies or en- dorsements, either expressed or implied, of the Air Force Research Laboratory or the U.S. Government.