new core A5 projects: 8106-8108启动/Big Adv Program Ends January 31

金鹏

今天开始新的many-core projects( projects for large-core machines)启动,
仍旧使用2.27版的A5内核(不排除未来版本升级),参数还是 bigadv
虽然仍旧沿用-bigadv标志参数和服务器IP地址(与普通smp区别),
但连接的项目不是原来的BA而是现在新的many-core projects,PPD也暴跌至BA包的1/3  

Re: Updates thread

by kasson » Mon Feb 16, 2015 11:44 pm

Projects 8106 and 8108 are moving to advanced; project 8107 is on beta.

Re: Updates thread

by kasson » Thu Feb 12, 2015 12:51 am

Projects 8106 and 8108 are now on beta once more, with recalculated points and k-factors. According to reports thus far, the PPD is roughly in line with other A3 work units that run well on many cores.

Re: new core A5 projects: 8106-8108

by kasson » Thu Feb 12, 2015 12:49 am

We are now releasing project 8108 to beta:
Points: 7349.0
Preferred: 4.3 days
Deadline: 54 days
k-factor: 3.91

1. [21:15:41] + Attempting to get work packet
   
2. [21:15:41] Passkey found
   
3. [21:15:41] - Will indicate memory of 32127 MB
   
4. [21:15:41] - Connecting to assignment server
   
5. [21:15:41] Connecting to http://assign.stanford.edu:8080/
   
6. [21:15:44] Posted data.
   
7. [21:15:44] Initial: 8F80; - Successful: assigned to (128.143.231.201).
   
8. [21:15:44] + News From Folding@Home:
   
9. [21:15:44] Loaded queue successfully.
   
10. [21:15:44] Sent data
    
11. [21:15:44] Connecting to http://128.143.231.201:8080/
    
12. [21:15:50] Posted data.
    
13. [21:15:51] Initial: 0000; - Receiving payload (expected size: 24253725)
    
14. [22:10:57] - Downloaded at ~7 kB/s
    
15. [22:10:57] - Averaged speed for that direction ~59 kB/s
    
16. [22:10:57] + Received work.
    
17. [22:10:57] Trying to send all finished work units
    
18. [22:10:57] + No unsent completed units remaining.
    
19. [22:10:57] + Closed connections
    
20. [22:10:57]
    
21. [22:10:57] + Processing work unit
    
22. [22:10:57] Core required: FahCore_a5.exe
    
23. [22:10:57] Core found.
    
24. [22:10:57] Working on queue slot 01 [February 1 22:10:57 UTC]
    
25. [22:10:57] + Working ...
    
26. [22:10:57] - Calling './FahCore_a5.exe -dir work/ -nice 19 -suffix 01 -np 32 -checkpoint 15 -verbose -lifeline 3094 -version 634'
    
27. 
    
28. [22:10:57]
    
29. [22:10:57] *------------------------------*
    
30. [22:10:57] Folding@Home Gromacs SMP Core
    
31. [22:10:57] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
    
32. [22:10:57]
    
33. [22:10:57] Preparing to commence simulation
    
34. [22:10:57] - Looking at optimizations...
    
35. [22:10:57] - Created dyn
    
36. [22:10:57] - Files status OK
    
37. [22:10:59] - Expanded 24253213 -> 33024540 (decompressed 136.1 percent)
    
38. [22:10:59] Called DecompressByteArray: compressed_data_size=24253213 data_size=33024540, decompressed_data_size=33024540 diff=0
    
39. [22:10:59] - Digital signature verified
    
40. [22:10:59]
    
41. [22:10:59] Project: 8106 (Run 0, Clone 44, Gen 0)
    
42. [22:10:59]
    
43. [22:10:59] Assembly optimizations on if available.
    
44. [22:10:59] Entering M.D.
    
