Is This The End Of Moore’s Law?
Moore’s law is unsustainable. This statement is the elephant in the room of a lot of electronics discussions as we rapidly approach a few different landmarks in semiconductors. Researchers are
In order to further the discussion, The Information Technology and Innovation Foundation held a panel discussion and invited some of the most prestigious semiconductor experts. The panel included: Ahmad Bahai, Chief Technology Officer, Analog Business, Texas Instruments; Mark Bohr, Director of Process Architecture and Integration, Intel; Professor Sanjay Banerjee, Director of the Microelectronics Research Center, University of Texas-Austin; Robert Colwell, Director, Microsystems Technology Office, Defense Advanced Research Projects Agency; and Brian Toohey, President & Chief Executive Officer, Semiconductor Industry Association. The discussion, titled, “Are Advancements in Computing Over? The Future of Moore’s Law” was moderated by Robert D. Atkinson, President of ITIF.
Having Intel represented makes sense, given that Moore’s law was named after Intel co-founder Gordon Moore and his 1965 prediction that computer power would double every two years. Up until now, this has proven to be a pretty spot on prediction (though the time frame is actually between 12 and 18 months) with the end result being processing speeds that have increased over 1 million-fold since 1965. However, researchers are starting to realize that without some sort of intervention the technology will hit a few logistical and physical limits around 2020. This is using a silicon-based CMOS semiconductor as an example. Obviously, a decrease in the progress of Moore’s law will have a significant impact on the pace of most of the electronics including cellphones, cars, and anything else that uses the chips. Perhaps, even more alarmingly, they’re used in critical equipment for military and defense systems as well as in the medical field. The people who work in these fields are expecting the same rate of technological advancement as we’ve been experiencing for the past few decades, but it looks like that could all change if nothing is done.
The panel cited lack of funds and lack of semiconductor engineers as major challenges in solving the problem. Problems of this magnitude require large amount of money and a dedicated work force and researchers are struggling.
Collaboration across the government and industry is essential to solving the issue. This is a bedrock technology that’s been instrumental to the success of the majority of electronics, so it deserves the time and money. Unfortunately, though large corporations have been devoting some money to the cause, it’s probably not enough to solve the problem before it becomes critical.
In even more alarming statistics quoted by the panel, every second, 640 Terrabits of data are traveling through the internet and that number is doubling and quadrupling every year, so even if engineers were able to sustain Moore’s law, it wouldn’t really matter because it can’t keep up with those advancements. In three years, the amount of video will be 16 times as much as we’re viewing now. Not even close to something we’re capable of sustaining.
Moral of the story? We’ve got a lot of data and a lot of people depending on researchers to find a solution to a looming problem that could really impact manufacturing and health. The good news is that panels like this bring the discussion in to the light. The bad news? It might already be too late.
Listen to the whole discussion, here: http://www.itif.org/media/are-advancements-computing-over-future-moore-s-law#audio