Posts by Alez

1) Message boards : Cafe MilkyWay : The Scottish Boinc Team - Old and new members, we have new forum and web pages. (Message 66431) Posted 12 May 2017 by Alez Post: The Scottish Boinc Team (established in 1999 as Scotland's Seti), one of the oldest and most successful BOINC teams crunching today have new web pages and forum available at https://tsbt.co.uk We crunch for all current BOINC projects and are always at the forefront of adding new projects that start up within BOINC. Our aim is to use our spare or dedicated computing power to advance human knowledge while enjoying some friendly competition amongst ourselves and with other teams. If you want to be part of a lively, friendly team, please come join us, everyone is welcome. Join the TSBT Team Forum. Take part in shaping your teams future, chat with other members, get help and advice or help others. We participate in all BOINCstats Team challenges and hold competitions both within our own team and with other rival teams. We maintain a very informative, user and newbie friendly, forum with our own team statistics, team and individual milestones, BOINC project and hardware information across all supported platforms and lots of help. Join us, you will be assured of a very warm welcome and it's easy to do. For those that havn't ben around for a while, have a look round our forum, the heart of the team. https://www.tsbt.co.uk/forum/ Guests can read also read most parts of the forum without registration.Theres a guest forum where you can stop by and say hi. Registered members will be able to immediately post and be granted access to additional areas of the forum. Active TSBT members can gain access to the private team areas by request. Full instructions on joining are available along with the forum rules on the website. News updates from all active projects are also part of the forum along with other features in development. All old members and anyone thinking of joining The Scottish Boinc Team, or if you're just browsing have a look at the heart of our team. Everyone is welcome. https://www.tsbt.co.uk/forum/
2) Message boards : Number crunching : Feeder Down (Message 62369) Posted 23 Sep 2014 by Alez Post: still down :(
3) Message boards : Number crunching : CPU/GPU Comparison (do we need CPU apps when GPU app is available) (Message 61930) Posted 21 Jun 2014 by Alez Post: Arivald, I have a very mixed selection of cpu's / gpu's , ranging from i7's driving multiple 7970's to a single core p4 driving a 660ti simply because it can. What systems I choose to dedicate to Boinc is my choice. How much I choose to spend on electricity is my choice. What you choose to crunch on which systems and how much you choose to pay is your choice. Your choice is not my choice. My choice should not be forced on you as neither should your choice be forced on me. If a system is still capable of producing meaningful work within the due date decided by the project in question then whether or not I choose to run that system is my choice. Take a minute to reflect on the words ' my choice ' as they apply to me as much as they apply to you. oh, and if you want power efficiency , stop running Boinc and turn your system off.
4) Questions and Answers : Windows : Computation Errors on Cataylist 14.4 (Message 61909) Posted 17 Jun 2014 by Alez Post: nope, error not caused by speed , over clocking etc,. etc. This has been a long running issue. With 5xxx and 6xxx series cards you either have to remove milkyway v2 from your preferences and run mod fit or use cat 12.8 or earlier. All the 13 and 14 drivers cause these cards to error here with a memory violation.
5) Message boards : Number crunching : Computation Errors on Cataylist 14.4 (Message 61908) Posted 17 Jun 2014 by Alez Post: This has been a long running issue. With 5xxx and 6xxx series cards you either have to remove milkyway from your preferences and run mod fit or use cat 12.8 or earlier. All the 13 and 14 drivers cause these cards to error here with a memory violation.
6) Message boards : Number crunching : adequately supply or WUs: PLEASE: increase max_wus_in_progress or enable the Computing preference settings (Message 61660) Posted 28 Apr 2014 by Alez Post: I am currently running 5 different computers on the Milkyway challenge at boincstats. Between them they have 11 GPU's, some of which get through a unit in less than 30 seconds. Not a single one of these computers has had or ever has had any issues getting enough work to complete. There is no problem with the work fetch on the server. As you have been told many times now update the version of Boinc that you are running.
