BANGALORE, India—Details of the proposed semiconductor wafer fab to be set up in India are set to be announced by the end of the year at the latest, with talks between the federal government and interested companies getting down to brass tacks.
"The last time the government announced its scheme to have a wafer fabrication plant in India, well, it did not happen," said Sachin Pilot, minister of state for communications and information technology.
Pilot said India's government is more flexible than it had been a few years ago, when a government-backed scheme to locate a chip fab in India failed. "We are far more willing to meet [a potential investor] halfway down the line," Pilot said. Pilot said the government is currently discussing what technology node an Indian wafer fab would be capable of and how long that technology is likely to remain relevant.
Speaking at the India Semiconductor Association’s (ISA) annual summit, Pilot said a decision will likely be made in the next few months, and at the latest, by the year-end.
India's government has said it plans to locate a wafer fab in the country and that the fab would be the beginning of a chip manufacturing ecosystem in India.
"The government will answer this problem by investing in and do whatever else is needed for a fab," said R. Chandrashekhar, secretary of India's Department of Information Technology. "Significant progress has been made in our talks with potential contenders."
Chandrashekhar said India's imminent National Electronics Policy would address the possibility that imports of electronic components and products would overtake the amount spent on petroleum imports within a decade.
The seventh annual summit of the ISA, like the first one, is grappling with the issue of a lack of an electronics manufacturing ecosystem in the country. Speaking at the summit, Aart de Geus, chairman and CEO of EDA and IP vendor Synopsys Inc., said the question of whether India should have electronics manufacturing would be somewhat addressed due to the success of the Akash tablet, the low-cost, Android-based tablet designed by U.K. firm Datawind and built by India-based Quanta, which is distributed to Indian college students and sold commercially at a subsidized price.
"This brings up the question of electronics manufacturing—power availability, distribution, Wi-Fi access, software, fabless semiconductor design—and India must decide which areas it wants to differentiate itself in before spending billions on a fab," de Geus said. "The [pent-up] demand for the tablet computer begs the question of manufacturing in India."
I am bit pessimistic about this development. The decision which is taken by one government is never honored by the new party that comes into power. India will go to election in next couple of years and I am sure government will forget about this project.
If India is serious about this they really need to partner up to accomplish it. They could partner with an established foundry that could bring all the infrastructure, or they would have to partner with all the infrastructure partners (Equipment, EDA, Yield services, process developers, etc.) if they have a chance to pull it off.
Call me a pessimist..but I think it is too late for India to catch up on semiconductor manufacturing. The talks over establishing Fab started 4-5 years back...nothing materialized so far other than talking about the probable upcoming decisions in the conferences.
Let's lock on the date, end of 2012...we"ll review the status again and I'm sure it will be status quo.
Let us say we hear that the fab will happen by the end of 2012. Next step is to get the facilities ready, install, qual process and measurement tools, finish hiring and get to run first Si, followed by yield learning. What matters is, is there any yield. Since there is no experience in this path in India, are we expecting the fab to produce chips by 2015? If so, what type of products and design rule? Will there be any profit to make a difference!
This is too late already. Memory manufacturers are struggling, logic will probably end up at 14nm production by 2014..There is absolutely nothing India can do. Just stop these gimmicks and do something else. Sorry, it is not a frustration from my end, but there is no realistic estimate mentioned anywhere in these proposals.
The Ambani group took over the Fab City and planned a Reliance Fab, they hired some ex-fab guy to come in and do it... But all it was another land deal... can't trust them either... Reliance sold the land at a profit and left fab city in hyderabad high and dry.. thats the record....
While I am sure the issue can be resolved, you really cannot run a fab with the current (no pun intended), frequency of power outages. Even the glitching caused by switching back and forth from local generators could damage WIP.
India needs their private business leaders such as Ambanis, Tatas and Birlas to take interest in such projects. Then only such huge infrastructure projects can be successful. Otherwise it will be another government non-starter.
With the current govt riddled with a lot of corruption charges and a lot of factions opposing many of the upcoming projects ( Jaitapur Atomic power project for example) there is little hope that the FAB project will take off.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.