45. [22:11:05] Mapping NT from 32 to 32
    
46. [22:11:08] Completed 0 out of 250000 steps  (0%)
    
47. [22:21:14] Completed 2500 out of 250000 steps  (1%)
    
48. [22:30:56] Completed 5000 out of 250000 steps  (2%)
    
49. [22:40:34] Completed 7500 out of 250000 steps  (3%)
    
50. [22:50:14] Completed 10000 out of 250000 steps  (4%)
    
51. [22:59:53] Completed 12500 out of 250000 steps  (5%)
    
52. [23:09:32] Completed 15000 out of 250000 steps  (6%)
    
53. [23:19:11] Completed 17500 out of 250000 steps  (7%)
    
54. [23:28:51] Completed 20000 out of 250000 steps  (8%)
    
55. [23:38:30] Completed 22500 out of 250000 steps  (9%)

复制代码

new core A5 projects: 8106-8108

by kasson » Mon Feb 02, 2015 4:41 am

We have a few new many-core projects released and upcoming.
Project 8106 is going straight to full FAH, as it's similar to some work units we've run before and is a control for the new projects.
Project 8107 simulates membrane fusion using tethers analogous to work by our experimental collaborators.
Project 8108 simulates what happens to membrane fusion when the fusion proteins are defective.

More details on all of these soon. There may be a few bumps in the road here--we've had a perfect storm of some hardware failures in our testing infrastructure just as we're trying to get these out to make sure there's enough work for many-core machines. All of these projects should be pretty stable--we've run them on some big supercomputers (in many cases using a lot more cores than on FAH). But we want to get all the details right for you.

Project 8106 is as follows:
5856.0 points
Preferred deadline: 4 days
Final deadline: 43 days
k-factor: 3.92

Re: Updates thread

by kasson » Mon Feb 02, 2015 4:35 am

new project on full: 8106
8107 and 8108 on beta shortly.

===================================================================

Big Adv Program Ends January 31
January 31, 2015 by Vijay Pande ·

As we have previously announced, the bigadv (BA) program will reach end-of-life on January 31, 2015.  We would like to thank all the donors who have contributed to the program.  Throughout the history of Folding@home, donor participation has enabled us to tackle hard problems and engage in bold experiments.  The BA program is an example of both of these––the projects in BA are one that simply could not have been addressed otherwise.  We are still analyzing the results of these projects, but the preliminary data already yields some exciting results that we are comparing to experiments.

Although the BA program is ending, we recognize that the many-core systems previously used to run BA may not perform optimally on all work units in the Folding@home ecosystem (although they will do quite well on most).  In an ideal world, we would have each project performance-benchmarked on a wide variety of systems and a dynamic allocation scheme that matches clients to projects that perform well on their hardware while ensuring a distribution of client capability across Folding@home scientific priorities.  However, this sort of matching is more technically involved (and places larger demands on the assignment server) than we are able to offer at this time.  At BA end-of-life, in order to make sure that these many-core get WUs best suited for them, clients continuing to use the BA flags will be directed to large work units, although these will carry the normal Folding@Home points scheme rather than a BA scheme.  We are planning to expand the diversity of these work units to include a variety of “large” simulation problems, but at this point we are not making statements as to the anticipated longevity of this scheme.

Thank you once again for contributing to Folding@home, whether the BA program or any of our other initiatives.  All our past and future scientific achievements are due to your participation and generosity

米国时间刚过2月1号2时吧? BA木有了

1. [01:56:21] + Attempting to get work packet
   
2. [01:56:21] Passkey found
   
3. [01:56:21] - Will indicate memory of 32127 MB
   
4. [01:56:21] - Connecting to assignment server
   
5. [01:56:21] Connecting to http://assign.stanford.edu:8080/
   
6. [01:56:25] Posted data.
   
7. [01:56:25] Initial: 0000; + Could not authenticate Assignment Server response
   
8. [01:56:25] Connecting to http://assign2.stanford.edu:80/
   
9. [01:56:25] Posted data.
   
