My homepage is going through an early life diapery crisis. Please pardon the mess and send someone to clean it up quick.
Enrolled as a graduate student at Texas A&M University pursuing MS in Computer Science.
Currently working under Dr. Vivek Sarin on high performance computing. Investigating novel means to optimize the sparse matrix vector multiplication routine. Method adopted is to convert data into code, ie: remove one level of indirection (from the core instruction in the"for loop" that iterates through all matrix non-zeros) in order to reduce cache misses and speed up the performance.
Was doing research under Jennifer Welch in the field of theoretical distributed computing, dealing specifically with failure detection in partially synchronous systems and associated areas. Really interesting stuff and may look to continue in it in the future hopefully
My erdos number starts taking up a finite value (though I am not aware which). Paper titled "Parallel Sparse Matrix Vector Multiplication using Greedy Extraction of Boxes" gets accepted at HiPC'10. Glee!!
Will look to expand on my ideas and try to better them; maybe even give them a nice shape...maybe pi or the dzz or d(x^2+y^2) orbitals.
"On a small obscure world somewhere in the middle of nowhere in particular-nowhere,that is, that could ever be found, since it is protected by a vast field of unprobability to which only six men in this galaxy have a key-it was raining."