New challenges
Power draw, density and cost of ownership came to the front — raw performance stopped being the only metric.
Machines had grown to the point where the electricity and cooling bill shaped the architecture as much as peak FLOPS did. Much of the programme was about facilities engineering rather than computation.
Data volumes grew alongside, and storage became a track of its own.
Talks
That same week, on 8–10 October, the 2nd International Conference on High Performance Computing ran in the same building — a separate event with separate registration.
Thursday 11 October · Workshops Igor Sikorsky KPI, Building 1
01
CUDA HPC Workshop
NVIDIA
02
News from the HPC Advisory Council
HPC Advisory Council
03
State-of-the-art interconnect technologies for efficient HPC systems
HPC Advisory Council
04
Extending application scalability through MPI optimisation
HPC Advisory Council
05
Designing clusters around emerging technologies
HPC Advisory Council
06
PBS Pro Users Workshop: maximizing HPC utilization with PBS Professional
Alexander Franke · Altair
Friday 12 October · Technology day KPI, Building 1, 37 Peremohy Ave.
07
Opening remarks
Volodymyr Savyak · organiser of HPC DAY
08
Best practices for accelerating innovation with high performance computing
Philippe Trautmann · HP
09
New advancements in HPC interconnect technology
Tong Liu · Mellanox
10
Kepler architecture. Performance. Efficiency. Availability
Dmitry Konyagin · NVIDIA
11
HP CMU — easy to use cluster management utility
Sebastien Cabaniols · HP
12
Extreme performance networks from Extreme Networks
Alexey Grigoriev · Extreme Networks
13
Altair’s Compute Manager: your gateway to HPC cloud computing with PBS Professional
Alexander Franke · Altair
14
Intel Many Integrated Core Architecture & Parallel Studio
Andrey Semin · Intel
15
Where to store petabytes?
Dmitry Alexeyev · FUJIFILM
16
Computing clusters: construction, installation, service
Volodymyr Osnach · Technology Supercomputers Laboratory