1) Message boards : News : New Nbody Version 1.50 (Message 63566) Posted 14 May 2015 by HireMe.geek.nz Post: Both the MT and single CPU versions get stuck at 99.999% and never finish on x86 Windows systems. You need to release a bug fix version soon, as the current situation results in work being done but no CPU credit for it happening as you have to abort the improperly terminating work units.

2) Message boards : Number crunching : Add "trickles" to all applications as work units taking longer as time goes on... (Message 57926) Posted 15 Apr 2013 by HireMe.geek.nz Post: Add "trickles" to all applications as work units are taking longer as time goes on... My view is that the trickles should be at 25% 50% 75% 90% But, if it is known that the work unit is going to take a while a 10% or 15% interval should be available. Sending only the CPU time since the last trickle would be fine. At some later date adding some program internal state debugging information to the trickle up message would be a plus. The "Trickle API" is at -- https://boinc.berkeley.edu/trac/wiki/TrickleMessages -- https://boinc.berkeley.edu/trac/wiki/TrickleApi

3) Message boards : Number crunching : Questionable Cruncher Discussion Thread (Message 46488) Posted 7 Mar 2011 by HireMe.geek.nz Post: Your current settings Allow up to ~10000 work units per day! This is taken from the linked page from the post above this post : "Maximum daily WU quota per CPU : 9948/day" !!!!!!!!!!!!!!!!!!!!! Tweak it down to 100 -- for everyone. !!!!!!!!!!!!!!!!!!!!! That will probably be a better fix, and the dodgy computers will drop out... No fuss, no muss. This will not be the first project to do this.

4) Message boards : Application Code Discussion : Wierd compatabiliy issues beyond SSSE3 ... (Message 46487) Posted 7 Mar 2011 by HireMe.geek.nz Post: Via http://en.wikipedia.org/wiki/SSE5 ... Compatibility AMD's SSE5 extension bundle does not include the full set of Intel's SSE4 instructions making it a competitor to SSE4 rather than a successor. This ambiguity complicates software development. It is recommended practice for a program to test for the presence of instruction set extensions by means of the CPUID instruction before entering a code path which depends upon those instructions to function correctly. For maximum portability, an optimized application will require three code paths: a base code path for compatibility with older processors (from either vendor), a separately optimized Intel code path exploiting SSE4 or AVX, and a separately optimized AMD code path exploiting SSE5. Due to this proliferation, benchmarks between Intel and AMD processors increasingly reflect the cleverness or implementation quality of the divergent code paths rather than the strength of the underlying platform.

5) Message boards : Application Code Discussion : {SSE ... SSSE3 (etc...)} optimization vs {MMX ... SSE2} (Message 46486) Posted 7 Mar 2011 by HireMe.geek.nz Post: FPU / SSE / SSE2 / SSE3 / SSSE3 (etc) optimization vs FPU / MMX / SSE As a screensaver is coming, and the core science application will have to be optimized (at least at the command line interface of the compiler) for SSE and beyond yet again. The SSE optimization so far has paid off, but more experimentation and tries at SSE3 and SSSE3 should be tried. I don't know about user uptake of SSE4 or SSE5 at this time. As for the other more advanced instructions, one will have to wait. I would almost expect Intel to totally dump MMX in the future -- there is not an infinite amount of room for all the logic gates.

6) Message boards : Number crunching : Straw poll : What x86 extra instructions can your CPU do? (Message 46485) Posted 7 Mar 2011 by HireMe.geek.nz Post: Straw poll : What x86 extra instructions can your CPU do? Myself, I am not in love with CISC or the kludge the x86 has become but the vast majority of CPUs used by people are x86. BOINC availability outside of x86 is limited, so that does not help. So for clarity : http://en.wikipedia.org/wiki/X86 has these extra instructions http://en.wikipedia.org/wiki/Template:Multimedia_extensions that can be found by http://en.wikipedia.org/wiki/CPUID My PC has the following ... Number of cores 2 (max 2) Number of threads 2 (max 2) Name Intel Pentium E2140 Codename Conroe Specification Intel(R) Pentium(R) Dual CPU E2140 @ 1.60GHz Package (platform ID) Socket 775 LGA (0x0) CPUID 6.F.D Extended CPUID 6.F Core Stepping M0 Technology 65 nm Core Speed 1600.2 MHz Multiplier x FSB 8.0 x 200.0 MHz Rated Bus speed 800.1 MHz Stock frequency 1600 MHz Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, EM64T L1 Data cache 2 x 32 KBytes, 8-way set associative, 64-byte line size L1 Instruction cache 2 x 32 KBytes, 8-way set associative, 64-byte line size L2 cache 1024 KBytes, 4-way set associative, 64-byte line size FID/VID Control yes FID range 6.0x - 8.0x Max VID 1.238 V

7) Message boards : Number crunching : Will there be GPU (CUDA) support for the screensaver? (Message 46291) Posted 19 Feb 2011 by HireMe.geek.nz Post: Will there be GPU (CUDA) support for the screensaver? I hate to suggest having split versions of the screensaver, but CUDA support can be done -- as far as I know -- using one codebase that can adapt to there being a GPU or not.

