AI \u2013 a growth engine for semiconductors<\/strong><\/p>\n\n\n\n It is mind-boggling\nto see how fast the semiconductor industry is advancing and how much investment\nand research goes into AI and machine learning (ML). Given the enormous\ninvestment both at the corporate level and in venture capital, there is little\ndoubt in my mind that AI will be at the core of the next industrial revolution\njust as much as in the next \u201ckiller products\u201d.<\/p>\n\n\n\n It has been this way\nfor the past 40 years \u2013 whether it\u2019s the Palm Pilot, the smartphone or the\nsmart watch \u2014 they all started with limited capabilities and were fairly\nclunky. Thanks to Moore\u2019s law, the next generation is twice as beautiful, sleek\nand powerful, and makes us want to open our wallets.<\/p>\n\n\n\n In 2018, venture capital funding of fabless semiconductor\ncompanies in AI exceeded the level for the entire industry in 2002. Like a\nseismic force, companies are creating domain-specific AI and deep learning\ncapabilities. By an informal analysis, there are around 40 companies\nspecializing on vision applications and about the same number focused on\ncloud\/HPC AI and AI applications on the edge, respectively. <\/p>\n\n\n\n There\u2019s reason to\nassume that we will continue to see more and more powerful, power-efficient and\nhighly specialized chips for AI and ML applications, enabling more and more\naccurate applications with amazing real-time responsiveness.<\/p>\n\n\n\n The semiconductor playing field is shifting <\/strong><\/p>\n\n\n\n At the same time,\ncompanies not known as \u201cchip designers\u201d or manufacturers are increasingly entering\nthe scene: not only Google, Facebook, and Amazon, but also Ford, Lockheed\nMartin, Boeing and Northrup Grumman, just to name a few.<\/p>\n\n\n\n The simple reason for this focus is the potential for\ndifferentiation. For any manufacturer of a consumer or industrial product,\nhaving the fastest, most robust and most power-efficient hardware can give them\nseveral years of lead. For example: advanced driver assistance systems (ADAS),\nindustrial vision applications, autonomous control and other disciplines.<\/p>\n\n\n\n Especially when it\ncomes to advanced applications and AI, systems companies may have an advantage\nto overcome one of the big bottlenecks: the right data. <\/p>\n\n\n\n Availability of data is still a limiting factor for many\nAI applications<\/strong><\/p>\n\n\n\n Oftentimes, we are\nled to believe there is an abundance (if not a deluge) of data. However, there is\nquite a number of highly-relevant industrial applications of AI and ML where\nthat\u2019s not the case at all.<\/p>\n\n\n\n Take autonomous\ndriving as an example.<\/p>\n\n\n\n![\"\"](\"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/19\/2020\/01\/notifier-app-seidel-180910-eci_medium.jpeg\")
<\/figure><\/div>\n\n\n\n![\"\"](\"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/19\/2020\/01\/Companies-creating-domain-specific-AI-1.jpg\")
<\/figure><\/div>\n\n\n\n