7) Message boards : Number crunching : Why am not getting any work ? (Message 61558) Posted 19 Apr 2014 by Alez Post: If you are still stuck, what worked for me ( with a 6970 and a 6990 which wouldn't work with later drivers ) was to run amdcleanuputility ( downloaded from AMD site ) Then install cal 12.10 and SDK v2.8. Install the drivers from AMD, not those provided by windows.
8) Message boards : Number crunching : adequately supply or WUs: PLEASE: increase max_wus_in_progress or enable the Computing preference settings (Message 61505) Posted 11 Apr 2014 by Alez Post: I am looking into the impact of more WUs in progress on the efficiency of the server algorithm. But more specifically are others having issues with the WUs not be generated often enough or requesting new WUs. Jeff and yet somehow I'm managing close to 1 million a day on the current challenge on Boincstats and not running out of units ? and never had an issue getting work to keep a minimum of 7 computers crunching ? no problem getting units to crunch here.
9) Message boards : Number crunching : R9 290 (Message 61498) Posted 10 Apr 2014 by Alez Post: I am currently running a AMD R9 - 270X (Pitcairn) with little problem getting WU's. Even though this card is rated 40% faster in the specs released it is running 50% slower than the 7900 series AMD's. I think the problem relates to the programming of the WU's. Regards JohnG It's also got to do with the card itself a 270X has 1280 stream processors, while a 7970 has 2048. That means a 7970 is going to process things roughly twice as fast, it is NOT exactly linear, as a 270X gpu will. A 280X has 2048 while a 290X has 2816 stream processors. If you think of stream processors like little tiny cpu cores you can see why more stream processors means faster crunching. This obviously ONLY applies to AMD cards as Nvidia cards use what they call 'cuda cores' and it is MUCH different. Also if a card is over clocked compared to another card it can make up some of the difference as the processors will be working faster. The new R series AMD cards are generally clocked faster then the cards they were based on making then slightly faster. Actually the difference between the 7970 and the 270X is far, far greater than the single precision Gflops and number of stream processors would suggest The reason is simple, the 7970 is 1/4 double precision, the 270X is only 1/16th double precision. In double precision the 7970 is rated at 947.2 Gflops, the 270X at only 260 Gflops. Thats why the 270X is disappointing on Milkyway as it's single precision is quite good (for gaming ) but it's double precision is artificially crippled to nVidia levels. The 280X is rated at 1/4 double precision and the 290X at 1/8. There is nothing wrong with the units, it's simply that the 270X is very poor at double precision calculations.
10) Message boards : Number crunching : Benchmark results - times wanted for any hardware, CPU or GPU, old or new! (Message 61490) Posted 8 Apr 2014 by Alez Post: Yes, 2 units running, 1 on each card. All the projects I have run on it so far treat it as 2 x 6970's basically. Think Moo might be the only project that would treat it as a dual card unless anyone can tell me different. I'm still very disappointed with the DP performance of the R9 270x despite it being above a Titan on the list :)
11) Message boards : Number crunching : Benchmark results - times wanted for any hardware, CPU or GPU, old or new! (Message 61471) Posted 6 Apr 2014 by Alez Post: HD 6990 ( 830 Mhz ) stock not uber mode - 165 secs windows 7 R9 270x ( 1000 Mhz ) - 407 secs windows 7
12) Message boards : Number crunching : Benchmark results - times wanted for any hardware, CPU or GPU, old or new! (Message 61321) Posted 4 Mar 2014 by Alez Post: must have missed that memo too, still makes the r9 270 disappointing for an ATI and the 610 the slowest card so far, but it never misses a beat. Don't recall it ever throwing an error :)
13) Message boards : Number crunching : Benchmark results - times wanted for any hardware, CPU or GPU, old or new! (Message 61318) Posted 3 Mar 2014 by Alez Post: or you take the 50's and the 76's and call it 63 :) I only posted the numbers because at the time you had no data from either of my three cards in this machine, and only looked back 2 pages ( which were all nearly 50's ). How you choose to interpret it, that's your call and your welcome to use it as you see fit I also have 7750's and 7770's running in various systems but you already have data from them.