10. [01:56:25] Initial: 0000; + Could not authenticate Assignment Server 2 response
    
11. [01:56:25] + Couldn't get work instructions.
    
12. [01:56:25] - Attempt #1  to get work failed, and no other work to do.
    
13. Waiting before retry.
    
14. [01:56:45] + Attempting to get work packet
    
15. [01:56:45] Passkey found
    
16. [01:56:45] - Will indicate memory of 32127 MB
    
17. [01:56:45] - Connecting to assignment server
    
18. [01:56:45] Connecting to http://assign.stanford.edu:8080/
    
19. [01:56:50] Posted data.
    
20. [01:56:50] Initial: 0000; + Could not authenticate Assignment Server response
    
21. [01:56:50] Connecting to http://assign2.stanford.edu:80/
    
22. [01:56:54] Posted data.
    
23. [01:56:54] Initial: 0000; + Could not authenticate Assignment Server 2 response
    
24. [01:56:54] + Couldn't get work instructions.
    
25. [01:56:54] - Attempt #2  to get work failed, and no other work to do.
    
26. Waiting before retry.
    
27. [01:57:15] + Attempting to get work packet
    
28. [01:57:15] Passkey found
    
29. [01:57:15] - Will indicate memory of 32127 MB
    
30. [01:57:15] - Connecting to assignment server
    
31. [01:57:15] Connecting to http://assign.stanford.edu:8080/
    
32. [01:57:21] Posted data.
    
33. [01:57:21] Initial: 0000; + Could not authenticate Assignment Server response
    
34. [01:57:21] Connecting to http://assign2.stanford.edu:80/
    
35. [01:57:32] Posted data.
    
36. [01:57:32] Initial: 0000; + Could not authenticate Assignment Server 2 response
    
37. [01:57:32] + Couldn't get work instructions.
    
38. [01:57:32] - Attempt #3  to get work failed, and no other work to do.
    
39. Waiting before retry.
    
40. [01:57:54] + Attempting to get work packet
    
41. [01:57:54] Passkey found
    
42. [01:57:54] - Will indicate memory of 32127 MB
    
43. [01:57:54] - Connecting to assignment server
    
44. [01:57:54] Connecting to http://assign.stanford.edu:8080/
    
45. [01:57:59] Posted data.
    
46. [01:57:59] Initial: 0000; + Could not authenticate Assignment Server response
    
47. [01:57:59] Connecting to http://assign2.stanford.edu:80/
    
48. [01:58:00] Posted data.
    
49. [01:58:00] Initial: 0000; + Could not authenticate Assignment Server 2 response
    
50. [01:58:00] + Couldn't get work instructions.
    
51. [01:58:00] - Attempt #4  to get work failed, and no other work to do.
    
52. Waiting before retry.
    
53. [01:58:51] + Attempting to get work packet
    
54. [01:58:51] Passkey found
    
55. [01:58:51] - Will indicate memory of 32127 MB
    
56. [01:58:51] - Connecting to assignment server
    
57. [01:58:51] Connecting to http://assign.stanford.edu:8080/
    
58. [01:58:57] Posted data.
    
59. [01:58:57] Initial: 0000; + Could not authenticate Assignment Server response
    
60. [01:58:57] Connecting to http://assign2.stanford.edu:80/
    
61. [01:59:01] Posted data.
    
62. [01:59:01] Initial: 0000; + Could not authenticate Assignment Server 2 response
    
63. [01:59:01] + Couldn't get work instructions.
    
64. [01:59:01] - Attempt #5  to get work failed, and no other work to do.
    
65. Waiting before retry.
    
66. [02:00:28] + Attempting to get work packet
    
67. [02:00:28] Passkey found
    
68. [02:00:28] - Will indicate memory of 32127 MB
    
69. [02:00:28] - Connecting to assignment server
    
70. [02:00:28] Connecting to http://assign.stanford.edu:8080/
    
71. [02:00:33] Posted data.
    