8) Message boards : MilkyWay@home Science : Lopsided Milky Way, hint Solar System not from Milky Way originally... (Message 43720) Posted 11 Nov 2010 by HireMe.geek.nz Post: It is my understanding from Classical Mechanics and Relativistic Mechanics that gravity wells or centres (like those created by the galactic core) create a very precisely aligned sheet of stars and solar systems that orbit in sync with the galactic plane. Yes, the Earth has a 23 degree tilt and that has to be filtered out with respect to all observations (where galactic plane alignment is critical). Yes, orbital resonances with the solar system are an important part of all research into solar system origins. Yes, you need double precision maths to analyse all of this. Yet, most astronomy (visible or radio) is in the Southern Hemisphere. This placement effect is the sticky point that must be explained somehow with some certainty. However, proof of the solar system osselating up and down the galactic plane is not easy to obtain due to the very long timeframes involved -- even if statistical mechanics can show the ossalatory effect with great certainty. One galaxy cannibalising another galaxy is a systemic process that is measured in Gegayears -- and the mechanics of such are still very new to scientific research. The cannibalization of the Solar System could have taken place as far back as 65 millions to 80 millions of years ago -- long enough to have substantial quenching of redshift and other orbital data. The Milky Way has a more powerful gravity centre, so it can quench better than dwarf galaxies. The stellar "metals" composition argument that is to distinguish star populations from each other (as a way to prove or disprove external origins) is interesting but can't be totally conclusive. The metals composition argument is subject to bell curve star formation probability problems. I would like to see a secondary or tertiary arguments using other physics to show more conclusively that the Sagittarius Solar System origin is bunk. If this origin idea can be disproven by 3 separate and different scientific methods that would be optimal. My view is that a second argument might be found in the stability or instability of the Oort Cloud. -- http://en.wikipedia.org/wiki/Oort_cloud Tertiary arguments are welcome...

9) Message boards : Application Code Discussion : Application "trickles" would be nice ... (25%, 50%, 75%) (Message 43605) Posted 9 Nov 2010 by HireMe.geek.nz Post: Application "trickles" would be nice ... (25%, 50%, 75%) -- but only if application runs for more than 3 hours : Myself I only wish the trickles to return : CPU seconds GPU seconds Diagnostic program data (I have not looked at the source code so I will not put anything down here) ... but probably only 16 parameters or less. Format, for futureproofing ... XML

10) Message boards : Application Code Discussion : SSE2 application behaviour (Win32, Vista, 32/64 CPU) ~120% better is nominal (Message 43604) Posted 9 Nov 2010 by HireMe.geek.nz Post: My CPU is this : OS Name Microsoft Windows Vista Home Premium Version 6.0.6002 Service Pack 2 Build 6002 System Model SR5 250 NX System Type X86-based PC Processor Intel(R) Pentium(R) Dual CPU E2140 @ 1.60GHz, 1600 Mhz, 2 Core(s), 2 Logical Processor(s) BIOS Version/Date American Megatrends Inc. 5.14, 8/10/2007 SMBIOS Version 2.4 +++ VIA INTEL CPU IDENT APP +++ Intel(R) Processor Identification Utility Version: 4.23.20100812 Operating System: 6.0-6002-Service Pack 2 Number of processors in system: 1 Current processor: #1 Active cores per processor: 2 Processor Name: Intel(R) Pentium(R) CPU E2140 @ 1.60GHz Type: 0 Family: 6 Model: F Stepping: D Revision: A3 Maximum CPUID Level: A L1 Instruction Cache: 2 x 32 KB L1 Data Cache: 2 x 32 KB L2 Cache: 1 MB Enhanced Intel SpeedStep(R) Technology: Yes MMX(TM): Yes Intel(R) SSE: Yes Intel(R) SSE2: Yes Intel(R) SSE3: Yes Intel(R) SSE4: No Enhanced Halt State: Yes Execute Disable Bit: Yes Intel(R) Hyper-Threading Technology: No Intel(R) 64 Architecture: Yes Intel(R) Virtualization Technology: No Expected Processor Frequency: 1.60 GHz Reported Processor Frequency: 1.60 GHz Expected System Bus Frequency: 800 MHz Reported System Bus Frequency: 800 MHz ++++++++++++ Outcome ++++++++++++ Your SSE2 application is running about 120% better than the more bare bones application that was distributed. Tick rate % (old application) : 0.001 per second (mostly invariant) Tick rate % (SSE2) : ~0.002 per second, (-/+) 10% variance

11) Message boards : MilkyWay@home Science : Lopsided Milky Way, hint Solar System not from Milky Way originally... (Message 43603) Posted 9 Nov 2010 by HireMe.geek.nz Post: The lopsided Milky Way with respect to the Solar System: From what I understand it is a hint that the Solar System not from Milky Way originally ... possibly the entire system originated in the calmer Sagittarius Galaxy. Solar systems originating in the Milky Way, if celestial mechanics mean anything should not be lopsided at all. Where exactly can one get ahold of research relating to this?

©2024 Astroinformatics Group