I think this is more a question of architecture Can you write simple code that takes a bit of data process this and puts the data back on the stack, if so you could run say 1400 process (aws or the likes) for 1 min each day
You could also afford to run less efficient code as AS3 on the server Time is expensive CPU is cheap Sent from my iPhone > On 14 Aug 2016, at 05:54, Justin Mclean <[email protected]> wrote: > > Hi, > >> Therefore, I assumed that a single threaded process used one virtual core on >> a CPU. Is that >> incorrect? > > Yep. > >> Thanks for putting Spark on the map for me, I had never even heard of it, >> but is it as fast as a C program? > > On a single core probably not, but it's much more scalable and could easily > finish the quicker given enough cores/CPU. How many it would need I couldn’t > guess at. But I done a little bit of Apache Spark programming and found it > quite fast. On my laptop things like work out what are the most common words > in the entire works of Shakespeare took 10 seconds or so including read and > parsing the text. > > One other option on a single machine for fast number crunching is to use the > GPU for calculations [1], but I’ve not played about with that. > > Justin > > 1. http://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html