72. [02:00:33] Initial: 0000; + Could not authenticate Assignment Server response
    
73. [02:00:33] Connecting to http://assign2.stanford.edu:80/
    
74. [02:00:36] Posted data.
    
75. [02:00:36] Initial: 0000; + Could not authenticate Assignment Server 2 response
    
76. [02:00:36] + Couldn't get work instructions.
    
77. [02:00:36] - Attempt #6  to get work failed, and no other work to do.
    
78. Waiting before retry.
    
79. [02:03:26] + Attempting to get work packet
    
80. [02:03:26] Passkey found
    
81. [02:03:26] - Will indicate memory of 32127 MB
    
82. [02:03:26] - Connecting to assignment server
    
83. [02:03:26] Connecting to http://assign.stanford.edu:8080/
    
84. [02:03:27] Posted data.
    
85. [02:03:27] Initial: 0000; + Could not authenticate Assignment Server response
    
86. [02:03:27] Connecting to http://assign2.stanford.edu:80/
    
87. [02:03:27] Posted data.
    
88. [02:03:27] Initial: 0000; + Could not authenticate Assignment Server 2 response
    
89. [02:03:27] + Couldn't get work instructions.
    
90. [02:03:27] - Attempt #7  to get work failed, and no other work to do.
    
91. Waiting before retry.
    
92. [02:08:55] + Attempting to get work packet
    
93. [02:08:55] Passkey found
    
94. [02:08:55] - Will indicate memory of 32127 MB
    
95. [02:08:55] - Connecting to assignment server
    
96. [02:08:55] Connecting to http://assign.stanford.edu:8080/
    
97. [02:08:56] Posted data.
    
98. [02:08:56] Initial: 0000; + Could not authenticate Assignment Server response
    
99. [02:08:56] Connecting to http://assign2.stanford.edu:80/
    
100. [02:08:57] Posted data.
     
101. [02:08:57] Initial: 0000; + Could not authenticate Assignment Server 2 response
     
102. [02:08:57] + Couldn't get work instructions.
     
103. [02:08:57] - Attempt #8  to get work failed, and no other work to do.
     
104. Waiting before retry.
     
105. [02:19:40] + Attempting to get work packet
     
106. [02:19:40] Passkey found
     
107. [02:19:40] - Will indicate memory of 32127 MB
     
108. [02:19:40] - Connecting to assignment server
     
109. [02:19:40] Connecting to http://assign.stanford.edu:8080/
     
110. [02:19:40] Posted data.
     
111. [02:19:40] Initial: 0000; + Could not authenticate Assignment Server response
     
112. [02:19:40] Connecting to http://assign2.stanford.edu:80/
     
113. [02:19:42] Posted data.
     
114. [02:19:42] Initial: 0000; + Could not authenticate Assignment Server 2 response
     
115. [02:19:42] + Couldn't get work instructions.
     
116. [02:19:42] - Attempt #9  to get work failed, and no other work to do.
     
117. Waiting before retry.
     
118. [02:22:56] - Autosending finished units... [February 1 02:22:56 UTC]
     
119. [02:22:56] Trying to send all finished work units
     
120. [02:22:56] Project: 8104 (Run 0, Clone 60, Gen 255)
     
121. [02:22:56] - Error: Could not get length of results file work/wuresults_00.dat
     
122. [02:22:56] - Error: Could not read unit 00 file. Removing from queue.
     
123. [02:22:56] + Sent 0 of 1 completed units to the server
     
124. [02:22:56] - Autosend completed
     
125. [02:41:10] + Attempting to get work packet
     
126. [02:41:10] Passkey found
     
127. [02:41:10] - Will indicate memory of 32127 MB
     
128. [02:41:10] - Connecting to assignment server
     
129. [02:41:10] Connecting to http://assign.stanford.edu:8080/
     
130. [02:41:10] Posted data.
     
