The 7 things I learned at Semicon West(Gallery below)
Alex Allison works for Camstar Systems, Inc. - a leading provider of Manufacturing Execution Systems. Alex operates within a market development role at Camstar Systems and took the time to write this post for Manufacturing Geek to recap his first experience at Semicon West.
1. It’s BIG. Really, really big (yet smaller than years past)
It was my first Semicon West experience and I had been forewarned about being overwhelmed. Over 30,000 of your closest friends and 720 exhibitors attend Semicon West annually. Yet, despite the huge crowds my experienced colleagues informed me that the turnout was a mere fraction of its golden years.
2. 450 mm Wafers – Thanks but no thanks
Bigger wafers mean higher die to wafer ratios, increased efficiency, and greater productivity, which could transfer into cost reductions. History reflects the adoption of larger and larger wafers, but the problem is that the fabs who have invested the cash to move from 150 to 200 and then to 300 mm wafers have not seen adequate ROI to invest in even costlier machinery. I wouldn’t hold your breath for 450 mm wafers.
3. Where are all the equipment companies?
I was shocked at the lack of large equipment because I was looking forward to seeing and playing with large pieces of machinery. I had heard tales of full scale state of the art lithography and etching machinery that companies like AMSL would ship in with engineers and mechanics weeks before the event to have up and running by the time the show kicked off. This year AMSL had a video that depicted its 193 nm scanner. The vast majority of booths were devoted to accessory suppliers or small low cost equipment such as linear systems and vacuums.
4. Ingots are quite cool!
Ingots are the starting block in which wafers and subsequently chips are made. The growing process begins by refining 99.9999999% pure silicon crystal (sand) into a crucible and submitting it to extreme heat where by a solid silicon seed is lowered by cable into the molten material and slowly pulled (spun) out. Once cooled the ingot can be cut into wafers and sent to a fab for FE (front-end) and BE (back-end) manufacturing. These ingots vary in size but some of the largest can be over six feet in length and 450 mm in diameter. The one I am pictured with here was made by West Coast Quartz (WCQ) company and was said to weigh in at 254 KG and had to be moved with a forklift!
5. India may be the new semiconductor breeding ground
The semiconductor industry may have started in Texas and then California U.S. of A. but it quickly moved across to Europe and ultimately became dominated by Asia due to globalization. Many of the top tier fabs exist in Malaysia, Japan, and Taiwan, while fabless companies are growing by building Asia foundries. China’s semiconductor market had limited success, and bymost of the money and capital was starting to dry up. By domino effect there has been a resurgence of the US semiconductor industry and the market seems poised to accept a new region. It seems that the scales are already tipping in India’s favor. While adequate water and a steady power grid have been hurdles some of the biggest players: Samsung, Intel, and LG are already investing their future in India. Won’t be long before others catch on.
6. Abolishing Moore’s Law
Moore’s law says - over the history of computing hardware, the number of transistors in a dense integrated circuit doubles approximately every two years. For 50+ years Moore’s law defined technological pace, but the bottleneck is that the machinery (most notably lithography tools) can’t keep up with the science. It is commonplace that fabs are using outdated 8-10 year old lithography tools because the cost to replace is exponential. Don’t forget about the basic principles of supply and demand. Very few devices justify the need to have more and more transistors per chip as phones and PC’s have plateaued in size – the only technology we want smaller is flash memory.
7. Analog chips are still relevant
With the digital age moving at light speed it may seem that anything outdated will pass by the way side, but analog chips and subsequently analog technology is not going anywhere . . . at least for a while. We still rely too heavily on dating technology such as temperature, speed, pressure, flow, and human voices which are all controlled by analog devices such as displays, speakers, microphones, antennas, telephone, etc. Companies are still making large profits off these old technologies. While it may not be a strategy for the future I believe that the devices and the technology that support it will be around much longer than people think. Just ask the 100 + million Americans who still have a landline to give it up for a digital phone and see what happens – they might very well send you a fax reply.
Semicon Westwill go down in my book as a success as I was able to converse with tons of people and hear their stories which was the very reason I went in the first place. I believe that Semicon West serves as a great temperature check on industry news and technology but the days where fab companies and big machine makers attend has passed. After all is said and done I will always remember the time I tried to bear hug a 254 KG ingot!