schreiber@schreiber.asd.sgi.com (Olivier Schreiber) (12/07/90)
>From schreiber Wed Dec 5 17:00:15 1990 >Received: by schreiber.asd.sgi.com (5.52/900721.SGI) > for schreiber id AA06130; Wed, 5 Dec 90 17:00:15 PST >Date: Wed, 5 Dec 90 17:00:15 PST >From: schreiber (Olivier Schreiber) >Message-Id: <9012060100.AA06130@schreiber.asd.sgi.com> >To: dongarra@cs.utk.edu, schreiber >Subject: Your policy on linpack 100*100. Thanks. >Status: RO >What is the policy about linpack 100x100? >I think there is some disagreement within the benchmarking community, >witness this exchange on the comp.benchmarks newsgroup: >>>>>> On 26 Nov 90 19:13:42 GMT, schreiber@schreiber.asd.sgi.com (Olivier Schreiber) said: >Olivier> In <MCCALPIN.90Nov22090957@pereland.cms.udel.edu> mccalpin@perelandra.cms.udel.edu (John D. McCalpin) writes: >>[...] It is only "fairly" exact since the vendor is >>allowed to re-write the BLAS-1 (Basic Linear Algebra Subroutines) >>however s/he wants in order to improve performance. [....] >Olivier> Reading the Dongarra CS-89-85 report, my understanding is that >Olivier> for the LINPACK 100x100 test, one is not allowed to modify the >Olivier> BLAS-1 Basic Linear Algebra Subroutines either as they are >Olivier> contained in the source code obtained from netlib. >Dongarra certainly seemed to have intended that the vendors produce >optimized BLAS for their machines. The only "requirement" seemed to >me to be that the source/type of the BLAS be disclosed. >In the early days, the report contained separate sections for the "All >Fortran" cases and the "Coded BLAS" cases. I do not believe that this >distinction is maintained in the current (and much abbreviated) >report. >-- >John D. McCalpin mccalpin@perelandra.cms.udel.edu >Assistant Professor mccalpin@brahms.udel.edu >College of Marine Studies, U. Del. J.MCCALPIN/OMNET >-- >Olivier Schreiber schreiber@asd.sgi.com Tel(415)335 7353 Technical Marketing MS 7L580 Silicon Graphics Inc., 2011 North Shoreline Blvd. Mountain View, Ca 94039-7311 From @dasher.cs.utk.edu:dongarra@cs.utk.edu Wed Dec 5 19:31:42 1990 Received: from giraffe.asd.sgi.com by schreiber.asd.sgi.com (5.52/900721.SGI) for schreiber id AA09264; Wed, 5 Dec 90 19:31:42 PST Received: from sgi.sgi.com by giraffe.asd.sgi.com via SMTP (5.64-bind 1.5+ida/900721.SGI) for schreiber@schreiber.asd.sgi.com id AA25981; Wed, 5 Dec 90 19:31:40 -0800 Received: from DASHER.CS.UTK.EDU by SGI.COM via SMTP (5.64-bind 1.5+ida/900410.SGI) for schreiber@schreiber.asd.sgi.com id AA04677; Wed, 5 Dec 90 19:31:37 -0800 Received: by dasher.cs.utk.edu (5.61++/2.6c-UTK) id AA20764; Wed, 5 Dec 90 22:33:11 -0500 Date: Wed, 5 Dec 90 22:33:11 -0500 From: dongarra@cs.utk.edu Message-Id: <9012060333.AA20764@dasher.cs.utk.edu> To: schreiber Subject: Re: Your policy on linpack 100*100. Thanks. Status: R Olivier is correct. The Fortran source code for SGEFA, SGESL and the BLAS are not to be changed for the 100x100 Linpack benchmark. The code should be run as distributed from netlib@ornl.gov. No changes are allowed, no ``finger prints'' on the code, not even changes to the comments. I hope that is clear. In the old days I had a table with ``coded Blas'', but that got too confusing. Today its all Fortran. Jack -- Olivier Schreiber schreiber@asd.sgi.com Tel(415)335 7353 Technical Marketing MS 7L580 Silicon Graphics Inc., 2011 North Shoreline Blvd. Mountain View, Ca 94039-7311