MANY ASTROPHYSICAL PLASMAS ARE SO HOT AND DIFFUSE THAT THE COLLISIONAL MEAN FREE PATH IS LARGER THAN THE SYSTEM SIZE. PERHAPS THE BEST EXAMPLES OF SUCH SYSTEMS ARE LOW-LUMINOSITY ACCRETION FLOWS ONTO BLACK HOLES SUCH AS SGR A* AT THE CENTER OF OUR OWN GALAXY OR M87 IN THE VIRGO CLUSTER. TO DATE THEORETICAL MODELS OF THESE ACCRETION FLOWS ARE BASED ON MAGNETOHYDRODYNAMICS (MHD) A COLLISIONAL FLUID THEORY SOMETIMES (BUT RARELY) EXTENDED WITH NON-MHD FEATURES SUCH AS ANISOTROPIC (I.E. MAGNETIC-FIELD-ALIGNED) VISCOSITY AND THERMAL CONDUCTION. WHILE THESE EXTENSIONS HAVE BEEN RECOGNIZED AS CRUCIAL THEY REQUIRE AD HOC ASSUMPTIONS ABOUT THE ROLE OF MICROSCOPIC KINETIC INSTABILITIES (NAMELY FIREHOSE AND MIRROR) IN REGULATING THE TRANSPORT PROPERTIES. THESE ASSUMPTIONS STRONGLY AFFECT THE OUTCOME OF THE CALCULATIONS AND YET THEY HAVE NEVER BEEN TESTED USING MORE FUNDAMENTAL (I.E.~KINETIC) MODELS.THIS PROPOSAL OUTLINES A COMPREHENSIVE FIRST-PRINCIPLES STUDY OF THE PLASMA PHYSICS OF COLLISIONLESS ACCRETION FLOWS USING BOTH ANALYTIC AND STATE-OF-THE-ART NUMERICAL MODELS. THE LATTER WILL UTILIZE A NEW HYBRID-KINETIC PARTICLE-IN-CELL CODE PEGASUS DEVELOPED BY THE PI AND CO-I SPECIFICALLY TO STUDY THIS PROBLEM. A COMPREHENSIVE KINETIC STUDY OF THE 3D SATURATION OF THE MAGNETOROTATIONAL INSTABILITY IN A COLLISIONLESS PLASMA WILL BE PERFORMED IN ORDER TO UNDERSTAND THE INTERPLAY BETWEEN TURBULENCE TRANSPORT AND LARMOR-SCALE KINETIC INSTABILITIES SUCH AS FIREHOSE AND MIRROR. WHETHER SUCH INSTABILITIES ALTER THE MACROSCOPIC SATURATED STATE FOR EXAMPLE BY LIMITING THE TRANSPORT OF ANGULAR MOMENTUM BY ANISOTROPIC PRESSURE WILL BE ADDRESSED. USING THESE RESULTS AN APPROPRIATE "FLUID" CLOSURE WILL BE DEVELOPED THAT CAN CAPTURE THE MULTI-SCALE EFFECTS OF PLASMA KINETICS ON MAGNETOROTATIONAL TURBULENCE FOR USE BY THE ASTROPHYSICS COMMUNITY IN BUILDING EVOLUTIONARY MODELS OF ACCRETION DISKS. THE PI HAS ALREADY SUCCESSFULLY PERFORMED THE FIRST THREE-DIMENSIONAL KINETIC SIMULATION OF THE MAGNETOROTATIONAL DYNAMO (PUBLICATION IN PREPARATION).FOR THE FIRST TIME GLOBAL KINETIC SIMULATIONS OF MAGNETOROTATIONAL TURBULENCE WILL BE ALSO PERFORMED SPANNING MORE THAN TWO ORDERS OF MAGNITUDE IN RADIUS. THESE SIMULATIONS WILL ALLOW THE GLOBAL STRUCTURE OF COLLISIONLESS ACCRETION FLOWS TO BE COMPUTED FROM FIRST PRINCIPLES AND COMPARED AND CONTRASTED WITH THAT FOUND IN PRIOR MHD MODELS. SPECIAL ATTENTION WILL BE PAID TO WHETHER VERTICAL STRATIFICATION RESULTS IN THE FORMATION OF A HOT MAGNETIZED CORONA AND TO WHETHER SIGNIFICANT NON-THERMAL PARTICLE ACCELERATION OCCURS (AS IMPLIED BY NON-THERMAL SPECTRA OBSERVED IN MANY SYSTEMS). FINALLY TO MAKE COMPARISONS TO EXISTING AND UPCOMING SUBMILLIMETER AND X-RAY ASTRONOMICAL OBSERVATIONS THE ELECTRON THERMODYNAMICS AND EMISSION WILL BE MODELED. THIS WORK COMPLIMENTS ONGOING NUMERICAL STUDIES USING MHD IN STRONG-FIELD GENERAL RELATIVITY WHICH SEEK TO DIRECTLY CONNECT THE PROPERTIES OF SIMULATED BLACK-HOLE ACCRETION FLOWS IN CURVED SPACETIME WITH THE OBSERVED MM/SUB-MM EMISSION.WHAT MAKES THIS AMBITIOUS PROPOSAL TENABLE IS THE WIDESPREAD AVAILABILITY OF HPC RESOURCES THE VAST IMPROVEMENT IN NUMERICAL ALGORITHMS FOR PLASMA KINETICS AND THE EMERGING CONSENSUS THAT THE DETAILED PLASMA PHYSICS OF THE UNIVERSE MUST BE UNDERSTOOD IN ORDER TO ADVANCE RESEARCH IN MANY FRONTIER AREAS OF THEORETICAL ASTROPHYSICS. THE THEMES THAT THIS PROPOSAL TACKLES ARE BROAD AND FAR REACHING: THE NATURE OF BLACK-HOLE ACCRETION THE MATERIAL PROPERTIES OF HIGH-BETA MAGNETIZED PLASMAS THE ACCELERATION OF PARTICLES BY TURBULENCE THE EFFICIENCY OF MAGNETIC DYNAMO IN A COLLISIONLESS PLASMA THE INTERPLAY BETWEEN FLUID AND KINETIC SCALES AND THE IMPACT ALL OF THIS PHYSICS HAS ON THE OBSERVED EMISSION. BUT WE BELIEVE THAT THEY ARE ALSO ADDRESSABLE IF A SINGLE PHYSICAL PROCESS ENCAPSULATING THESE THEMES -- NAMELY KINETIC MAGNETOROTATIONAL TURBULENCE -- IS CONSIDERED. THIS IS WHAT WE PROPOSE TO DO.