Posts by FalconFly
|
61)
Message boards :
Number crunching :
A journey: iGPU slowing CPU processing
(Message 1547345)
Posted 25 Jul 2014 by  FalconFly
Post: Ah, okay... Running a similar setup (albeit with a rather slow mobile Core i3) myself since just a few days, I also found out the performance ratio on SETI comparing CPU and iGPUs is unexpectedly close. So in this case, you're right as the iGPU doesn't contribute so much to the overall score when bundled with fast CPU cores. |
|
62)
Message boards :
Number crunching :
SETI@Home Wow!-Event 2014
(Message 1546696)
Posted 24 Jul 2014 by FalconFly
Post: Count me in... long time no crunching here ;) |
|
63)
Message boards :
Number crunching :
A journey: iGPU slowing CPU processing
(Message 1546694)
Posted 24 Jul 2014 by FalconFly
Post: I wasn't referring CPU performance. The impact of CPU performance when the iGPU is already bottlenecking the RAM bandwidth is not very important. The overall performance loss caused by the iGPU not getting max. RAM bandwidth is the key. The CPU cores anyway hardly contribute much to the total output, unless it is a very potent CPU coupled with a rather slow iGPU (although that can depend on the project). The potential output gain from running CPU cores is typically more than offset but the significantly reduced iGPU performance under these conditions. |
|
64)
Message boards :
Number crunching :
A journey: iGPU slowing CPU processing
(Message 1546022)
Posted 22 Jul 2014 by FalconFly
Post: As far as I have observed, iGPUs causing massive slowdown of other CPU tasks is simply caused by it downright crippling the RAM bandwidth. iGPUs that use System RAM easily peak-load the entire RAM bandwidth all on their own. Adding CPU tasks that both occupy cache and then also compete for the little RAM bandwidth just slows down everything. iGPUs are usually better left running all on their own (no CPU tasks at all, except the one keeping the GPU loaded). That typically yields the best overall performance - albeit it does leave some CPU cores unused. Although that feels like a waste, shared memory GPUs (i.e. AMD APUs or CPU-integrated intel GPUs) force the user to make that decision or reserve the CPUs for another non-CPU intensive BOINC project. On the positive side, having the CPU cores run idle reduces power consumption while the iGPU yields optimum total performance in terms of output. Also, once the iGPU is on its own, its output performance on the project will also maximize. The only exception to that rule would be tasks that fit nearly entirely into the CPU caches and do not compete with RAM bandwidth to any significant extent. Quick example : Dual-Channel DDR3-1866 yields about some 14-16GB/sec realistic max. sustained performance under typical conditions and a standard system overhead (*). That bandwidth is easily consumed completely by modern iGPUs all on its own. Adding any other RAM-Intensive tasks will divide that figure by the total amount of cores competing for its bandwidth. In the end, all of them would operate under extremely RAM bandwidth-starved conditions, yielding comparably poor performance on both CPU and iGPU runtimes. (*) Note that Dual Channel does not double the RAM bandwidth in any way, although marketing never stopped suggesting that. It's maximum sustained advantage over single channel is approx. 15-20%. This won't change until someone develops and integrates a fully functional RAID 0 RAM-controller (RAM striping), which doesn't exist so far. PS. I've just recently put my notebook to use for SETI. As it both holds an intel HD4000 as well as an NVidia GT720M 1GB, I've set the Core i3-3110M 2.4GHz GPU to run only the two GPU tasks. The HD4000 is given the max. available RAM bandwidth of dual channel DDR3-1600, while the GT720M is working off its own VRAM. The four logical CPU cores (2x phys + 2 HT) keep those loaded to the optmium extent while keeping their resulting RAM bandwidth competition to a minimum, and having the Caches all for themselfes to feed the GPUs. The remaining CPU idle time is reducing the thermal stress on the notebook's cooling system, which is highly desirable. |
|
65)
Message boards :
Number crunching :
Emergency fund drive for the project.............
