Intel Corp. may have announced groundbreaking at its first front-end
300mm fab in Asia (Dalian, China) but more specifically, according to
Intel in its press release; this symbolises Intel's manufacturing
leadership, ability to cultivate engineering talent, acceleration of
the growth of the IT ecosystem as well as Intel's culture of
environmental leadership.
When Fab 68 goes into production in 2010, Intel believes the fab ($4
billion total investment) will spur the revitalisation of industries in
China's Northeast region, in part by the need of equipment and
materials suppliers to have support operations close by.
Heavy
stuff indeed for an average-sized 300mm facility (15,000 square meter
clean room) that will not be ramped straight away. When it does ramp it
will have to use trailing-edge technology!
I know news of this nature receives a certain level of hype, but can Intel really claim all this and get away with it?
Firstly, Intel has chosen a greenfield site far, far away from developing clusters in China such as Shanghai and Beijing.
This
is the first time Intel has not built a 300mm fab within an existing
cluster, which adds risk and costs. However, the site selection experts
at Intel rarely get their sums wrong so the decision to establish a fab
base in Dalian is intriguing.
Perhaps Intel is concerned about
the security of processing and design IP. Isolation may be perceived by
the company as its best tactic to protect its IP, but I remain curious
as to why they picked this location!
The mere fact that this is
a greenfield site will mean that China's regional agencies will be
investing in all the key infrastructure supporting a 300mm fab will
require. Secure reliable power supply, good road and airport links as
well as fast track customs and distribution facilities are required,
but how good is the airport and its connections?
Equipment and
materials suppliers and a whole host of related support industries will
have to set up offices, facilities and staff local to the fab, which
typically generates around 5,000 jobs in addition to those required by
Intel directly.
This can be looked on as an up-front cost for
those suppliers, but the purchasing power of Intel means that few will
refuse to bear that expense.
This all results in a significant
boost to the local economy over the lifetime of Fab 68, rendering
Intel's economic impact and the companies attracted by the endeavour as
significant.
Again, because Fab 68 is on a greenfield site,
Intel has to invest money in the local university educational system if
it is going to find the right pool of engineering skills.
Not
surprisingly, Intel also announced that it was partnering with the
Dalian University of Technology and the Dalian Municipal Government to
establish the Semiconductor Technology Institute. This would mean
provision of a complete 200mm wafer processing tool set for training
purposes.
Although I am sure Intel will want to be guaranteed
the cream of the technology institute, support industries will also
need skilled staff. The recruiting scope has to be made larger to
provide for more than just the needs of Intel.
However, Intel
has to start from scratch, meaning that in that respect, the company is
‘helping to cultivate engineering talent,' necessary to get the fab off
the ground.
The point made about environmental stewardship at
its fabs is nothing new from Intel. Regardless of where in the world it
operates and regardless of the type of facility, the company has global
EHS strategies, policies and methodologies. Fab 68 will be no
different.
However, the wording in the press release gives the
impression that some or all existing fabs in China are not as
environmentally friendly as Intel will be with Fab 68.
Since
when can Intel claim the high moral ground on environmental leadership
in the semiconductor industry - or any industry, for that matter?
Intel
claimed that Fab 68 would ‘accelerate the growth of China's information
technology (IT) ecosystem.' Two broad terms - IT and ecosystem - that
have become almost meaningless these days!
I won't ask Intel to
actually explain that one or quantify the ‘acceleration' element across
China as that is probably a job for a team of economists to determine
sometime in the future.
The final claim made by Intel for Fab 68 centred on the manufacturing leadership such a facility gives Intel.
Yes,
Intel has more fabs than its direct competition, but it is being caught
up with and surpassed when you throw in memory manufacturers to the
equation. The fact that the fab will be trailing-edge makes such claims
almost meaningless.
That said, a $4 billion investment over a
period of years while investments continue to take place in other Intel
clusters is of course impressive and duly noted.
Perhaps the
key aspect of Intel's move to build a 300mm fab in China is that it's
the last piece of the jigsaw to be put in place in the country. Intel
has two assembly and test plants in Shanghai and Chengdu, along with
R&D centres and labs in Beijing, Shanghai and elsewhere.
As a foreign company, that and that alone is very impressive.