TAS-member thread: October 2005

Page 11 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
14 of 34 eligible votes are in (41% of all eligible [as of now! if more join, than this changes!]) :)

Seti@home - BOINC: -----------------> 9 votes (64% of those who voted, 26,5% of all who are eligible to vote!!) ;)
LHC@Home - BOINC -----------------> 1 vote
rerun Rosetta@home - BOINC: ----> 3 votes ;)
rerun Predictor@home - BOINC: ---> 1 vote

no vote given: -----------------------> 20 votes! :( (the non voters win??!!??)

As it is now at least 7 votes are needed to change the outcome of the poll ...

2½ days are left! Go, Go, GO, GOOOO and ........................... vote please!?!

Let' s get votin'!!! :D
Should really a minority (no matter how you look at it: either 9 or 14) of TAS decide? :shocked:
 
rise

rise

Diamond Member
Dec 13, 2004
9,116
46
91
i dunno, how many active TAS members do we have? it looks to me that only ~14 or so ran rosetta this month, no?
 
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
11 TAS-members have crunched Rosetta@home in October. :D
On the list there are 36 member accounts, two of which are very inactive (they were created by Johan when setting up the site) - that leaves 34 members.
During the last 6-7 months or so, I have seen posts from 24 members(which indicates some kind of activity ...) ;)

I have PM'ed to 33 of the members (I did not PM to myself ...), one of those has left TeAm Anandtech. That leaves 33 members, including myself :)
14 have voted, that is 42%.

In the previous poll 11 members voted ...
 
BlackMountainCow

BlackMountainCow

Diamond Member
May 28, 2003
5,759
0
0
LOL, Peter, you really do love stats and numbers, don't you? ;)

Let me guess, a cup of coffee and a peanut butter & jelly bread , right?

:)

Thx for you efforts man!
 
amdxborg

amdxborg

Diamond Member
Aug 27, 2002
6,790
23
81
lol Well voting is looking good.. :D Thanx for setting up everything Peter! :beer:;)
 
kb3edk

kb3edk

Senior member
Jul 11, 2004
494
0
0
Originally posted by: rise4310
:)

i didn't realize how much i took the x2 for granted until i started setting up his stuff on my , errr his, 3000+ venice. wow, what a difference :Q

these are great chips.
Click to expand...

I couldn't agree with you more... the X2 is quite possibly the best processor ever for distributed computing!

1: Fits in most Socket 939 motherboards (unlike Intel dual cores that require a new board)
2: Affordable for the home user, unlike previous SMP architectures which were intended for the enterprise setting and were priced accordingly
3: Helps save space for multi-computer DC "farms" - for big city apartment dwellers like me this is crucial!
4: Barely noticable increase in heat/cooling requirements over single core Socket 939 CPUs
5: Literally twice the points per day output when given sufficient RAM! You might think that this wouldn't be the case, since for example, the 2.0 GHz, 512KB L2 cache Athlon 64 is a 3200+ in single core, 3800+ in dual core. However all this shows is how weird AMD's rating system can be :confused:

The only thing stopping me from buying another X2 is wanting to wait until the new AMD Socket M2 comes out next year! :cool:
 
BlackMountainCow

BlackMountainCow

Diamond Member
May 28, 2003
5,759
0
0
Originally posted by: kb3edk

The only thing stopping me from buying another X2 is wanting to wait until the new AMD Socket M2 comes out next year! :cool:
Click to expand...

easy way around that: Buy the Asrock 939 Dual SATA II board! I've got one and it not only offers both PCIe and full AGP support (yip, ULi chipset) but it also has a "future slot" that offers support for a riser card that will hold a M2 socket. So: Have an X2 now but also have the option for an M2 later on!

:thumbsup:
 
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
Don't say that, mrwizer. According to the stats you should pass me in 1 day - 24 - 29 hours. That is enough - if I stick to my pace.
Anyhow: "It ain't over until the fat lady sings" ;) whoever said that ...
Crunch on, you may very well catch me on the last day! Would be fun though, if I put on the last proverbial coal ... ;) :p
Now, where the h..k is the shovel? :D
 
GLeeM

GLeeM

Elite Member
Apr 2, 2004
7,199
128
106
Originally posted by: petrusbroder
11 TAS-members have crunched Rosetta@home in October. :D
On the list there are 36 member accounts, two of which are very inactive (they were created by Johan when setting up the site) - that leaves 34 members.
During the last 6-7 months or so, I have seen posts from 24 members(which indicates some kind of activity ...) ;)

I have PM'ed to 33 of the members (I did not PM to myself ...), one of those has left TeAm Anandtech. That leaves 33 members, including myself :)
14 have voted, that is 42%.

