workloads

A few minutes of googling suggests that they often run at a 95% hit rate on typical workloads and a 99% hit rate on compute workloads.

These supercomputers are highly engineered for low network latency between nodes, which is necessary for many scientific workloads.

It isn't a fair comparison though, since a supercomputer with that throughput isn't working on incoherent serial workloads.

Though you might want to offload heavy workloads to a droplet/linode if you go for the new Macbook.

So much of the effort in modern HPC systems used for Hadoopy workloads goes to latency hiding.

It's one of those cases when an emulator or hypervisor outperforms bare metal under some workloads.

In fact when you start thinking abstractly like this, other tech such as Riak makes sense for a lot of workloads.

Kubernetes focuses on _what_ Docker should run, and how to move those container workloads around.

NoSQL is ancient, and the value of NoSQL is in specialized data models and workloads.

Ruby performance has not improved substantially since 2.0, and we should almost always be faster for straight-line workloads.

To add to the others, in a nutshell: - Mesos is a generalized, low level framework for distributing workloads over multiple nodes.

Postgres never updates data in-place, and certain workloads can never be as fast on Postgres as using a different storage engine such as MySQL/TokuDB.

However,when simulating a nuclear explosion or a weather system (maybe that's what you mean by "real" workloads?), the heavy node-to-node communication makes the supercomputer much, much better suited.