(Message 1056074)
Posted 14 Dec 2010 by FalconFly
Post: Duty calls once again I see... ;) Thank you for your gift of $400.00 via Mastercard on 12/14/2010. |
|
66)
Message boards :
Number crunching :
Noobie CUDA GPU temperature worry
(Message 946185)
Posted 9 Nov 2009 by FalconFly
Post: Small add-on : Seems some stock GPU fans don't like 24/7 ops as well. The fan on my Club3D HD4890 "Superclocked Edition" now begins to make rattling noises in the speed regime it is now usually in (~50% rpm). That's just a mere 2 weeks into GPU crunching with the otherwise fairly new card (<3 months), due to its rather frequent speed changes to the fan (it has a rather sensitive fan control logic that will quickly alternate speed levels depending on GPU load). ...another example how 24/7 ops can easily hurt the Video card, in my case caused by a ~1.50$ stock cooler part - the fan which obviously isn't upto that job due to cheap design. I'll have to look into an alternative GPU cooler, otherwise I see that Video Card going dead in less than 14 days, caused by the weakest part in the chain. |
|
67)
Message boards :
Number crunching :
Noobie CUDA GPU temperature worry
(Message 942794)
Posted 25 Oct 2009 by FalconFly
Post: OK... So... Not at all unfortunately. What 24/7 ops does to the hardware is mostly invisible to the typical owner not knowing or honoring the hardware limitations : the decreased lifespan of its components. Especially the capacitors have a pretty well defined lifespan, which almost exclusively depends on running hours and are very sensitive to operating temperature. As these cards are designed to operate a limited amount of time per day, its equivalent expected lifespan will normally be several years - even when used under merely average conditions. Put under 24/7 load, however, this lifespan is quickly reduced to a fraction of the intended period - further decreased by high operating temperatures. You'll see the same when going to harddrives, there are only few normal desktop drives that carry a 24/7 specification. For server drives, this specification is standard and the drives were built for that purpose. 24/7 vs. "normal use" may sound just like marketing but has important design differences based on the different specifications. You'll find this forum has plenty of reports of dead video cards. GPU failure is seldom the cause, literally burned VRMs or blown capacitors usually are. Therefor my advice : - there's far more to running a video card 24/7 than just keeping GPU temps in mind - VRMs and especially capacitors will need significantly increased cooling to preserve their lifespan (the reason many modern cards have dedicated, additional temperature sensors i.e. for VRMs and VRAM - as their temps are just as critical as GPU temps, the Fan Control logic has to be able to react to these as well) You can choose to ignore these requirements and may get lucky. Else you will destroy your xxx$ Video card due to lack i.e. of 10$ additional cooling or a plain inadequate case. |
|
68)
Message boards :
Number crunching :
Noobie CUDA GPU temperature worry
(Message 940746)
Posted 17 Oct 2009 by FalconFly
Post: Just be aware that conventional Video cards are neither designed not specced for 24/7 data crunching and its ensuing thermal stress. For a few, irregular hours of gaming each day, such temperatures may be okay and will result in a normal life span of the device. But as a permanent load, critical parts like VRM modules, condensators and to some extend the VRAM will experience a reduced life span. With these going defective, a still intact GPU (which can accept such temperatures for long periods of time) becomes worthless. The formentioned components are the weakest links in that chain. For anything that goes like 24/7 GPU crunching, one should have clearly above-average case cooling, specifically a cooling that directly benefits the video card. Stuffed into generic 0815-cases or given inadequate cooling for the job, those modern cards will run for a variable while, then die. |
|
69)
Message boards :
Number crunching :
Pondering Stealth Results
(Message 939129)
Posted 11 Oct 2009 by FalconFly
