> The problem is that most software applications couldn't take advantage of all those cores, even after investing hundreds of millions of dollars in computer science.
This is such a tired argument. All these problems have been long solved in Computer Science. Intel had 4000 CPUs in its Paragon System in the 199Os.
The software has been around long enough, my particular lineage is : Newsqueek, Alef, Limbo, Go.
The commercial world chased the Mhz (2 oxen or 1024 chickens? as Cray put it). It's not my fault you all thought C++ and Java was the way to build the future. It wasn't in 1995, it isn't now.
Im not sure that electricity was the main driver here either. From my point of view the impetus was building a processor that was high performance and not needing active refrigeration to keep it from burning up.
Real trend, silly framing: if power bills hadn't been the limiting factor, progress would have accelerated to the point where something else was, and solving that "would have saved the computer industry".
3 comments
[ 1.8 ms ] story [ 23.8 ms ] threadThis is such a tired argument. All these problems have been long solved in Computer Science. Intel had 4000 CPUs in its Paragon System in the 199Os.
The software has been around long enough, my particular lineage is : Newsqueek, Alef, Limbo, Go.
The commercial world chased the Mhz (2 oxen or 1024 chickens? as Cray put it). It's not my fault you all thought C++ and Java was the way to build the future. It wasn't in 1995, it isn't now.