131. [02:41:10] Initial: 0000; + Could not authenticate Assignment Server response
     
132. [02:41:10] Connecting to http://assign2.stanford.edu:80/
     
133. [02:41:12] Posted data.
     
134. [02:41:12] Initial: 0000; + Could not authenticate Assignment Server 2 response
     
135. [02:41:12] + Couldn't get work instructions.
     
136. [02:41:12] - Attempt #10  to get work failed, and no other work to do.
     
137. Waiting before retry.
     
138. [03:23:56] + Attempting to get work packet
     
139. [03:23:56] Passkey found
     
140. [03:23:56] - Will indicate memory of 32127 MB
     
141. [03:23:56] - Connecting to assignment server
     
142. [03:23:56] Connecting to http://assign.stanford.edu:8080/
     
143. [03:23:58] Posted data.
     
144. [03:23:58] Initial: 0000; + Could not authenticate Assignment Server response
     
145. [03:23:58] Connecting to http://assign2.stanford.edu:80/
     
146. [03:23:58] Posted data.
     
147. [03:23:58] Initial: 0000; + Could not authenticate Assignment Server 2 response
     
148. [03:23:58] + Couldn't get work instructions.
     
149. [03:23:58] - Attempt #11  to get work failed, and no other work to do.
     
150. Waiting before retry.
     

复制代码

==============================
好消息据潘德说以后会有大SMP项目,大概是替代BA的?
否则那些大婶们的多路服务器跑SMP小包几十分钟一个包,的确郁闷

[–]VijayPande-FAHF@h Director 5 指標 1 月 前
We are looking into setting up projects for large-core machines. The benchmarking would be done along the current SMP client benchmark scheme.

https://www.reddit.com/r/foldingathome/comments/2o7zka/what_is_in_the_future_for_the_ba_folders_come/

Re: Last day of bigadv today ...

by kasson » Sun Feb 01, 2015 5:08 am

Please note Prof. Pande's blog post as of today (linked above and below). The -bigadv flags will no longer be linked to the bigadv program (i.e. no separate points scheme) but will direct clients to work units that should scale well on large-core machines. The diversity of these work units is expected to expand in the future. At the moment, we don't have a roadmap we can share for this particular mechanism, but we will continue to do our best to ensure that large-core machines are matched with work units that they perform well on.

金鹏

@ocw
今天开始新的many-core projects( projects for large-core machines)启动,
仍旧使用2.27版的A5内核(不排除未来版本升级),
虽然仍旧沿用-bigadv标志参数和服务器IP地址(与普通smp区别),
但连接的项目不是原来的BA而是现在新的many-core projects,
新项目两个变化: 1/接包大小从原来的30兆变为现在的23兆
                       2/ PPD得分从原来的37万增至现在43万(同样参数的2P_2687W)

金鹏

ocw 发表于 2015-2-2 09:43
4P 太耗電了, 2p 還可以.
3 台 4p 的電費可以跑 9 張 980,  129w ppd vs 315w ppd.

排除网络造成延迟,4P_6272应该有55万+的PPD,
到时我看看估算准不准


这个12.04的鸡血系统下应该能装V7客户端跑显卡吧?

金鹏

wpf999 发表于 2015-2-2 10:09
弱弱的问一句，新core A5对核数有什么要求？ 什么样的U能按时完成？

前面讲了,目前内核还是老的2.27版

项目方向也许有所调整,貌似原来的BIGADV走错路了

不清楚接到A5内核包的条件是不是原来的24T+,那个12T+机器可以试试

8106目前奖励期限4天,过期期限43天,貌似奖励因数与BA明显不同.范围幅度更宽广

感觉12T+都能跑完,分数高低而已

金鹏

Re: Updates thread

by kasson » Thu Feb 05, 2015 3:01 am

Projects 8106-8108 have been pulled from assign. We're re-calculating points yields. There was a mistake in points assignment that also triggered a safety feature in the stats code. We're fixing all these problems and will be *slowly* re-releasing through the beta process as everything comes online. Many apologies for the problems here.