Post: I am missing such functions since the days of SETIwatch (logging top Gaussian/Pulse/Triplet for every single WorkUnit into local database). By now I originally had expected that some BOINC-wide logging would exist (either per Project Site or with a dedicated BOINC tool). But apart from isolated and project-specialized tools requiring manual work (CPDN) or Projects like Rosetta (detailled results summary) usually nothing is shown concerning results. Maybe we see something like that when NTPCKR is fully operational but I'm not too hopeful because of other priorities. |
|
70)
Message boards :
Number crunching :
Welcome to the 10 year club
(Message 938138)
Posted 7 Oct 2009 by FalconFly
Post: Almost forgot it... 2 days ago I turned 10. Actually unbelievable how fast the time runs. Strangely I still remember that 5 Oct 1999 23:08:23 UTC (00:08:23 Local time) it was a cool and dark night, lights were toned down already in my room as I originally prepared to go to bed. I couldn't rest, however, because I previously already had trouble signing up onto that "SETI Screenaver" and I didn't intend to give up that easily this day. Moments later (after another attempt to create the account), I noticed the Program had downloaded data and started to work. I looked at that Screensaver until 2am in the night before actually going to bed. Back in these days, my AMD Athlon 500 (pretty fast machine for that time) easily took some 13-14 hours to complete a WorkUnit. Being hooked up to a 56k Modem at an expensive internet rate, every Up- and Download was manual business. It wasn't until a month later that I discovered the much faster ways to compute data first by tweaking the Screensaver and eventually switching to the CLI Client. Two month later, I managed to network my 2nd PC (AMD K6-2 400MHz) and allow it to crunch data as well (a WorkUnit took some 16 hours on that rig vs. ~9 on the Athlon). Equipped with 2 CPU-Cores totalling 900MHz of computing power, I went into near fulltime-crunching these days. Now - 10 years later - I look back and find I'm crunching SETI data for the seti.nl 10-year gauntlet, watching over 24 Systems (sporting 92 CPU Cores) dawing some 2.1kW electrical power which comfortably heats an entire house. Still feels like a very lonely business though and explaining colleages or friends what this permanently humming, spanning over 2 rooms lineup of expensive computers does... still creates the same weird expression on their faces (commonly going along with a rather helpless "uh-huh"). But someone's got to get the job done. And if we don't - who would ? |
|
71)
Message boards :
Number crunching :
Weird can't upload Problem...
(Message 936106)
Posted 26 Sep 2009 by FalconFly
Post: Awesome, that did the trick. I already had this file in the BOINC data directory after installation on the System years ago, but somehow it got lost (?) (would have never thought of that File going AWOL, rather searched everywhere else). Thanks for the quick help :) |
|
72)
Message boards :
Number crunching :
Weird can't upload Problem...
(Message 936087)
Posted 26 Sep 2009 by FalconFly
Post: Got it on only a single of my Hosts (Link) : SETI@home 26.09.2009 19:33:17 Started upload of 26au09ad.14272.14791.14.10.220_1_0 --- 26.09.2009 19:33:31 Project communication failed: attempting access to reference site SETI@home 26.09.2009 19:33:31 Temporarily failed upload of 26au09ad.14272.14791.14.10.148_0_0: HTTP error --- 26.09.2009 19:33:33 Internet access OK - project servers may be temporarily down. I already tried resetting Project, updated BOINC from 6.4.5 to 6.6.40 but it doesn't change anything. Network connection is fully normal, a quick run of POEM resulted in absolutely normal behaviour. I even tried flushing the DNS cache and did several System reboots - all to no avail. All my other Systems running over the same pipe have no problems running and Down/Uploading SETI work. Basically it can download work, process it but can't upload. BOINCview shows 100% Upload complete where it aborts and backs off over and over. Does anyone know such a weird behaviour with SETI only and possible fixes (?) |
|
73)
Message boards :
Number crunching :
- POLL - : SETI@home on ATI GPU..
