Performance Quiz #6 -- Looking at the sixth cut

re: Performance Quiz #6 -- Looking at the sixth cut

mark — Thu, 19 May 2005 22:15:40 GMT

Thanks. This has been an excellent series, Rico. Showing how powerful the CLR is, as well as showing some concrete examples of how to measure and improve within its walls. I think this series needs to become a cohesive article for MSDN.

re: Performance Quiz #6 -- Looking at the sixth cut

Eric W. Bachtal — Fri, 20 May 2005 00:25:33 GMT

Great series of posts. It was wonderful to see the managed code perform so well right out of the gate. And, your subsequent optimization was simple and effective. Thanks for taking the time and sharing this with us.

re: Managed vs Unmanaged

Hugo Batista — Fri, 20 May 2005 15:23:05 GMT

re: Performance Quiz #6 -- Looking at the sixth cut

C++ guy — Fri, 20 May 2005 19:25:07 GMT

Great series of posts. The CLR is very impressive.

Optimizing managed C# vs. native C++ code

Jonathan Hardwick — Sat, 21 May 2005 02:43:30 GMT

Raymond Chen (aka "fixed more Windows bugs than you've had hot dinners") and Rico Mariani (aka "Mr .NET...

re: Performance Quiz #6 -- Looking at the sixth cut

Nemanja Trifunovic — Sat, 21 May 2005 15:41:08 GMT

>Six versions and quite a bit of work later, >we've been soundly trumped

Excuse me, but this is not fair. Raymond has never tried to beat any "managed" version (I bet he doesn't even know about your effort - you started it after he went to vacation) - all he wanted to prove is that to achieve optimization you need to measure and not to gues where the bottleneck is.

re: Performance Quiz #6 -- Looking at the sixth cut

ricom — Sat, 21 May 2005 22:37:58 GMT

I assure you Raymond doesn't need any defense :)

Remember I work with Raymond, we talked about this series of postings at some length quite some time ago and then again when this got to the top of his queue.

It's all supposed to be an educational illustration. The competitive language is just a facade to make it less of a dry read.

The outcome was never really in question, the journey is the illustration.

Actually I think the fact that Raymond just did what he thought was best (well in advance of my comments) at each stage makes it more interesting, not less.

.NET Efficiency

Alan's Corner — Tue, 31 May 2005 16:26:15 GMT

So I was reading through one of my favorite MSDN blogs (http://blogs.msdn.com/oldnewthing/)

And he...

re: Performance Quiz #6 -- Looking at the sixth cut

Keith J. Farmer — Wed, 08 Jun 2005 01:18:27 GMT

So -- excluding startup time -- what's the performance comparison like?

It seems to me that (total - startup) ~ 0.03s, about half of the total time for v6. If startup for v6 is ~0.00s, then the actual working bits still seem twice as fast in the CLR.

re: Performance Quiz #6 -- Looking at the sixth cut

ricom — Wed, 08 Jun 2005 08:56:45 GMT

No, they each do the job in about 30ms. The CLR has 30ms more startup overhead basically.

Check out the space analysis for the reasons why.

re: Performance Quiz #6 -- Looking at the sixth cut

Ross — Sun, 12 Jun 2005 18:42:43 GMT

That is kind of interesting.

Do you have figures on the difference in memory usage?

re: Performance Quiz #6 -- Looking at the sixth cut

Ross — Sun, 12 Jun 2005 18:50:57 GMT

Oops, just discovered the next blog entry :)

Just because I don't write about .NET doesn't mean that I don't like it

The Old New Thing — Mon, 31 Jul 2006 17:00:26 GMT

I'm just not an expert.

Console.ReadLine() :: C++ is ugly after you’ve been doing C# for a while

Console.ReadLine() :: C++ is ugly after you’ve been doing C# for a while — Fri, 05 Jan 2007 02:42:19 GMT

PingBack from http://console.writeline.net/blog/?p=6