Reconfigurable Data Flow (1998) [pdf]

"Submitted, rejected and still posted online by the Wild-and-Crazy Session at ASPLOS 1998". We are now at the twenty years anniversary and the ideas are still wild and crazy and increasingly relevant. Reconfigurable Data Flow are now the best way to solve problems which challenge the performance and capabilities of more traditional machines.

As Max Plank observed: "A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it." Or because there is a technology shift and the only way to solve the problem is to rethink the architecture of the machine used to solve the problem.