STAC-A2 on IBM Power System S824 with IBM Power8 processors
STAC-A2 on Phi-enabled HP Apollo-based ProLiant XL250a Gen9 server
kdb+ 3.2 with NVMe-based Intel SSDs on Lenovo System x3850 X6 under STAC-M3
STAC-A2 on AWS: Initial price-performance comparisons of EC2 instances
2-socket Power8-based stack sets multiple records
Leading-edge performance for this amount of CPU cores and server memory on an Intel architecture
Initial price-performance comparisons of ten different EC2 instance types.
Fastest kdb+ system in 7 of the 17 response-time benchmarks
Cloud-based solution sets 5 of 17 response-time records.
Topics included new approaches to realtime analytics, technology learnings from other latency-sensitive industries, new approaches to compute-intensive workloads, tips for developing and deploying parallel applications, advances in low-latency networking, and new approaches to financial time series data and data for regulatory reporting.
Some people learn their word-of-the-day from calendars and web sites; I am lucky enough to interact with financial services technologists who bring their own colorful language. As an example, the response I got from one when I asked about SQL-on-Hadoop: chimera.
Our recent research on “big data” use cases uncovered many large financial services firms using Hadoop and other technologies to build “multi-tenant” analytics platforms—that is, shared environments meant to support multiple independent applications. In some cases, these platforms are being built in advance of demand, trying to preempt individual businesses from deploying