Page 1 of 1

why big pharmaceutical company does't create supercomputer ?

Posted: Wed Jul 24, 2013 10:20 am
by yavuz
hi,
does big companies have super computer for folding ?
for example Pfizer 2012 income is 14 billion $. (source:wikipedia)
if folding calculations is very important for new drugs,
why Pfizer does not buy 100.000 gtx690 and build largest supercomputer with cost of 200 million$ only (estimate) ?

i don't claim that FAH is not very important but i am curious about this topic.
thx for reply

Re: why big pharmaceutical company does't create supercomput

Posted: Wed Jul 24, 2013 1:39 pm
by Joe_H
I don't know whether or not any of the large pharmaceutical companies have computers doing simulations similar to F@H. But F@H is involved in a level of base research that many companies in all areas of technology either do not do at all or do a limited amount. From my experience large companies' research and development groups tend to use the results of base level research done at academic institutions to guide their research among many candidates for future products. In the case of drugs, the published results of F@H may be used to select from a group of potential drugs for further research.

Re: why big pharmaceutical company does't create supercomput

Posted: Wed Jul 24, 2013 2:31 pm
by k1wi
Big Pharma almost certainly use super computers to get their work done. Either on their own private systems or by renting capacity.

Whether they are tapping into the power of GPUs in the same way as FAH, I am not sure. Not only is GPU computing cutting edge technology, but it may require different modelling to the types used by F@H.

Re: why big pharmaceutical company does't create supercomput

Posted: Wed Jul 24, 2013 3:14 pm
by 7im
Did you google this question?

Re: why big pharmaceutical company does't create supercomput

Posted: Wed Jul 24, 2013 3:27 pm
by Jesse_V
I'll second Joe_H's comments. From what I've read, this is how things go. Basic research has to be done before you can narrow down the list of drug candidates, which has to be done before you can choose one or two and start trials.

Re: why big pharmaceutical company does't create supercomput

Posted: Wed Jul 24, 2013 4:43 pm
by 7im
Also, big pharma is profit driven, so their research is targeted at only those drugs with potential to make a lot of money because the development and approval process is so expensive. That pushes big pharma in to different areas of research than what FAH is doing.

So yes, they all have or rent time on super computers to do research, but their priorities are different, hence why they don't run fah. It's not a comment on the worthiness of fah either way.

Re: why big pharmaceutical company does't create supercomput

Posted: Wed Jul 24, 2013 5:20 pm
by NookieBandit
@ yavuz, here's a link that specifically describes the use of supercomputers for drug discovery and some of the players in the industry that rent supercomputer horsepower. http://tinyurl.com/mehqcxl

Extending on Joe_H and Jesse_V's comments, it would be interesting to know if a drug company has actually taken result sets from F@H and used that as the "rule set" for feeding their supercomputer systems in order to shorten drug discovery time.