[Cryptography] Speculation considered harmful?

jamesd at echeque.com jamesd at echeque.com
Fri Jan 5 02:02:33 EST 2018


On 1/5/2018 2:32 PM, Tony Arcieri wrote:
> The "swarm of puny cores" approach is a bad fit for a lot of problem 
> domains, so I'm curious to see if RISC-V implementations can evolve into 
> something a bit closer to what we'd use for typical "business 
> logic"-heavy server workloads.

Single thread of execution maxed out a long time ago.

We still have not figured out how to take proper advantage of multi core 
and multi thread.

But we are improving our ability to deal with distributed processing. 
In principle, most problem domains could be dealt with by a swarm of 
puny cores.  It is just that we lack a suitable language and a suitable 
processing model.  C++ is still struggling to integrate multi processing 
into the model computer that the compiler targets.

Node.js, an interpreted Javascript language, has a good start on the 
problem, as do the latest operating systems implementing mapreduce,

C++ and Rust are still wandering lost in the woods, despite their 
gigantic efficiency advantage over interpreted Javascript.

Mapreduce frequently gives stupendously poor performance, even though it 
is scalable.  We don't have a high level way of addressing multiprocessing.


More information about the cryptography mailing list