Additional Information Available Upon Request

Analog Co-Processors for Complex System Simulation and Design

• This research is an exploration as to how computationally intensive engineering and physics problems can be more rapidly solved using our CMOS analog computer chips.
• Our high-speed analog computer chips based on RF/mm-wave circuits greatly accelerate the solution of linear and non-linear partial differential equations arising in many applications.  Those include simulations for electromagnetics, magnetohydrodynamics (such as for nuclear fusion), and fluid mechanics.
• Advances in artificial intelligence (AI) may depend on neural networks trained using algorithms from physics simulations, such as solutions of the wave equation, such as for  recurrent neural networks. Our analog computer chips enable high-speed, low-power solutions of the wave equation.
• What simulations are taking you much too long to wait?   Contact Ocius Technologies to ask about a custom analog computer chip for an order of magnitude speedup, with low power.
• Technical experts: Drs. S.I. Hariharan and PI Arjuna Madanayake
• Funded by a DARPA STTR Phase II grant
• Review a Brief White Paper from early in  2020  here
  

---

---

More accurate proton Magnetic Resonance Spectroscopy (MRS)

---

• ​MRS takes data using a standard MRI scanner, but takes it in a way that proton spectroscopy can be used to identify brain metabolites and their concentration.   It is not yet completely accepted by clinicians, in part because of concern over MRS accuracy.
  
• The transformational algorithm we have developed is based on the complex-valued nature of the data, instead of only its absolute value.  This enables a much simpler, faster algorithm with many advantages.  Currently, Ocius Tech is preparing a joint NIH STTR application for funding.
  
• This project is a cooperative project headed by CTO Dale Mugler and medical personnel from the Medical University of South Carolina in Charleston.