Fantastic deep dive as always, thanks for doing such stellar work!
On that note, any chance we might get a teardown/history of Cray architectures in the future? Specifically the Cray-1 and 2?
To throw a more serious challenge your way: How about a write-up on the original Frank Rosenblatt Perceptron? I know finding an original Mark-1 part would be close to impossible but it blows my mind that they were successfully doing real-time visual classification in 1957 with an electro-mechanical machine (potentiometers and motors) using a 20x20 "video" feed with some learning algorithm that was not based on backprop.
Comments
No vacuum tube questions.
Fantastic deep dive as always, thanks for doing such stellar work!
On that note, any chance we might get a teardown/history of Cray architectures in the future? Specifically the Cray-1 and 2?
To throw a more serious challenge your way: How about a write-up on the original Frank Rosenblatt Perceptron? I know finding an original Mark-1 part would be close to impossible but it blows my mind that they were successfully doing real-time visual classification in 1957 with an electro-mechanical machine (potentiometers and motors) using a 20x20 "video" feed with some learning algorithm that was not based on backprop.