(Message 936026)
Posted 26 Sep 2009 by FalconFly
Post: I. YES - I have an ATI GPU |
|
74)
Message boards :
Number crunching :
Optimization
(Message 906304)
Posted 11 Jun 2009 by FalconFly
Post: Other than disabling unneeded Tasks and not using unneeded Software running in the Background - not much you could do. Just make sure you don't run CPU-intensive tasks like 3D Screensavers - which includes the BOINC screensaver option - to optimize performance. Viewing your computer Details in the Computer Listing will show you the average CPU efficiency. It should be 99.x% for dedicated systems and in the high 90's for typical systems, unless you're running Games or alike (tasks using 100% CPU power) while BOINC is still running. For max. performance, most people forget the easiest way to achieve it : Let BOINC run at all times and leave the system running 24/7. A slower CPU with 24/7 uptime can easily beat a high-end system running only 1 hour a day. There are a few smaller tweaks (BIOS related, disabling CPU power saving functions or all unneeded Visual Effects within Windows) but those either come with a drawback or require in-depth knowledge. Performance gains usually are present but quite low, for most people it's not worth the effort while potentially risking system stability. PS. I'd advise to stay clear of so-called "optimizer Programs" that claim to make your System faster (your Game Booster shouldn't affect BOINC performance however). Most of these freeware/shareware "optimizers" either do nothing for performance or even cause problems, resulting in actually decreased performance. ---- edit ---- Forgot the most essential thing : It looks like your're not using an optimized CPU Application or modified GPU Application. For the CPU that means you're wasting alot of performance and your GPU will from time to time encounter the so-called VLAR-Bug (a series of workunits that will block the GPU for longer periods of time) Go to the Lunatics Website and download the appropriate Unified Installer. After selecting the proper functions for your CPU + GPU crunching, it will install the latest optimized/modified applications, which should increase your overall performance massively. |
|
75)
Message boards :
Number crunching :
CLOSED** SETI/BOINC Milestones (tm) XVII **CLOSED
(Message 904964)
Posted 7 Jun 2009 by FalconFly
Post: ** Falconfly ** Just revisited the Thread for the 8 Million, that sure went quick :D I managed to get to about M1.3 (130k RAC) with some 800k pending, clocking in 11th place overall in RAC. Not too shabby for a mere few weeks production that didn't even run at full efficiency (had some troubles with the CUDA boxes). Power consumption was indeed quite high, peak usage was some ~4800W of pure crunching power. Since the 10 year anniversary race is over, I'm in the final process of winding down. Now it's time for optimization of power consumption again. PS. Current power bill is 312 Euro or 442 US$ per month :p |
|
76)
Message boards :
Technical News :
Upwards and Onwards (May 28 2009)
(Message 902964)
Posted 2 Jun 2009 by FalconFly
Post: I see from the latest Technical News that intel donated six machines, my best guess is that one will serve as NTPCKR. |
|
77)
Message boards :
Technical News :
Upwards and Onwards (May 28 2009)
(Message 902848)
Posted 2 Jun 2009 by FalconFly
Post: Do I understand correctly that they do have a NTPCKR machine already available ? ... Does "Ahhh..." mean "Yes" ? |
|
78)
Message boards :
Technical News :
Upwards and Onwards (May 28 2009)
(Message 902599)
Posted 1 Jun 2009 by FalconFly
Post: Do I understand correctly that they do have a NTPCKR machine already available ? (I didn't see the mentioned Webcast and it's still posted as required hardware in the Hardware/Case Donation Thread) |
|
79)
Message boards :
Number crunching :
CPU cooling and cleaning - a cautionary tale
(Message 902419)
Posted 1 Jun 2009 by FalconFly
Post: How about showing the smoker the heatsink and fan? Politely explain that that same stuff is in their lungs and they need to stop smoking! Unfortunately such dust can and will collect anywhere, regardless of smoking or not. The only way to prevent this are dust filters in front of the Fans and other Air intakes, plus placing the PC not on the ground will help alot. I've seen "non-smoking" Systems that looked alot worse than above pictures, simply due to location and being never opened. |
|
80)
Message boards :
Number crunching :
"Can't Merge Hosts - They're Incompatible"
(Message 901704)
Posted 30 May 2009 by FalconFly
Post: That is in the server code... He can't change that... Of course he can, albeit only temporarily. client_state.xml <os_version></os_version> Shouldn't be needed though. |
Previous 20 · Next 20
©2025 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.