Lars is a polymath scientist in AI ultra high performance computing, machine learning (eight fundamental machine learning patents), RF/uW, analog mixed signal integrated circuits and algorithms, solver intractable problems for the DoD and industry through novel invention, development and application of hardware and software. Lars has built a vast career in AI, machine learning, novel algorithms, quantum physics, therapeutic small molecule drug discovery, ab initio quantum chemistry, brain cancer fundamental mechanism basic research, microelectronic, optoelectronic, superconducting and analog/hybrid VLSI design, extremely low temperature and semiconductor solid state physics and quantum computing research. He was the founder and director of the (GTE) $10M Advanced Machine Intelligence Laboratory where he was awarded the company research award for his patented machine learning technologies. His machine learning technology was responsible for the first large scale neural network DoD system. Lars is inventor of the SPECTRE hybrid neuron neuromorphic VLSI machine.
The SPECTRE Multi-Chip Module on dielectric (MCM-D) is implemented using micro-cooled industrial diamond or on a superconducting thin film for ultimate performance. SPECTRE module estimated performance is 10^15 neural computations per second.
SPECTRE enables ultra performance unbiased unsupervised reinforcement machine learning. The SPECTRE proof of principle application is data dependent cryptanalysis.
© LARSX And All Patents Pending, No Unauthorized Use Of Content Without Written Permission From LARSX ©