The economy collapsed when technology moved ahead of business and industry found that society could only absorb new technology at a finite rate. Within the EDA and semiconductor industry, our blind faith in Moore's Law contributed to the imbalance. For many years, investors aggressively funded both the supply and the demand sides of Moore's Law. Optimism and the abundance of cheap capital propelled the movement from 0.25 micron to 0.09 micron at an astounding rate.
However, the design gap between semiconductor-manufacturing capability and EDA's ability to deal with increasing complexity still exists. The tyranny of Moore's Law will widen the design gap as finer geometry makes the design problem tougher, which in turn triggers an even more feverish pursuit of Moore's Law to recover lost performance due to inefficient tools.
We should put Moore's Law on hold until we fully exploit the current generation of silicon technology through better tools, methodology and business models.
Let's fix the ASIC design handoff model to reduce the cost and risk of ASIC design. Allow designers to get physical all the way to GDSII, or to stay logical, with silicon virtual prototyping (SVP) tools bridging the gap. Let's get SVP and physical synthesis working together well so that we can conclude the RTL-to-GDSII generation and move to the next-higher level of design abstraction.
Let's create alternative silicon architectures besides FPGA and full-custom ASIC that are inherently free of deep-submicron issues to address emerging market requirements.
Let's cooperate to create an intellectual-property infrastructure to make universal reuse a reality.
Despite the financial carnage, the Internet bubble was beneficial to technology. We may never again see such a massive concentration of human and financial capital devoted to advancing technology. What's left behind is a vast, fully paid-for infrastructure in communication and silicon technology that can be used to build the next wave.
We have the technology and, hopefully, the wisdom to implement the next wave of convergences that can truly change the way we work and live-and can do so profitably. The convergence of voice, data and video over an Internet Protocol network. The convergence of communications, computers and entertainment. The convergence of biology, electronics, medicine and mechanical engineering. The best has yet to come. So, who cares if Moore's Law dies tomorrow?