Charles Oliver Nutter schrieb:
> Jochen Theodorou wrote:
>> I think there is not enough data to see a trend. I modified your test, 
>> made it run from 1-20 threads and for 50 loops, making an average time 
>> containing the time it took to execute all threads and put these in a 
>> diagram. I used a Q6600 Quadcore intel CPU with java 1.6.0_03-b05 on 
>> Linux 2.6.22-14-generic #1 SMP x86_64 GNU/Linux. What I can see is that 
>> the time constantly goes up until 4 Threads are reached, my number of 
>> CPUs. Using 5 threads is takes less time than using 4, but after that 
>> the time looks more or less constant.
>>
>> This looks quite scalable to me.
> 
> You did notice the code splits the number of iterations by the number of 
> threads, right? At the very least, the amount of time taken to run all 
> threads should go down.

why? If I have 1 CPU and two threads, each of them running t seconds, 
shouldn't the time then be 2*t? Everything over 2*t would be the thread 
handling overhead. since you let 1 Thread run n iteration taking time t 
and then 2 Threads with n/2 iterations... shouldn't the total time then 
still be t?

If you want to tell me that running with 3 threads takes longer than 
running with one thread and this should not happen, then yes, probably, 
but that is not what I understand from scalability. I do understand that 
if you go to big n and the total time increases much, then the 
scalability is bad. If you look at 7 or 20 Threads, then this does not 
happen. Only when the number of threads is lower this happens. So you 
could say the JVM in 1.6 does not scale well for a lower number of 
Threads, but does for a high number of threads...

> Run the same test in Java 5 and you'll see that Java 6 does not scale 
> properly at all for this test.

I will later

bye blackdrag

-- 
Jochen "blackdrag" Theodorou
The Groovy Project Tech Lead (http://groovy.codehaus.org)
http://blackdragsview.blogspot.com/
http://www.g2one.com/

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "JVM 
Languages" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/jvm-languages?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to