Common HPC Resource for all researchers in India
0 have signed. Let’s get to 100!
Please forgive the way of presentation of this request letter: it is not a petition. The change.org website in our case provides a convenient way for researchers in India to express their support with the ideas given below, and hopefully some will add their individual perspectives.
- Provide enough computing resources for chemistry, physics, biology, mathematics, climate change and others.
- Introduce new platform - CompaS - Computing as a Service
- Researchers to concentrate only on research problems, but not on computing challenges.
- User friendly trainings and workshops on regular basis at different locations.
- Proper 3 level electricity backup and 2 level network connectivity with plug and play nodes for seamless computations.
- Better to initiate by the Indian Government rather by a corporate or private entity.
Need of CompaS
- Many premier institutions in India have in-house/central/group wise HPC facility. But then how can other research/academic research groups could get access to HPCs for a nominal fee.
- Lots of scientific research instruments left unused in many institutions and if those funds could be utilised for a common facility, then many researchers across India will get benefited for even more productive research.
- Running simulations or models with different atomic configurations and parameters. Each simulation takes atleast 8 hours to complete, and researchers require >100 iterations but due to limited resource more time taken.
- When a researcher is about to submit the last simulation, they need to finalize the research article. To their suprise, ‘qstat’ shows 800 tasks got submitted in the last hour and their job is on the queue leads to more waiting time.
- Researcher thought their code/job would run for 3-6 hours. On assumption, declared it might go as long as 4, just to be safe. But it never made it. After 4 hours, their task died at 98% completion, and they are now without results and penalized for abusing the cluster policy or wall-time. How could they know?
- If researchers have some weekend calculations they want to run. After some failed ‘qsub’ commands they realize computing resource is down for maintenance.
- Complete list of applications/softwares/codes/libraries installed in the HPC. For reference: https://help.rc.ufl.edu/doc/Applications to cater all research area.
Our execution strategy incorporates advanced compute cluster computers with high performance computing with advanced interface to researchers who can perform their scientific calculations via web browser, mobile apps and seamlessly monitor the job updates without any restrictions or limits or breakdowns for successful research findings. Proven hardware and network technologies, extremely qualified personnel, and a highly responsive approach to managing our cluster computing technology.
We are aware that C-DAC is working on National Supercomputing Mission (NSM). The NSM is being jointly implemented by the Department of Science and Technology (DST) and Department of Electronics and Information Technology (DeitY) at an estimated cost of '. 4,500 crore over seven years. Of this, so far the DST has released . 85' crore and DeitY, . 25 crore. But then this is a new initiative for shared supercomputing resources. (Taken from official webpage)
Today: Girinath is counting on you
Girinath Pillai, PhD needs your help with “Common HPC Resources for all researchers in India : To be part of NSM”. Join Girinath and 52 supporters today.