Most recent change on $Date: 2011/03/23 21:45:39 $ (UT) [Y2K test]
(Xeon 2 dual 64-bit processor, Speed 3.0 GHz, Memory 4 Gb, Cache 4096 Kb, RAID0 file system;
31DEC05 compiled with g77 2.95 dynamic link on RedHat Enterprise at CfA)
(Xeon 2 dual 64-bit processor, Speed 3.0 GHz, Memory 2 Gb, Cache 4096 Kb, RAID0 file system;
31DEC06 compiled with Intel ifort on RedHat Enterprise at CfA)
NOTE: a new benchmark test, called Y2K, has been developed. It is an adjustment of the existing DDT with an altered set of tasks (
IMAGR), and a 10-fold change in the size of the problems. Its goal is to take on the order of an hour to run on fairly new desktop hardware. However, it now takes 3:43 on a 3.06 GHz dual-cpu Intel Xeon with the Intel binary installation. It takes about 2:05 on the new record holder.
The Y2K and DDT (Dirty Dozen) suites are reliable ways of determining two things:
The new AIPS Memo 104
describes the new test and gives initial performance results. It was
developed on the
31DEC00 version of AIPS. See the 31DEC05 and 31DEC06
pages for information on obtaining and maintaining your very own copy
of the current releases. For more details on the earlier DDT, see AIPS
Memo 85 which has the information on the last iteration of the DDT
test. This was modified for the
15JAN94 release of AIPS
(this is not the most recent version; see the AIPS Home Page for info on recent releases). There are
other documents available too.
The AIPSMark00 is defined as:
Am(00) = (40000 ÷ TY2K(L))
where TY2K(L) is the total run time in seconds
of the Y2K LARGE test. This run time was defined initially by
Glendenning and Hunt as the time in seconds from
RUN DDTEXEC through
PRINTING ANSWERS, ERRORS, OTHER IMPORTANT
MESSAGES). The multiplier 40000 was chosen to make the
new AIPSMark00 about the same as the old
AIPSMark93 on many machines.
The AIPSMark93 is defined as:
Am(93) = (4000 ÷ TDDT(L))
where TDDT(L) is the total run time in seconds for the LARGE DDT suite as defined by Glendenning and Hunt. Chris Flatters devised a formula for predicting a computer's AIPSMarks based on its measured SPECfp(95) and SPECint(95) performance. An extensive list of SPEC results is maintained on the Web by John DiMarco at the University of Toronto. Chris finds that the following relation gives a fair, ball-park estimate of the AIPSmark that can be expected on a machine with known SPEC(95) benchmarks.
AIPSMark93 = 0.898 × SPECfp(95) + 0.11 × SPECint(95) - 1.665
Install the latest version of AIPS; see for example
Start running AIPS and enter
EXPLAIN Y2K and
EXPLAIN DDT. These are available on the web for the
DDT.HLP. The process
consists of several steps:
Don't panic; 31DEC10 versions of the Y2K and DDT data files are
available (remember to right click for Save as or Save link as) via anonymous
ftp. The older Y2K data files are available (remember to right
click for Save as or Save link as) via anonymous
ftp and the older DDT data files are also available via anonymous
ftp. But please look at the file sizes (use the
dir command in your ftp program) before
pulling these across. Some of them are large: the aggregate size of
all 33 DDT files is about 65 Megabytes while the Y2K files add up to
217 Megabytes. The
needed to compile and execute the tests are provided with the
31DEC00 and later versions of AIPS. Note that DDT does
not run on modern version of AIPS since MX and other tasks involved
have been removed from the system.
The results published in these pages do NOT constitute any sort of an endorsement by NRAO or Associated Universities, Inc. (NRAO's parent organization) of any particular vendor. The AIPS web pages at NRAO/CV do not and will not contain links to any hardware vendors' web pages.
In addition, many of the results reported here were relayed to us by others, and while all of these were carried out in a uniform way to the best of our knowledge, there will inevitably be variations introduced by factors such as changes in the AIPS code over time, system loads, different disk controllers (e.g. IDE vs. SCSI), and so on. The intent here is to share AIPS users' experiences with various platforms among the Astronomical Community.
As reported above, the Y2K LARGE test can now run in as little as 5.0 minutes on some computers. Therefore, we have put together a HUGE version of the Y2K test. It uses Carilli/Perley Cynus-A X-Band data from D, C, B, and 4 frequencies of A configurations of the VLA. The 2674490 visbilities are Cleaned in four resolutions by IMAGR before the self-cal (replacing APCLN since it does not do multiple resolutions). The self-cal is amplitude and phase with the square root of the normal weights. The following Clean is also four resolutions done with IMAGR. The VTESS (and the one in LARGE) use the trick of setting FLUX = -F/10 where F is about the correct flux. With this trick VTESS is quite convergent rather than being completely erratic.
