Intel Continues With More Big-Time Optimizations To The Linux Kernel

[unwind-protect]

So are we behind or ahead of Windows with improvements like this? Has all this stuff been communicated to Microsoft ahead of time? Or are all OSes randomly finding these things?

[sinepgib]

So are we behind or ahead of Windows with improvements like this? Has all this stuff been communicated to Microsoft ahead of time? Or are all OSes randomly finding these things?

This is software development in a nutshell. Windows is just less public about it. Nobody writes perfect software, so it definitely will contain inefficiencies, which you sometimes will find and fix, and sometimes not.

[unwind-protect]

This is software development in a nutshell. Windows is just less public about it. Nobody writes perfect software, so it definitely will contain inefficiencies, which you sometimes will find and fix, and sometimes not.

For all we know FreeBSD could be way ahead of everybody else

[timofonic]

For all we know FreeBSD could be way ahead of everybody else

Wasn't OpenBSD?

[GreenReaper]

In terms of things like "hyoerthreading is bad mojo, let's turn it off", I guess so.

[microcode]

Exactly. You'll have a lot of devs and admins now making the case in certain datacenter arenas and other network heavy setups kind of saying "see, I told you so" instead of pouring tons of $$ into hardware upgrades to fix the issue when it's just an oversight in code being revealed by a bottle neck. Do tools exist to hunt for these kinds of bottlenecks or is it just a dev staring at something and trying to figure out why something should be going way faster than it is but isn't?

Well, I think there are a lot of these that are known issues as well as many that can be discovered methodically by figuring out a theoretical maximum throughput for given hardware, and checking if there's a difference in order of magnitude between that and the actual throughput.
Latency is harder to work out methodically.

[liam]

So are we behind or ahead of Windows with improvements like this? Has all this stuff been communicated to Microsoft ahead of time? Or are all OSes randomly finding these things?

Regarding concurrency? Pretty far behind. If nothing else, nt & .NET can watch over huge numbers of connections while using very little cpu, and, since they are (or were) designed to be used together, you'll see very efficient thread pool implementations. I think one of the big reasons behind this is that Windows is better at tracking tasks with context.