[LLVMdev] LLVM benchmarks against GCC

["?=Valery A.Khamenya=?koi8-r?Q?"]

Hi all, 

i was thinking that this question was not good
right after relese 1.0, but now perhaps it is OK...
if not, then I am sorry. 

So, what about current status of benchmarks?

I mean comparison to gcc.

I have looked at 
http://llvm.cs.uiuc.edu/testresults/X86/

Unfortunatelly graphs lines are hardly for human eye, 
but tables are OK.

I give my own interpretation for April 26, 2004 benchmrking 
tests, please don't beat me if i am wrong, but correct me.
Focus of my attention is the time of execution of the program,
i.e., fields GCC/CBE and GCC/LLC.


For me it looks like following.

----------------------------------------------------
1. Programs/External:

 a) CBE code is already comparable with GCC code 
    (some tests are slower, but some quicker.)

 b) LLC code is still rather slower then GCC code


2. Programs/MultiSource

 a) CBE code is already rather quicker then GCC code

    some tests are still (moderate) slower, 
    but some are much quicker (up to 5 times).

 b) LLC code is still rather slower then GCC code.

    However some tests show up to 5 times speed up


3. Programs/SingleSource

 a) CBE code is rather quicker then GCC code 

    some tests are still (moderate) slower, 
    but some are much quicker (up to 6 times).

 b) LLC code is comparable with GCC code or even rather quicker .

    some tests are still slower (up to 5 times), 
    but some are much quicker (up to 5 times).
    But tendence rather to "LLC code is quicker"
----------------------------------------------------

Overall impression: 
1) CBE code is already rather quicker then GCC
2) LLC code is rather slower then CBE, but comparable to GCC

BTW, guys, why not to focus more attention on slow
tests like SPEC/CFP2000/179.art/179.art or 
UnitTests/2002-10-09-ArrayResolution? 

or maybe it is already under hard work? :)

---
Valery A.Khamenya