Effective Priors in Diffusion Image Science of the Brain

Diffusion magnetic resonance imaging allows us to measure microscopic features of the brain without using “wet lab” techniques. This opened, now for over two decades, a vast array of investigations into the relationships between brain structure and psychological and psychiatric behavior. We will present a brief introduction to the imaging technique and then show how we can infer boundaries of various tissues and pathways within the brain by effectively using histological and Markov priors.

Computer Architecture Seminar: Making Graphics Processing Unit Memory Systems More General Purpose

Graphics Processing Units (GPUs) are being re-purposed to perform general-purpose computations as they offer the potential for better performance and and lower energy than conventional CPUs for some workloads. Effective use of GPGPUs, however often requires the programmer and/or runtime software to explicitly manage the logical name (address) and/or physical location of data. It is our hypothesis--shared by some--that GPGPUs can be made more generally effective if all CPU-GPU program threads can access data via a uniform virtual address backed by hardware data movement support.

Computer Architecture Seminar: DataFlow SuperComputing

This presentation analyses the essence of DataFlow SuperComputing, defines its advantages and sheds light on the related programming model. DataFlow computers, compared to ControlFlow computers, offer speedups of 20 to 200 (even 2000 for some applications), power reductions of about 20, and size reductions of also about 20. However, the programming paradigm is different, and has to be mastered. The talk explains the paradigm, using Maxeler as an example, and sheds light on the ongoing research in the field. Examples include Physics, Engineering, DataMining, FinancialAnalytics, etc.

Disengaged Scheduling for Fair, Protected Access to Fast Computational Accelerators

(Joint work with Konstantinos Menychtas and Kai Shen)

With GPUs and other computational accelerators assuming an increasing
share of the workload on modern machines, we argue that the operating
system must assume responsibility for accelerator scheduling and
resource management. This goal is particularly challenging on systems
that allow user-level applications to interact directly with the device.

Inside the Pure Storage FlashArray: Building a High Performance, Data Reducing Enterprise Storage System from Commodity SSDs

The storage industry is currently in the midst of a flash revolution. Today's smartphones, cameras, and many laptops all use flash storage, but the $30 billion a year enterprise storage market is still dominated by spinning disk. Flash has large advantages in speed and power consumption, but its disadvantages (price, limited overwrites, large erase block size) have prevented it from being a drop-in replacement for disk in a storage array.


Subscribe to UW-Madison Computer Sciences Department RSS