| |
 |
| Paper: |
Amdahl”s Laws and Extreme Data-Intensive Scientific Computing |
| Volume: |
442, Astronomical Data Analysis Software and Systems XX (ADASSXX) |
| Page: |
405 |
| Authors: |
Szalay, A. |
| Abstract: |
Scientific computing is increasingly revolving around massive amounts of
data. In astronomy, observations and numerical simulations are on the verge of
generating many Petabytes of data. This new, data-centric computing requires a new look
at computing architectures and strategies. The talk will revisit Amdahl's Law establishing
the relation between CPU and I/O in a balanced computer system, and use this to analyze
current computing architectures and workloads. We will discuss how existing hardware
can be used to build systems that are much closer to an ideal Amdahl machine. Scaling
existing architectures to the yearly doubling of data will soon require excessive amounts
of electrical power. We have deployed various scientific test cases, mostly drawn from
astronomy, over different architectures and compare performance and scaling laws. We
discuss an inexpensive, yet high performance multi-petabyte system currently under
construction at JHU. |
|
|
|
|
|