金鹏

Re: new core A5 projects: 8106-8108

by kasson » Thu Feb 05, 2015 3:09 am

Many apologies for the problems here. I was trying to ensure work unit availability for a smoother transition post-bigadv; in retrospect, it would have been better to accept a "gap" period for many-core-optimized work units and make sure that the rollout process worked properly.
What happened here was the following:
1. In benchmarking the work units, I mis-combined formulae and ended up with a k-factor that was substantially too high.
2. The very high bonus values that resulted triggered a safety threshold in the stats code, resulting in no bonus points at all for some work units.
3. We had a hardware failure on one of the benchmarking machines, slowing correction of this process.

We are in the process of fixing all three of these. Once these are done, we will proceed with a slow beta rollout. Be advised that PPD values will be lower--they were inflated substantially by the benchmarking mistake I made. Many apologies.

金鹏

Re: new core A5 projects: 8106-8108

by kasson » Thu Feb 05, 2015 9:37 am

The missing-bonus returns have now been rerun and should have appropriate credit.

金鹏

wpf999 发表于 2015-2-5 13:17
斯坦福大学的研究员也能代错公式，这不科学。。。

跑了一天的7504小包,才13万PPD,跑100分钟一个包

Re: new core A5 projects: 8106-8108

by kasson » Thu Feb 05, 2015 1:49 am

We've shut off assigns for these projects until we can resolve problems. Thanks for your patience.

PS we have records for all returned work units.

Re: new core A5 projects: 8106-8108

by kasson » Thu Feb 05, 2015 9:37 am

The missing-bonus returns have now been rerun and should have appropriate credit.

金鹏

Lynt 发表于 2015-2-5 11:00
这是要减分的节奏？

K因数大幅缩水

Re: new core A5 projects: 8106-8108

by kasson » Sat Feb 07, 2015 11:35 am

We are getting ready to restart project 8106 on beta. k-factor is 3.92; all other numbers are unchanged. 8107 and 8108 will follow in the next several days as our benchmark re-runs finish.

The changes at this time are only in project availability, no core changes.

金鹏

wpf999 发表于 2015-2-7 20:23
双2678w现在多少PPD？

加了bigbeta参数在接8106新包,下载速度坑,大小从原来的23兆变成现在29兆+

Re: new core A5 projects: 8106-8108

by EXT64 » Sat Feb 07, 2015 1:03 pm

That PPD looks much more in line with standard SMP. A couple comparisons:

8106 -> 130k PPD
8822 -> 139k PPD
7504 -> 120k PPD
8814 -> 133k PPD

So I would say benchmarking was a success.

Edit: for reference this is a 2P E5-2670.

金鹏

cuda 发表于 2015-2-7 20:57
8106原先的k-factor是39.2，现在改成3.92，之前是点错了小数点？
由于PPD和k-factor的1/2方成正比，那么8 ...

兄弟精确43*.316=13万
我现在32用了30T跑的12.8万PPD

金鹏

Re: Updates thread

by kasson » Thu Feb 12, 2015 12:51 am

Projects 8106 and 8108 are now on beta once more, with recalculated points and k-factors. According to reports thus far, the PPD is roughly in line with other A3 work units that run well on many cores.

金鹏

wpf999 发表于 2015-2-13 13:10
fah官方其实是为了鼓励大家用GPU，又快又省电

32T也放弃了,全面转跑WCG刷牌子,未来就留O版捐助的64C跑鸡血下8106-8108

Re: new core A5 projects: 8106-8108

by kasson » Sat Feb 14, 2015 2:24 am

Thanks for the reports thus far. 8107 is also up:
Points: 7358.0
Preferred deadline: 4 days
Final deadline: 54 days
k-factor: 3.92

金鹏

Re: Updates thread

by kasson » Mon Feb 16, 2015 11:44 pm

Projects 8106 and 8108 are moving to advanced; project 8107 is on beta.