The Mac OS/X 2 GHz dual-headed machine listed at the top of this page does the HUGE test in 3519 seconds (2753 seconds of cpu). My Dell 1.3-GHz Linux box takes 9090 seconds (6416 seconds of cpu). The UV data sets and images for HUGE are available along with the other 31DEC04 data sets.
The test results from AIPS Memo 104 are summarized in a separate table.
Machine Machine Compiler DDTL Y2KL Y2KL cpu AMark run on compiled on type sec sec min 2000 ------- ----------- -------- ------- -------- -------- ----- 1 4 4 . A GNU 121-124 1126 16.28 35 2. 4 4 GNU 115-120 1060 15.14 38 3. 4 4 (b) IBM 81- 83 748 9.58 53 5. 4 5 (gen) IBM 77- 79 748 9.55 53 7. 4 4 (gen) IBM 78- 83 759 9.57 53 11. 5 5 A GNU 69- 73 582 9.29 69 12. 5 5 GNU 68- 70 558 7.77 72 13. 5 5 (a) IBM 33- 36 310 4.60 129 13. 5 5 (b) IBM 32- 38 305 4.42 131 15. 5 5 (gen) IBM 33- 34 300 4.46 133 16. 5 5 (gen -O5) IBM 34- 39 313-322 4.66-4.82 >124 The G5 used is hakone a 2.0 GHz, dual-headed machine with 8 Gbyte ram, 3 Terabytes disk! The G4 used is pippin a 1.33 GHz, single-headed, rack-mounted machine with 1 Gbyte of ram. Putting aips on a "disk image" slows things down (e.g. 47 sec DDT on the G5 with the IBM compiler, 90 sec on G4) IBM options (a) are -qthreaded -qunroll=auto -qtune=auto -qarch=auto (b) are -qtune=auto -qarch=auto all including (gen) use -O3. The A GNU is the Apple-provided version of the GNU compilers while GNU is one installed by Wes (~ 3.4.1).
This table is an attempt to show the best results for the most advanced hardware on which we or others have been able to run the DDT and AIPS. It is incomplete, but will be filled in as time permits.
|Hardware||Operating System||Compiler||AIPS Version||Am(93)||Flatters' prediction||Notes|
|Intel Pentium III Xeon 1.7GHz||Red Hat Linux 7.1||gcc 2.95.3||
||85.0||?||1 Gbyte memory, dual processor, two 10,000-RPM 70G SCSI disks.|
|Intel Pentium 4 1.5 GHz||Linux 2.2 (RH7.0)||g77/gcc 2.95||
||69.0||?||Single processor, 384 MBytes memory, 3x 15,000-RPM 18GB drives
(two striped as
|Alpha Digital/Compaq XP1000||Red Hat Linux 6.2||DEC Fortran||
||58.8||?||512 MB memory, 4MB Cache, 18.2G SCSI disk|
|Alpha Digital/Compaq DS-20||OSF/1 4.0D||DEC Fortran||
||33.61||?||2 Gbytes memory, RAID disks, guest Compaq/CSA system.|
|SGI Origin 2000||Irix 6.4||SGI||
||30.3||16.4/cpu?||4 300-MHz R12K CPUs, speed parameter set to 30.|
|AMD Athlon 600MHz||Linux 2.2.13||gcc 2.95.2||
||27.2||?||SuSE 6.3, 512 MBytes memory, 2 GBytes swap, 2 EIDE
|Intel Pentium III 800||Linux 2.2 (RH6.1)||g77-egcs 2.91.66||
||25.6||?||Single processor, 128 MBytes memory,
|HP C360 (PA-8500 chip)||HP-UX 10.20||HP f77 (not f90)||
||24.6||?||512 Mbytes main memory, Ultra-SCSI and SCSI Fast/Wide disks.|
|Intel Pentium III 550 Xeon||Linux 2.2 (RH6.0)||g77/egcs 1.1||
||21.6||?||Single processor, 768 MBytes memory, 4 x U-2 SCSI 10,000 RPM disks|
|Sun Sparc Ultra 60||SunOS 5.6||Sunsoft||
||18.4||?||13.9 without /tmp trick
speed parameter 13.5
|Sun Sparc Ultra 30/300||SunOS 5||Sunsoft||
||15.0||16.1||10.0 without /tmp trick|
|Aspen Systems Alpha 21164/433 MHz||Linux 2.1.57||DEC Fortran||
||9.0||16.1||Binaries built statically under OSF/1 4.0, run in emulation mode.|
|IBM RS/6000 SP2||AIX 4||xlf/cc||
||4.7||7.1||machine not empty, compiler options non-optimal|
A full table of results is available along with the initial 1993 results.
Pointers to sections ordered by manufacturer.