You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
As AI/ML take off, what company/companies will provide the processors and electronics to dominate? I feel this will become an ecosystem, where the system improves the process as it learns. What do readers think the ultimate partnerships will be the winners since mechanics and electronics are two different areas of research. Any additions on how extreme automation will arrive and when? Medical robots I feel will outrun humans as they accumulate data from market wide systems that actually advance the core data base.
As AI/ML take off, what company/companies will provide the processors and electronics to dominate? I feel this will become an ecosystem, where the system improves the process as it learns. What do readers think the ultimate partnerships will be the winners since mechanics and electronics are two different areas of research. Any additions on how extreme automation will arrive and when? Medical robots I feel will outrun humans as they accumulate data from market wide systems that actually advance the core data base.
Nvidia will lead the robot revolution, in the cloud and on the edge, in partnership with TSMC and many others. Watch Jensen Huang's keynote for GTC 2025.
Nvidia will lead the robot revolution, in the cloud and on the edge, in partnership with TSMC and many others. Watch Jensen Huang's keynote for GTC 2025.
The bigger issue is likely to be that will take only very few humans to actually manufacture stuff, or code, or harvest crops. What happens when the global percentage of people doing manufacturing drops like it did on farms thanks to mechanized agriculture (from 70% in 1840 to 2% today in US). It's going to happen in spite of anything Trump thinks he can do. Just what companies need to do to stay profitable in the face of completion.
I would say if you are picking a semi company that you would expect to have a moat in robotics applications... I would pick STMicroelectronics. They are probably the leader in the kinds of microcontrollers that are used in robotics and other embedded systems. However they are not really a leader in AI.
They are probably the leader in the kinds of microcontrollers that are used in robotics and other embedded systems. However they are not really a leader in AI.
Pretty sure that physical and visual training of AI will be the route to breakthroughs in robotics and other autonomous stuff - that means much more complex "controllers" that can do requisite inference. The big difference from much of the current AI based on transformers and token algorithms is that we'll see variants of training and inference networks/algorithms/computation that are directed at latent space representations, computation and optimizations that are better suited for spacial analysis and decision.
Any thoughts on TSM becoming a robot builder as the closer to the task the processor is the better. Any thoughts, comments or additions on these thoughts appreciated. TSM would be able to provide a level of sophistication and expertise to have large influence on the robotic ecosystem. Maybe TSM just becomes the master of robotic controls ecosystems.
Daifuku is a Japanese sweet made with mochi (a type of rice cake) that is filled with a sweet filling, often red bean paste (anko). It translates to "great luck" in Japanese. Daifuku can also refer to a Japanese company, DAIFUKU that is a leading provider of material handling solutions and automated systems.
Daifuku is a Japanese sweet made with mochi (a type of rice cake) that is filled with a sweet filling, often red bean paste (anko). It translates to "great luck" in Japanese. Daifuku can also refer to a Japanese company, DAIFUKU that is a leading provider of material handling solutions and automated systems.
All AMHSes I've ever seen with my own eyes were of Japanese origin. I am not even sure if non-Japanese AMHSes exist in semiconductor fab industry these days.
The bigger issue is likely to be that will take only very few humans to actually manufacture stuff, or code, or harvest crops. What happens when the global percentage of people doing manufacturing drops like it did on farms thanks to mechanized agriculture (from 70% in 1840 to 2% today in US). It's going to happen in spite of anything Trump thinks he can do. Just what companies need to do to stay profitable in the face of completion.
Excellent observation, robotics and automation of all types is only set to grow at an ever-increasing rate as robotics and AI/ML take off with ever increasing speed. A no brainer if there ever was one.
Any thoughts on TSM becoming a robot builder as the closer to the task the processor is the better. Any thoughts, comments or additions on these thoughts appreciated. TSM would be able to provide a level of sophistication and expertise to have large influence on the robotic ecosystem. Maybe TSM just becomes the master of robotic controls ecosystems.
Pretty sure that physical and visual training of AI will be the route to breakthroughs in robotics and other autonomous stuff - that means much more complex "controllers" that can do requisite inference. The big difference from much of the current AI based on transformers and token algorithms is that we'll see variants of training and inference networks/algorithms/computation that are directed at latent space representations, computation and optimizations that are better suited for spacial analysis and decision.
Interesting that China has flipped the script from an autonomous driving push, to the government and industry touting safety and avoiding ridiculous claims, just before the Shanghai Auto Show 2025.
2025 Shanghai Auto Show: Chinese carmakers hit the brakes on auto-driving hype after crash
‘Safety’ is the big buzzword as firms adjust tone amid Beijing’s closer scrutiny after a fatal crash involving a driver-assistance system
