106th Day Of Lockdown

Maharashtra2237241231929448 Tamil Nadu118594711161636 Delhi102831742173165 Gujarat37636267441978 Uttar Pradesh2996819627313 Telangana2761216287313 Karnataka2681511100417 West Bengal2383715790804 Rajasthan2140416575472 Andhra Pradesh211979745252 Haryana1799913645279 Madhya Pradesh1562711768622 Bihar12525933898 Assam12523833016 Odisha10097670354 Jammu and Kashmir89315399143 Punjab67494554175 Kerala5895345228 Chhatisgarh3415272814 Uttarakhand3230262143 Jharkhand3018210422 Goa190311568 Tripura171612481 Manipur14307710 Himachal Pradesh107876410 Puducherry104351714 Nagaland6443030 Chandigarh4924017 Arunachal Pradesh270922 Mizoram1971390 Sikkim125650 Meghalaya94432
Science 28 Oct 2017 Keralite in team sig ...

Keralite in team signalling in-memory computing

Published Oct 28, 2017, 6:12 am IST
Updated Oct 28, 2017, 6:12 am IST
Dr Abu says finding would lead to highly energy-efficient computers.
Dr Abu Sebastian
 Dr Abu Sebastian

Kottayam: A team, comprising Ettumanoor native Dr Abu Sebastian and a group of scientists at the IBM Research Laboratory in Switzerland, has shown how to perform computations or calculations in the memory.

This is considered a significant step towards finding an alternate way to compute without consuming so much energy and area.


The finding titled ‘Temporal correlation detection using computational phase- change memory’ appeared in the prestigious journal ‘Nature Communications’ on October 24. Responding to Deccan Chronicle in an email interview, Abu explained the significance of the findings as follows:

“A big focus of research these days is on building highly energy efficient cognitive computers. People are trying to understand how to alter the architecture or organisation of computers so that they are more efficient. Many people are looking at the human brain for inspiration in an attempt to see how the brain computes in such an energy-efficient manner. The new way of computing we have found is in marked contrast to the existing way of computing based on the von Neumann architecture proposed in 1945 by the Hungarian American mathematician John von Neumann. In that architecture, there is a physical separation between where information is stored and where we perform calculations on them.  This means that we need to move around data whenever we need to calculate something. This is one reason why our computers, especially when used for data-centric artificial intelligence and related applications, which are very compute intensive, are highly inefficient,” he said in the interview.

The new way of computing which performs computations or calculations in the memory itself is expected to be a significant step towards addressing the energetic costs of computing.

Dr Abu received his Master’s and PhD degrees from Iowa State University in the USA. He is a scientist at IBM Research - Zurich since 2006.

Click on Deccan Chronicle Technology and Science for the latest news and reviews. Follow us on Facebook, Twitter

Location: India, Kerala, kottayam