Ct: C for Throughput Computing

As the computational power of Intel architectures grows (by reducing transistor size and adding some more core ), new programming tools are required to exploit this new power. I actually prefer a different approach/solution to the problems that affect micro-electronic at the current stage: I think ILP more than MT could be the right programming model for the next generation architectures, but at the same time think that MT is the most suitable by the "market" point of view. Probably there are a lot of "things to do" by means of MT before introducing programmable logic into the consumer-level core.
Here we find a good example of "improved C/C++" to make easier to exploit MT on Intel's core. A very interesting use case is described by this white paper on option pricing. Montecarlo algorithms take advantages from MT.