One can get lost in details about characters, corporate cultures and malfunction. The very big picture seems to be that the company that for decades defined wh

One can get lost in details about characters, corporate cultures and malfunction. The very big picture seems to be that the company that for decades defined what computing looks like could not cope with the end of the exponential era that it embodied. After the fact even outsiders know that in the post-exponential era there are two games in town: fiddling at the high end with complex to build and code parallel/vector units and similarly fiddling at the low end with very low-power units for ubiquitous mobile, iot etc. These bifurcations were no secrets or sudden events, but had long and observable incubation. Missing both those megatrends when you are the superpower at the very center of the industry seems (on the surface) like remarkable incompetence but that is maybe too easy. The high-end is particularly intriguing because Nvidia'approach feels more like a monster created by Intel's failure to deliver what the world needs than the optimal design for high-end computing. Understanding the deep causes of this falling from grace would invaluable for those wondering what the dizzying technology ride has in store for us.