lunes, 28 de marzo de 2016

Israel big data co Alooma raises $11.2m

Israeli big data startup Alooma has close a $11.2 million Series A financing round led by Lightspeed Venture Partners and Sequoia Capital, which also invested in the company's seed round. Alooma, which provides a modern ETL platform for data engineers and scientists, has raised a total of $15 million to date. Based in Tel Aviv, Alooma will use the funding to expand its sales and marketing team in the San Francisco Bay Area and to continue to strengthen its product.

Alooma observes that more and more companies today are maturing out of basic “out-of-the-box” analytics tools as they try to leverage all of their data - from web and mobile applications to NoSql and relational databases to log files, 3rd party SaaS vendors and more. But building an integrated data platform internally is a complex challenge that requires hiring data scientists and data engineers, spending huge amounts of money and many tedious hours capturing, transforming and validating the data.

Alooma CEO Yoni Broyde says, “There is no ‘one solution to fit them all’ in the data market and sooner rather than later, companies realize that their in-house data projects are difficult to manage, impose new risks and significantly hinder their ability to move fast as a business. We enable companies to easily connect all of their data together and make it accessible, usable and valuable to everyone in the organization”.

Alooma enables companies to build and run their own data platform as a service, greatly reducing the time and cost required to leverage disparate sources of data. Companies can act upon all of their data and gain unprecedented decision-making capabilities. Set up in minutes, Alooma lets companies connect all their data sources to their own AWS Redshift or other data warehouse and use Alooma’s Code Engine to clean, transform or enrich any event--all in real time.

One of Alooma's customer success stories, Invoice2Go has grown to handle hundreds of thousands of customers. To continue their growth, Invoice2Go leverages over 15 channel partners in their marketing efforts. However, understanding which partners are most effective requires gaining access to partner data--which is owned and managed by each respective partner leading to more than 40 unique data structures. Using Alooma, Invoice2Go is able to ingest the various types of data formats each partner provides without having to write a custom solution for new sources. “Alooma has improved our business by enabling us to move every piece of data into our Amazon Redshift and allow it to be joined with the rest of our core data,” said Bob Briski, Director of Data Infrastructure, Invoice2Go. “These are problems that are hard to solve and usually need a custom solution, but there’s nothing out there right that can do it, especially at the level that Alooma does it.”

Source

domingo, 24 de enero de 2016

Riphah team develops supercomputer architecture

A team of engineers at faculty of computing and applied science, Riphah International University, has successfully developed supercomputer architecture.
A supercomputer is composed of multiple processors, memory and I/O system while an interconnect mechanism has significantly complex architecture than the ordinary computers, informed the university spokesman yesterday.


The system supports CUDA, MPI/LAM, OpenMP, OpenCL and OpenACC as the programming models and allows to solve larger algorithms & numerical techniques, big data & data mining, bioinformatics & genomics, business intelligence & analytics, climate, and weather & ocean related problems, he said. He added that the benchmarking results show that the system can achieve up to 3.20 terra FLOPS, which can be extended up till 10 FLOPS.
Previously, only a few universities in Pakistan could have developed the supercomputer architecture. When performance of this supercomputer is compared with the existing systems in Pakistan, the results show that the proposed supercomputer architecture stands at the 2nd position. While comparing the performance per watt, this system is ranked at the 1st place.

The system supports libraries using interactive development environments such as Open CV (using codeblocks) used for embedded computer vision, Cuda (Nsight) for GPU parallel programming, ROCCC (C2HDL high level synthesis) eclipse and open MP and open ACC parallel programming models.
Pro Chancellor Hassan Muhammad Khan, Vice Chancellor Prof Dr Anis Ahmed and Executive Director Engr Asadullah Khan of Riphah International University have congratulated the team of engineers including Dr Tassadaq Hussain, Engr Tariq Ali and Abdur Rehman, who have put a lot of efforts to develop this system. With this, the university will be able to carry out complex research projects very efficiently, in order to contribute towards the development of society.

Published in The Nation newspaper on 22-Jan-2016

 

 

Source