21 Feb
2012
21 Feb
'12
7:18 p.m.
It was thus said that the Great Toby Thain once stated:
On 21/02/12 12:24 PM, Sean Conner wrote:
I think I used the wrong value when calculating the 21x figure. But
still, 12x is not bad (especially given 22 seconds vs. 4.5 minutes).
-spc
It was thus said that the Great Toby Thain once
stated:
Where do you get 21x from? I see a wall clock ratio of about 12x (still
amazing; can you explain why it's superlinear?) On 21/02/12 12:47 AM, Sean Conner wrote:
I did a test on another sub-section of our codebase at work (this just
happened to have a decently written Makefile). First, a non-parallel make
(same system as above, 8-core SPARC system):
real 4m36.002s
user 4m3.075s
sys 0m29.713s
Now, a parallel make:
real 0m22.330s
user 7m54.787s
sys 0m42.245s
Something like 21 times faster. It was thus said that the Great Alexey Toptygin
once stated:
>On Mon, 20 Feb 2012, Peter C. Wallace wrote:
>
>>Much much faster and lower power for multiple low latency coupled tasks
>
>Right. So call me ignorant, and give me some examples of multiple low
>latency coupled tasks that can't be easily implemented as a single low
>latency task. Why do you need 2+ cores talking to each other? Why not 1
>core that is 2+ times as fast?
Not an embedded system, but at work we have access to a multicore
SPARC
system (Sun, I don't recall the model since it's actually stashed away
in a
data center) with 8 cores. Doing a parallel make (it helps to have a
properly written makefile; I went to the trouble to do so for the part
I'm
responsible for) only takes 1/10th the time of a non-parallel make.
Impressive. I've yet to see a *super*linear speedup for parallel make,
myself.