Re: Poor performance with OpenMP



On 2011-01-16, nmm1@xxxxxxxxx <nmm1@xxxxxxxxx> wrote:
However, that wasn't the actual issue being considered in the thread.
It was the benefit of using hyperthreading. If a program that would
run entirely in L1 cache and use no communication gets only a factor
of 1.2, your code would be very unlikely to benefit at all.

My understanding was that the "optimal" use case for HT would be
programs that do lots of independent sequential (pseudo)-random memory
access (something like graph traversal, for instance), such that
execution is bound by memory latency. Note that I said latency, if
memory BW is the bottleneck the situation is, again, different, and HT
will obviously not help if a single thread is able to saturate the
memory BW.

OTOH, if the programs are cache-friendly enough that they essentially
run in L1, then a single thread ought to be able to keep the execution
resources quite busy and HT would be of little benefit.


--
JB
.



Relevant Pages

  • Re: [Full-disclosure] [Dailydave] What RedHat doesnt want you to know about ExecShield (without
    ... buffer overflow attacks by performing executable memory checks. ... This is not the case with ExecShield without NX. ... code execution, in the other you do not. ...
    (Full-Disclosure)
  • [NT] Defeating Microsoft Windows XP SP2 Heap Protection and DEP Bypass
    ... The following security advisory is sent to the securiteam mailing list, and can be found at the SecuriTeam web site: http://www.securiteam.com ... and bypassing DEP (Data Execution Prevention). ... Buffer overrun attacks are among the most common mechanisms, or vectors, ... a long string to an input stream or control longer than the memory ...
    (Securiteam)
  • Re: [SLE] Threaded Perl
    ... > Pentium-III Hardware because if the problem is essentially memory bound ... I'm not familiar with the breakdown of execution units and how they relate ... that patterns of primary storage access (especially overall L2 and L3 ... execution patterns, access to RAM is the limiting factor. ...
    (SuSE)
  • Re: Possible evidence of performance regression for 8.1-S (vs. 7.1)
    ... But cpu and memory driven tests all seem to be about the same and the disk io differences are pretty small. ... Test execution summary: ...
    (freebsd-performance)
  • Re: Can you write code directly in CIL ???
    ... I don't care if a GC occurs during the execution of my code. ... always needs all of this memory the whole time that it is executing. ... and the CLR isn't going to care what your function is doing. ... >>> You don't understand a fundamental concept to .NET and CIL. ...
    (microsoft.public.dotnet.languages.csharp)