How not to be a “Spherical Cow”: a quantitative approach

IMG_1466OK, now that I have got your attention with that catchy title, I hope to sustain it for a couple more minutes while you read this #hopefullynotsoboring blog post. My name is Kameswar Rao Vaddina (in short you can call me “Kamesh”). If you can pronounce “Transsendentaalifilosofia”, then my name should be easy peasy. I am a visiting researcher to TUT from Norway and will be working at the department of pervasive computing. For the next 2 and half months, I will be working on minimizing temperature of multicore systems. Continuous scaling of technology, higher on-chip integration, increased power density combined with consumers demanding faster and faster processors is posing some significant thermal challenges for modern microprocessors. Higher on-chip temperature, effects reliability of the system by causing timing uncertainties, reduces performance and increases cooling costs. Temperature is also a multi-scale problem. It occurs at chip, server, rack and datacenter level and can be optimized at different levels of system abstraction. My research mainly focuses on software-based high-level techniques and algorithms to optimize the temperature profile of the system. Still there?? Another catchy phrase time??

Finnish Sauna is nominated for Nobel peace price by Obama – Reuters wire.

Now, back to my blog post. I am not new to Finland. I did my PhD from University of Turku. Yes, from the great and noble city of Turku (can I say that here?). I worked on Thermal-Aware Networked Many-Core systems. My postdoc has been sort of a continuation of that work and I intend to eventually venture into making software applications intelligent and thermal-aware. I tried to be a bit funny in this post. I am not sure if it has succeeded or not. But, in reality I am a very boring and mundane person. I can prove it when you invite me for beer or something (no not Mustamakkara).

This entry was posted in new employee, news, research and tagged , , , , . Bookmark the permalink.