[lkml]   [1998]   [Nov]   [29]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: SMP scalability: 8 -> 32 CPUs

> By the way - this doesn't mean that we don't have an answer for those
> 16-1024 way SMP boxes; we do, that's what the Linux clustering effort
Please delete this mail if i am off-track.
But doesn't the linux kernel on these 16 way upward SMP boxes have 100's or
possibly 1000's of locks to make them scalable upto 16 or more on SMP boxes itself.

Does moving from say 16 to 64 way box means we are taking more locks ??
Or say the locks built in for 16 should work well for scaling for more
number of CPU's too. Excuse me for my ignorance, but to what extent is this dependent
on the hardware architecture ?

> is all about. As we run out of steam on the SMP front, you'll see more
> and more work on the clustering front to scale up well beyond what anyone
> else can do. I'm working on this with one of the government labs - they
> want to scale things to 10,000 CPUs and they darn well don't expect Linux
> SMP to do that :-)
Also just a general question from curiosity perspective, is the performance
of a cluster of these boxes better than a say 1024 way SMP box. This is assuming
we are able to scale it to say 1024 CPUs. This is vis-a-vis compared to a cluster
involving SMP boxes ( totally adding upto 1024 CPU's ).


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at

 \ /
  Last update: 2005-03-22 13:45    [W:0.093 / U:0.424 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site