14) Message boards : Number crunching : Benchmark results - times wanted for any hardware, CPU or GPU, old or new! (Message 61314) Posted 3 Mar 2014 by Alez Post: nope, times consistently at 50 secs. 680783869 509228308 3 Mar 2014, 12:57:49 UTC 3 Mar 2014, 19:53:37 UTC Completed and validated 50.42 5.90 213.76 Milkyway@Home Separation (Modified Fit) v1.28 (opencl_amd_ati) 680783868 509228307 3 Mar 2014, 12:57:49 UTC 3 Mar 2014, 19:51:20 UTC Completed and validated 50.24 5.37 213.76 Milkyway@Home Separation (Modified Fit) v1.28 (opencl_amd_ati) 680783868 509228307 3 Mar 2014, 12:57:49 UTC 3 Mar 2014, 19:51:20 UTC Completed and validated 50.24 5.37 213.76 Milkyway@Home Separation (Modified Fit) v1.28 (opencl_amd_ati) 680783867 509228306 3 Mar 2014, 12:57:49 UTC 3 Mar 2014, 19:52:29 UTC Completed and validated 50.44 6.02 213.76 Milkyway@Home Separation (Modified Fit) v1.28 (opencl_amd_ati) 680783866 509228305 3 Mar 2014, 12:57:49 UTC 3 Mar 2014, 19:51:20 UTC Completed and validated 51.31 6.05 213.76 Milkyway@Home Separation (Modified Fit) v1.28 (opencl_amd_ati) 680783865 509228304 3 Mar 2014, 12:57:49 UTC 3 Mar 2014, 18:40:20 UTC Completed and validated 51.61 6.08 213.76 Milkyway@Home Separation (Modified Fit) v1.28 (opencl_amd_ati) computer is here. http://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=479121 runs 1 x 7970 clocked to 1050, 1 x r9 270 clocked to 1000 and 1 x gt610
15) Message boards : Number crunching : Benchmark results - times wanted for any hardware, CPU or GPU, old or new! (Message 61294) Posted 2 Mar 2014 by Alez Post: HD 7970 51 secs R9 270 282 secs GT610 3500 secs all under win 7, boinc 7.2.39 The R9 results are suspect as milkyway reports 38 DP flops, default to 100 flops which is very incorrect. Will update to beta driver and try again.
16) Message boards : Cafe MilkyWay : Team Recruitment III (Message 60509) Posted 3 Dec 2013 by Alez Post: The Scottish Boinc Team Welcome to The Scottish Boinc Team ( established in 1999 as Scotland's Seti ), one of the oldest and we'd like to think, best Boinc teams around. After joining the team here, head over to our forum http://www.dunadd.co.uk/seti/forum/and Join us. We crunch for all Boinc projects and are always adding new ones as they come along. Our aim is to use our spare computing power to advance human knowledge while enjoying some friendly competition amongst ourselves and with other teams. So if you're Scottish by birth, ancestry, residence, emotional attachment or even if you've no link whatsoever with Scotland but just fancy being part of a lively friendly team, please come join us, everyone is welcome. http://www.dunadd.co.uk/seti/forum/ Take part in shaping your team, chat, get help and advice or help others. Participate in Team challenges and competitions both with ourselves and other rival teams. It's a very informative, user and newbie friendly, forum with stats, team and individual milestones, information and lots of help. Come join us, you will be assured of a warm welcome and it's easy to do. Come see our website and forum. Reading is free without registration. Team members can Register and will be able to immediately post and be granted access to additional private team areas.
17) Message boards : Number crunching : N-Body blues (Message 57062) Posted 28 Jan 2013 by Alez Post: I just aborted two with ridiculous run times but unfortunately didn't realise they were running in the first place. Will I get any credit for wasting so much CPU time ( over 300 hrs !!! ) http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=295130394 http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=294433850 also both those units are still out there wasting someone elses time. Can you not perform a server side abort ?
18) Message boards : News : Stopped N-Body Runs (Message 57061) Posted 28 Jan 2013 by Alez Post: I just aborted two with ridiculous run times but unfortunately didn't realise they were running in the first place. Will I get any credit for wasting so much CPU time ( over 300 hrs !!! ) http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=295130394 http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=294433850 also both those units are still out there wasting someone elses time. Can you not perform a server side abort ?