Fastcode MulInt64 B&V 0.3
Next message: David Knaack: "Re: Fastcode Logo Challenge"
Date: Tue, 21 Oct 2003 20:35:46 +0200
Released at
http://home3.gvdnet.dk/~marianndkc/index.htm
I need some guidance from you guys. How should we make the benchmark? It
looks like we have to inline all the functions in their own copy of the
benchmark function to get rid of the store to load stall? Which we need to
prove is there.
Regards
Dennis
Next message: David Knaack: "Re: Fastcode Logo Challenge"
Relevant Pages
- Re: Lisp vs. Java vs. C++ speed comparison time? [LONG]
... >>I don't see much that can be done with these, but you could inline ... > Depends on what sort of changes we're allowed to make to the ... rewriting REWRITE by moving the mutally ... I really think we should move on to a different benchmark. ...
(comp.lang.lisp) - Re: [v2.6.26] whats brewing in x86.git for v2.6.26
... Faster in a 100% bogus benchmark with the most unrealistic input data ... The input for the first 'benchmark' was indeed completely unrealistic. ... the speed of the generic bitmap implementation ... inline an optimized version, ...
(Linux-Kernel) - Re: Fastcode Pos B&V 5.5.1
... functions via function pointers and these will never be inlined, ... But it is problematic that we do not benchmark the true performanceof this ... Any inline function called via a function pointer will not be inlined. ... {$IFDEF CONDITIONALEXPRESSIONS} ...
(borland.public.delphi.language.basm) - Re: Compiling Regexp only once
... multiple line parsing) and that's where you can put the inline regexp ... I'm not sure what you mean by inline, do you mean a regexp literal like ... require 'benchmark' ... b.report "const match" do ...
(comp.lang.ruby) - Re: Fastcode MM B&V 0.48
... Dennis wrote: ... The last benchmark that reads "Usage Replay Benchmark" is unchecked, ... and even when it's checked it looks for the replay file MMUsage.log ... Regards, ...
(borland.public.delphi.language.basm) |
|
