Monday, October 25, 2010

Ubuntu Developer Summit/Linaro Developer Conference


I am currently participating in the Linaro Developer Conference.It is being held along with Ubuntu Developer Conference in Orlando(24-29 Oct10).Linaro is a Multi company initiative to standardise the Linux Kernel and middleware for the ARM based platforms.This will reduce redundancies and improve the time to market.

I had my presentation on Multimedia Codecs.It was received well by the audience.The presentation was about multimedia basics and the developement methodology of codecs and its best practices.This was for a wider audience from a variety of backgrounds to bring everyone to the same page.I will be further conducting sessions on Codec development for the next release cycle, day after tomorrow.The presentation is available at

https://wiki.linaro.org/WorkingGroups/Middleware/Multimedia/Specs/1105/OptimizeCodecBestPractices

Sunday, October 17, 2010

The necessity of National UID(Addhar)(?)

I was going through the UID documentation and was not able to find any strong reason as to go for a new numbering system considering that the already exisiting IT PAN(Permananet Account Number) could have done the same job comfortably.

In fact the Finance Ministry had decided to add biometric data to the PAN data base to plug leakages in terms of bogus cards.The PAN is already being used by crores of people.This number could as well have been used by other govt departments for delivering some other set of services.

The UID Authority will be paying Rs 50 for every new enrollment to the registrars.This money could have been saved for a few crore people,if PAN was used.

I am sure they would have done a feasibility study considering that Nilekani is heading the project.But,the govt(quasi govt) agencies more often that not have embarked on some redundant projects to shelve it later.The MAPIN project comes to mind.This was mandated for all mutual fund holders.Whereas PAN could have been used for the same.At the last moment MAPIN was scrapped and PAN was used in its place.The Voter ID card is one other redundant project.The mutual fund KYC(Know your Costomer) is another bogus policy being implemented.People can easily be tracket using PAN.

Nilekani is the Cenral government advisor for IT projects.Hope this will help prevent further redundant e-gov projects in the future.

Friday, October 9, 2009

Technology education and skills shortage in India.

According to a recent article in the Financial Times,
India is facing a shortage of civil engineers. If India
has to sustain a growth rate of 9% then the number
of civil engineers needs to rise threefold. India
currently produces around 10000-12000 civil engineers
per annum.

I have a number of observations in this regard.
i) Most of the good civil engineering students are
recruited by the IT companies from the colleges.
These companies are mostly Tier 1 service companies
with good training infrastructure. These candidates
are given training for 3-6 months and are redeployed
for software services work.
ii) The IT companies are forced to recruit candidates
from other domains due to the shortage of engineers
with IT specialization in engineering.
So, the shortage of civil engineers is further aggravated
due to their poaching by the IT companies. Secondly,
the crème de la crème of the civil engineers move out
of the domain into IT. This overall affects the quality of
civil engineering work in the country.

I think the problem can be solved by the massive
increase of technology education in the country.
Secondly, there should be proportionately more
IT seats in engineering colleges to meet the needs
of the IT industry. This will ensure skill set and
requirement match. It will also save the training
costs of the IT companies.

But, finally in the long run market will take care of the
shortage of civil engineers. The shortage will drive
up the salaries of the civil engineers.This in turn will
incentivize students to join the stream. This will also
stop the migration of civil engineers to the IT sector.

Thursday, September 17, 2009

Web 2.0 and politicians

First it was Omar Abdulla and now it's Sashi Tharoor.Omar's blog was hugely successful but due to a lot of unfavourable comments made on his blog,he quit the Web.Tharoor is in the face of a controversy over a tweet made on Twitter regarding MP expenditures.The partly is not taking it kindly.Whatever the arguments and counter arguments my interest is in the technical aspects of the whole issue.In fact Tharoor has become one of the ambassadors of the Congress Party in Web 2.0.The Congress has not been very effective in using the net to communicate or campaign.
They have a rarely updated website.The BJP fared far better in this aspect.Now if Tharoor is forced to quit twitter,I think it will also scare away other politicians from using the Web 2.0.It remains to be seen what is Tharoors response to the whole controversy.

Semiconductor companies and Airline companies.

Warren Buffet had said that the Airline industry is a strict no for him.I guess he might have said the same thing about the semiconductor/Electronics industry.Since joining the industry more than 7 years ago,I have seen the industry in bad shape.Even during the boom for 2005-06 ,companies were in bad shape.The argument then was that the problem is in over capacity/supply and that the situation will pass once demand goes up.But,it was not to be as the sub prime crisis and recession started after that.Earlier it was supply side problem and now it is a demand side problem.Though one must concede that a lot of consolidation(M&A) happened to to solve the supply side problem.But,unfortunately most of these M&A's were failures.I will just give a few example to illustrate it..
i)Motorola acquired Symbol.But,most of the senior management/talent had to leave the company.Considering that the key takeaway for a successful acquisition is human resources,it seems to me to be an utter failure.

ii)Motorola spinoff Freescale has been a sad story all the way.It is in the selling block now.

iii)Infineon is sustaining mostly through it lucrative memory business.

iv)Siemens mobile could not be saved after the acquisition by Benque.It was an utter failure

ii)Techtronics acquired Hughes and finally selling it.The very idea of creating synergy out of the vertical integration backfired.

iii) NXP is spinning off new companies ever other day.One of my friends joined Connexant.Then he got transferred to NXP as a part of an acquisition.Now he is in ST-NXP due to another M&A.All this within a 2 years time.

Wednesday, September 16, 2009

Mobile Phones and Connectivity-Solutions

I think that the solution to the problem as I posted in the last post is actually a very big opportunity for the semicon companies.Phones with higher computing power at affordable
rates should be made available the masses.This is possible by having a higher volumes
strategy with lower margins for higher computing power phones rather than treating them as premium products with higher margins.Secondly,the price of these phones can be brought down by using open platforms in them and selling them with customised software.So,that users only pay for what they need rather than making them pay for a host of exotic softwares on the phone which they will never use.

The second problem of poor quality connectivity will be solved by the coming of 3G services.

Tuesday, September 15, 2009

Mobiles and digital connectivity--problems

Looking at present statistics,mobiles seem to be the only gadget which has the highest penetration in India.Looking at the wide penetration of mobile phones,many experts are proposing to provide a host of services using the mobile phones.The service includes Internet connectivity mobile banking etc.No doubt the idea makes a lot of sense as it will provide access to a host of services to a wide array of citizens.But,I see two major problems with this idea.

(i)Low End Phones:-
Most of the phones being sold in India are entry level with very low computing power.They have very simple or no application processors with low computing power.The focus is more on the base band processing part for voice services.To run some applications,it needs to have a decent application processor.Looking at the current scenario,even mid range phones struggle to run ordinary applications.The so called higher end phones are exorbitantly costly.

(ii)Low speed Internet connectivity:-The supposedly 2.5 G services like GPRS provide very low speeds and practically unusable from mid range phones.Even if the connectivity is decent ,the data charges are exorbitant.Airtel charges 399 for downloading 500KB in Blackberry.

---Next post I will offer some solutions..