In the previous poll 11 members voted ...
Click to expand...

Thanks for setting up everything for TAS, Peter :thumbsup: Awesome job!

The TAS forums have always been open to anyone on the TeAm, except for one section (where we discussed the next project) for a while at the beginning to keep out spies - we were paranoid back then :confused:

For the first half year or so - if you wanted to join TAS you had to PM one of the leaders. I should have brought this up earlier, but didn't think it mattered that much - until your numbers started getting bigger and bigger. Those friendly people that posted on the TAS forum are not members unless they said they wanted to join.
 
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
Thanks, GLeeM, for the info. I will be more careful in future posts. ;)
It is not easy, since I am one of those late-comers and do not have all the history behind TAS. I would be happy if you could post some of the history in TAS forums somewhere.
OTOH, I do not think that openness is bad at all ... :) , coming from a country where openness is law (i.e. everything is public unless explicitly made secret by law). :Q
 
R

Rattledagger

Elite Member
Feb 5, 2001
2,994
19
81
Hmm, did Anandtech fall off the the net for some hours... :confused:

Anyway, SETI@Home and wu... old or new...


Matt Lebofsky posted many months ago, what then needed, he'll grab a tape more or less randomly from a box of tapes that have never been split before. After some initial tests, both projects starts splitting from the same tape, this to make sure SETI@Home/BOINC crunches correctly...

SETI@Home/BOINC will issue the same wu to 4 different users, and only if error/past deadline is more copies sent out. "Classic" on the other hand will send-out whatever is currently cached, and if there's a hold-up in production will just re-issue some wu indefinitely...

Also mentioned, the 3 servers acting as splitters runs one instance for each of the projects, meaning they're battling for resources...



Now, some months later, SETI@Home/BOINC has expanded to 5 splitter-servers, not sure but AFAIK "classic" is still on 3... if they're still splitting for "classic"...

Atleast going by "current progress", "classic" have been stuck on 227648070 wu for a couple weeks... It is a possibility the stats-pages have screwed-up, so let's take a quick look on whatever "classic" is currently sending-out...

Well, a quick test reveals they're handing-out 04no03aa.16415.18xxx and increasing, while SETI@Home/BOINC was handing-out 04no03aa in beginning of September...
Looking even closer, the Enterprise-queue have queued since 05.October, 04no03aa.16415.1266.* and 04no03aa.16415.32320.* while BBR queue VI has one as early as 17.September, 04no03aa.16415.22193.361066.238.


For anyone not remember the naming-conventions on seti-wu:
04no03aa.16415.22193.361066.238:
04no03aa: date started recorded, and aa is 1st tape started this day, ab 2nd tape and so on. ab can therefore largely be recorded the next day. ;)
16415: process-id on the splitter-process that split this wu, so can be the same for the whole tape. If re-split it will be different id, and the only difference to a SETI@Home/BOINC-wu is the process-id...
22193: last block split. Each tape has roughly 33k blocks, this means roughly 67% out in the tape.
361066: byte-offset from last block split or something...
238: one of the 256 pieces, each roughly 10kHz wide recorded at the same time.


All of this means, wu "classic" is to-day handing-out was already validated by SETI@Home/BOINC at start of September. Also, it strongly looks like "classic" isn't even splitting wu any longer, since the Enterprise-queue had wu from the start of tape and the end of tape as early as 5. October, but just re-loops the same bunch of wu again and again... This also means "classic" isn't doing anything useful any longer, except possibly heating your room... :evil:


Anyway, Matt Lebofsky posted this a fortnight ago:
At the current "burn rate" I have enough tapes near my desk for about 4-5 months. Soon we will be releasing a new client that will vastly increase the science analysis, and therefore vastly slow increase the turnaround time for each workunit. So these tapes will actually last about 2 years. Also there are some really old tapes that never got fully analyzed in classic for one reason or another that we will throw back into the "to do" pile.

As well, the new data recorder is nearing completion and we will hopefully hook it up with the multi beam receiver in Arecibo - which will mean a whole new batch of data and a new analysis engine.

So.. we're not starved for data.
Click to expand...


Bottom line is, SETI@Home/BOINC is splitting new tapes never distributed before. :cool:

 
petrusbroder

petrusbroder

Elite Member
Nov 28, 2004
13,348
1,155
126
Thanks a lot for the the timely information, Rattledagger! I will quote the gist of it (i.e. the last sentence) at TAS-forum, where the Q was asked ... :D
 
BlackMountainCow

BlackMountainCow

Diamond Member
May 28, 2003
5,759
0
0
Thx a lot for the answer Rattle! Now that fully answers my question and sheds some better light on SETI BOINC!

:thumbsup:
 
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom