Project 2665: whew!

[coo-coo-clocker]

This thing is brutal. :hitself:
I'm seeing 16m50s frame times for this WU. OUCH!
And it's only worth 1275 points. Urrrm.

[Kingcarcas]

Both of my rigs are getting Amber WUs, 3/400 PPD :(

[SparkyJJO]

I haven't gotten that unit yet...

[coo-coo-clocker]

[15:50:39] Preparing to commence simulation
[15:50:39] - Ensuring status. Please wait.
[15:50:44] - Starting from initial work packet
[15:50:45]
[15:50:45] Project: 2665 (Run 0, Clone 309, Gen 1)
[15:50:45]
[15:50:47] Assembly optimizations on if available.
[15:50:47] Entering M.D.
[15:51:06] ial work packet
[15:51:06]
[15:51:06] Project: 2665 (Run 0, Clone 309, Gen 1)
[15:51:06]
[15:51:09] 665 (Run 0, Clone 309, Gen 1)
[15:51:09]
[15:51:15] Entering M.D.
[15:51:21] Rejecting checkpoint
[15:51:22] Protein: HGG in water
[15:51:22] Writing local files
[15:51:30] Extra SSE boost OK.
[15:51:31] Writing local files
[15:51:31] Completed 0 out of 250000 steps (0 percent)
[16:08:13] Writing local files
[16:08:13] Completed 2500 out of 250000 steps (1 percent)
[16:25:23] Writing local files
[16:25:23] Completed 5000 out of 250000 steps (2 percent)
[16:42:32] Writing local files
[16:42:32] Completed 7500 out of 250000 steps (3 percent)
[16:59:42] Writing local files
[16:59:42] Completed 10000 out of 250000 steps (4 percent)
<Meanwhile, the following day...>
[19:14:22] Writing local files
[19:14:23] Completed 242500 out of 250000 steps (97 percent)
[19:30:48] Writing local files
[19:30:48] Completed 245000 out of 250000 steps (98 percent)
[19:47:13] Writing local files
[19:47:13] Completed 247500 out of 250000 steps (99 percent)
[20:03:38] Writing local files
[20:03:38] Completed 250000 out of 250000 steps (100 percent)
[20:03:38] Writing final coordinates.
[20:03:40] Past main M.D. loop
[20:03:40] Will end MPI now
[20:04:40]
[20:04:40] Finished Work Unit:
[20:04:40] - Reading up to 21310704 from "work/wudata_03.arc": Read 21310704
[20:04:40] - Reading up to 553372 from "work/wudata_03.xtc": Read 553372
[20:04:40] goefile size: 0
[20:04:40] logfile size: 203301
[20:04:40] Leaving Run
[20:04:45] - Writing 22073749 bytes of core data to disk...
[20:04:47] ... Done.
[20:04:47] - Failed to delete work/wudata_03.sas
[20:04:47] - Failed to delete work/wudata_03.goe
[20:04:47] Warning: check for stray files
[20:04:47] - Shutting down core
[20:06:47]

DRUM ROLL PLEASE....

[20:06:47] Folding@home Core Shutdown: FINISHED_UNIT
[20:06:47]
[20:06:47] Folding@home Core Shutdown: FINISHED_UNIT
[20:06:51] CoreStatus = 64 (100)
[20:06:51] Sending work to server
[20:06:51] + Attempting to send results
[20:08:28] + Results successfully sent
[20:08:28] Thank you for your contribution to Folding@Home.
[20:08:29] You have been a good boy.
[20:08:29] You will receive 6 credits and a tasty biscuit.


Elapsed time: 29:17:49
Q6600 @ 3,294 MHz

[Sir_Loin]

LOL! Nice bit of editing there coo-coo :ROTF: You just made my week :D

Early impressions of both the P2662 and P2665 leave much to be desired. What did your system pull for the next WU? I'm curious if S is weaving a few in here and there, or if they're dishing out a steady diet to certain systems.

Fortunately, I have not received either yet, but I expect that will soon change. :(

[sc00p]

LOL! Nice bit of editing there coo-coo :ROTF:
+1 Good one coo_coo :D :up:


BTW: I haven't got those either... but maybe it's this gimmick which MikeB12 taught us... clean regularly work-folder and FAHlog.txt, queue.dat, unitinfo.txt -files from your f@h installation location.

[SparkyJJO]

ouch, I just got hit with this project. 700PPD from it compared to 1800PPD on the 2653.

[coo-coo-clocker]

Good to see I'm not alone :rolleyes:

My C2D machine just picked it up today. Good news there is Fahmon is saying it will only take 26 hours on that machine :hm:

[Shadowtester]

I hope that wu remains a windows thing I am quite content with the 2605's that my Linux clients have been getting :)

[Polizei]

I've got another hour on this hell WU.

[Polizei]

Woke up to find out that the WU completed and I have yet another 2665 WU.

facepalm.jpg

[Marvin_The_Martian]

I'm chewing the same stuff, we're all affected :)


EOC stats are down, have been all day :(

Happy folding :up:

[The_Barnacle]

Yup, I've been runnng through several of them...

[SparkyJJO]

yikes anyone notice that these units when finished are like 20MB uploads? :eek:

[brewer265]

these are killers , my overclocked QX9650 is down to 1420 ppd with 2665. 3000+ ppd with most normal projects.

[Polizei]

I can't say what my average ppd is because I've crashed a couple WUs in trying to overclock, but I do know that with 2665, it is substantially less than what my rig is capable of.

I noticed something too... I was talking with Sparks about f@h memory usage and we concluded that 256MB should be fine, 512MB is plenty. What's the dealio with this picture though? When working on non 2665 projects, I noticed that memory usage was below 200MB, but now it's closing in on 450MB?

http://i2.photobucket.com/albums/y26/toms918/Misc/2665sucks.jpg

[The_Barnacle]

Yeah, I noticed the result file it had to return was over 22MB as well...

[sc00p]

512MB is plenty. What's the dealio with this picture though? :confused: for example... and IIRC... p2653 takes ~65M per fahcore_a1... running then dual-SMP -setup it makes well over ~520Mb. Windows XP on my pc uses ~340M, so at minimum I would use 1 G. Preferrably 2 Gees on my case :p:
But yes... that project 2665 is something to :sofa:
...yet I still haven't got any taste of it :shrug: :flame: :D

delete files I and Mike have mentioned regularly and see if it makes any difference on wu assingments

[Polizei]

But yes... that project 2665 is something to :sofa:
...yet I still haven't got any taste of it :shrug: :flame: :D

Just finished my second 2665 WU, got a different project number now - 3064. Back to the normal amount of points range too... 1753.

[MikeB12]

have yall seen the drama between hardfolding and 7im unfolding recently...
apparently they started a complaint petition of some sort... from what I can tell the whole wu 2665 debate is what kicked it off... all the finger pointing and drama is not my thing, so I won't sign the petition. it makes no sense to me to come down on a mod unless they have proof, and then it's a behind the scenes deal with an admin or site owner, not a public petition... anyway, I'm not gonna sign it...

I'm not sure if this is just users getting their panties in a wad over disagreements or if the petition has some validity to it... but here's some of the links if anyone wants to read it and decide for yourself...
hardfolding thread http://www.hardforum.com/showthread.php?t=1305957
petition http://www.petitionspot.com/petitions/7im
hardfolding thread http://www.hardforum.com/showthread.php?t=1305394
fah thread http://foldingforum.org/viewtopic.php?f=16&t=2561


this whole thing stinks of internet drama to me...

[SparkyJJO]

Polizei I'd say that 512MB is more than plenty for standard client (heck even 256MB seems to do ok) but for the SMP I highly recommend 1GB just becaues the SMP can use more than the 512MB would allow for. Besides, ram is cheap.

[Polizei]

Polizei I'd say that 512MB is more than plenty for standard client (heck even 256MB seems to do ok) but for the SMP I highly recommend 1GB just becaues the SMP can use more than the 512MB would allow for. Besides, ram is cheap.

:am:

I thought we were talking about SMP, not the regualr one.

[Marvin_The_Martian]

have yall seen the drama between hardfolding and 7im unfolding recently...
apparently they started a complaint petition of some sort... from what I can tell the whole wu 2665 debate is what kicked it off... all the finger pointing and drama is not my thing, so I won't sign the petition. it makes no sense to me to come down on a mod unless they have proof, and then it's a behind the scenes deal with an admin or site owner, not a public petition... anyway, I'm not gonna sign it...

I'm not sure if this is just users getting their panties in a wad over disagreements or if the petition has some validity to it... but here's some of the links if anyone wants to read it and decide for yourself...
hardfolding thread http://www.hardforum.com/showthread.php?t=1305957
petition http://www.petitionspot.com/petitions/7im
hardfolding thread http://www.hardforum.com/showthread.php?t=1305394
fah thread http://foldingforum.org/viewtopic.php?f=16&t=2561


this whole thing stinks of internet drama to me...

I don't like drama any more then you do, but I think I do remember previous complaints about the guy. Then again, I never had a problem with him ( eg I am not signing ).

[The_Barnacle]

Heh, I think the whole thing is kind of amusing. If people honestly think their effort and resources are not appreciated becuase of the point system, there are plenty of other projects like WCG that they could throw their weight behind. I guess I could understand for those who spent considerable time and money acheiving the status they have, and they don't want to start over elsewhere. But still, the points aren't the main purpose of these projects.

[Kingcarcas]

Wow and i thought the [H] was the asshat of the internet :rofl:I've been visiting the foldingforum quite often recently and i can't say i've noticed anything. I'll b on the look out :confused:

LOL there are a bunch of babies over there, i spent 6K on my folding rig whaaaa. It is a Beta client and like i always say even if people don't want to, nobody is forcing you to fold :down: Spend 6K on a home theater or something....

[coo-coo-clocker]

Wow and i thought the [H] was the asshat of the internet :rofl:I've been visiting the foldingforum quite often recently and i can't say i've noticed anything. I'll b on the look out :confused:

LOL there are a bunch of babies over there, i spent 6K on my folding rig whaaaa. It is a Beta client and like i always say even if people don't want to, nobody is forcing you to fold :down: Spend 6K on a home theater or something....

@Kingcarcas: ++1

There are some real egos at play with that story there. On the other hand, I wouldn't be pissing off the biggest team of volunteers on the whole program... They are contributing monster numbers over there, and that's something that is important even if a few of them are too big for their britches.

[[XC] 4X4N]

You guys are missing the point. Yes, running F@H is for science, but the point system is a way to keep it interesting. A little competition is fun. If it were not for points, there would be a hell of lot less people running any d/c project. I care about the science, I wouldn't fold otherwise, but I also care about the points. All anybody is asking for a fair system. I have never like the way F@H awards points, but every time something is brought up about it, the project acts like they could care less.

[The_Barnacle]

4X4N;2994766']You guys are missing the point. Yes, running F@H is for science, but the point system is a way to keep it interesting. A little competition is fun. If it were not for points, there would be a hell of lot less people running any d/c project. I care about the science, I wouldn't fold otherwise, but I also care about the points. All anybody is asking for a fair system. I have never like the way F@H awards points, but every time something is brought up about it, the project acts like they could care less.
Yeah, I hear waht you're saying. It's not like it costs them anything to be consistant with the points they award. It doesn't look good when the orgnizers of the project just stick their nose up at the contributors. I wouldn't be surprised to see a lot of people at [H] switching over to WCG.

[The_Barnacle]

Hey some good news about project 2665...

http://www.hardforum.com/showpost.php?p=1032503176&postcount=24

[The_Barnacle]

My only question then is are they going to go back and award points for 2665 WUs that have already been completed?

[coo-coo-clocker]

4X4N;2994766']You guys are missing the point. Yes, running F@H is for science, but the point system is a way to keep it interesting. A little competition is fun. If it were not for points, there would be a hell of lot less people running any d/c project. I care about the science, I wouldn't fold otherwise, but I also care about the points. All anybody is asking for a fair system. I have never like the way F@H awards points, but every time something is brought up about it, the project acts like they could care less.

4x4, I respect your opinion. Mainly because you've been so obviously dedicated to the program; that and you're my teammate. I mean your point is good in that points make it more interesting and generate participation and competition. And I do think that Stanford could show the community more respect by putting more time into better client code, better communications with the teams, and a more equitable/logical points system. I know I fuss now and then, but I don't place all that much value on the points really. If a given WU takes more time to crunch and gives less points, it doesn't bother me much. Everyone is "penalized" equally so the competitive playing field is level so to speak. I just don't see what all the fuss is about. Maybe they should just give 1 point per WU (regardless of complexity) and call it day. But what do I know - compared to you guys I am a noob at this... :shrug:

[SparkyJJO]

4X4N;2994766']You guys are missing the point. Yes, running F@H is for science, but the point system is a way to keep it interesting. A little competition is fun. If it were not for points, there would be a hell of lot less people running any d/c project. I care about the science, I wouldn't fold otherwise, but I also care about the points. All anybody is asking for a fair system. I have never like the way F@H awards points, but every time something is brought up about it, the project acts like they could care less.

Just curious, what don't you like about the point system?

[coo-coo-clocker]

Hey some good news about project 2665...

http://www.hardforum.com/showpost.php?p=1032503176&postcount=24

Now that they've 'fixed' the points, all three of my clients are now running 2665 :lol2: So yeah, they are streaming these things out...

[SparkyJJO]

I just downloaded another 2653

[Polizei]

2653 here.

[sc00p]

Yeah... I don't see so big problem here with some particular project which yields lower ppd. Isn't that same, I mean makes similar effect for all of us when this kind of wu is out, for all the teams folding... not just [H]ard?????

In the end this pointage system is better than with wcg IMO.

[MikeB12]

restored balance it is, peaceful is the universe again.

[sc00p]

restored balance it is, peaceful is the universe again.

I feel your force here...

[Stijndp]

I think the points are about fair and couldn't think of much more effective systems. What bothers me most is that Panda almost gives the impression they don't give a **** about folding. It's annoying how long we are allready running SMP's in beta. On the ps3 it's getting annoying how the stat servers not only break down constantly, but also their assignment servers!

[sc00p]

When expressions, impressions, affections, expectations etc. (hours in a day) do not meet... oh boy, I wouldn't want to be in charge of this big project which f@h have became.

[The_Barnacle]

So it looks like my last P2665 was worth 1920 points :up:

[Stijndp]

But how can you justify assignment servers going down? If the tech guys didn't come a day earlier to reboot the servers than planned, they would have been 3days down.

Or like how long did we got stuck with the GPU client and it's hardware requirement. It's still not fixed with no way to run it on NVIDIA while that's what 40% of people are buying. AMD is only 25%.

[SparkyJJO]

But how can you justify assignment servers going down? If the tech guys didn't come a day earlier to reboot the servers than planned, they would have been 3days down.

I work in a data center and servers go down randomly. Just the other day we had some strange issue crop up where half of everyone couldn't log in that morning, while others were fine, and sometimes changing the network port they were on made it work :shrug:

Or like how long did we got stuck with the GPU client and it's hardware requirement. It's still not fixed with no way to run it on NVIDIA while that's what 40% of people are buying. AMD is only 25%.

The reason on the ATI vs nvidia GPU client is the ATI cards design lends them to folding much better than nvidia cards. During the X1900/7900 days the 7900 wasn't much better than a 3GHz P4, while the X1900 was several times better, so naturally they developed that one. Now, I don't know about the 8800 as much, it does lend itself to folding better than the previous generation but it is more difficult than the ATI cards IIRC. Something about their design. I wish I knew all the details but that's the gist of it.

[Stijndp]

I work in a data center and servers go down randomly. Just the other day we had some strange issue crop up where half of everyone couldn't log in that morning, while others were fine, and sometimes changing the network port they were on made it work :shrug:

But if your data center takes itself serieus, you'll have someone standby in the room next to it or at least avaible within the hour.

[SparkyJJO]

But if your data center takes itself serieus, you'll have someone standby in the room next to it or at least avaible within the hour.

Yes, true... however the data center at First Energy I have a feeling is higher priority than perhaps the F@H servers. I'm just saying maybe they have their servers a lower priority than say main school servers? I don't know :shrug:

For the couple years I've been doing this there have only been a few downs here and there so while it is a tad frustrating at times it really isn't a big deal. Not like it affects us too much.. well OK I take that back, some of us are rather addicted and totally freak out when the CPU usage is 0% :shocked: :ROTF:

[MikeB12]

well OK I take that back, some of us are rather addicted and totally freak out when the CPU usage is 0% :shocked: :ROTF:

/\
LOL
have you ever awoken to a power blink in the middle of the night, noticed the clock blinking.... then gotten up just to log the kids back in and start the clients back up. http://www.scootplace.net/smiles/original/116.gif

[SparkyJJO]

/\
LOL
have you ever awoken to a power blink in the middle of the night, noticed the clock blinking.... then gotten up just to log the kids back in and start the clients back up. http://www.scootplace.net/smiles/original/116.gif

Nope I always find out in the morning :(

Thankfully I don't usually have power flickers, and if I do at least my main PC in on a UPS :D

[Stijndp]

Good idea actually for a program. When cpu cycles go below 99%, the volume will be turned up to 100% and the sound of a fog horn starts playing.

[SparkyJJO]

bwahahaha :lol:

thing about that is you'd have to hear that every time it finishes a WU :eek:

[[XC] 4X4N]

Just curious, what don't you like about the point system?

It has always been a crapshoot as to what wu you will get. I haven't ran the regular client for a while, but I seem to remember that the amber units took twice as long as the gromacs units, and awarded less points. Same thing now with the smp client. If you get a 2653, it's like hitting the lottery :)

I think Stanford should benchmark whatever set-up they like, with each class of wu for how long it takes to complete it. Then adjust the points accordingly since it is basically cpu/gpu time that is donated. If the benchmark machine completes a certain wu in 10 hours, it should worth 'x' amount of points. If a different wu on the same benchmark machine takes 20 hours to complete, award 2 times 'x'. Seems easy enough to me.

By the way, how does Stanford figure out how many points to award?

[Polizei]

If I were creating the f@h project from the start, I would do something like 75 or 100pts per 1 hour it takes to complete a WU.

Makes sense to me... :shrug:

[SparkyJJO]

If I were creating the f@h project from the start, I would do something like 75 or 100pts per 1 hour it takes to complete a WU.

Makes sense to me... :shrug:

Problem is the huge disparity between CPU speeds.

[conzymaher]

If I were creating the f@h project from the start, I would do something like 75 or 100pts per 1 hour it takes to complete a WU.

Makes sense to me... :shrug:

That would reward people with slower hardware making a lesser contribution? :D

[Polizei]

Establish a bench on a "middle of the road" computer.

Just got a 1920pt 2665 project.

[SparkyJJO]

Establish a bench on a "middle of the road" computer.

Just got a 1920pt 2665 project.

that would work.

gahh 70 second rev limiter again!

[Polizei]

that would work.

gahh 70 second rev limiter again!

Too much horsepower. Give some to me. ;)

[Stijndp]

4x4 that is exactly the way they bench the points for a WU.

How do you decide how much credit a work unit is worth?

Points are determined by the performance of a given machine relative to a benchmark machine. Before putting out any new work unit, we benchmark it on a dedicated 2.8 GHz Pentium 4 computer with SSE2 disabled. (more specifically, as reported by /proc/cpuinfo on linux: vendor_id : GenuineIntel, cpu family : 15, model : 2, model name : Intel(R) Pentium(R) 4 CPU 2.80GHz, stepping : 9, cpu MHz : 2806.438, cache size : 512 KB). This machine runs linux, so all WUs are benchmarked with the linux core. Note: Currently the linux and Windows fah_cores run at the same speed, so this does not impact points performance.

We plug the results of this into the following formula:

points = 110 * (daysPerWU)

where daysPerWU is the number of days it took to complete the unit. This equation was chosen to match the points for previous Gromacs WUs from the previous point system. The upshot is that Tinker WUs will be worth more than before we set up the new point system (i.e. before April 2004).

As you probably allready understand, there is no standard setup to bench. You have a million combinations of architectures.
And for the people who don't understand :p : The Amber doesn't use SSE while gromac does use SSE instructions. They are both benched on a machine where SSE is disabled. That's why you get an insane amount of points if you run gromacs on your SSE machine compaired to an Amber. People with an (very) old pc would infact get the same ppd.
How do you decide the credit value of SMP work units?

Points are determined by the performance of a given machine relative to a benchmark machine, similar to the CPU client benchmark process. Before releasing any new project (series of work units), we benchmark it on a dedicated Macintosh Pro with 2 - 2.33 GHz Dual Core Xeon processors. (more specifically, 2 Woodcrest 5140 processors with 4 MB cache (each), 5 GB FBDIMM Memory (667 MHz DDR2), 1.33 GHz Bus)

We plug the results of this benchmark test into the following formula:

points = 1760 * (daysPerWU)

where daysPerWU is the number of days it took to complete the work unit.

Here you see why some projects hurt on our consumer pc's while they actually run normal like on their benchmark pc and it's for instance the 2653 that runs faster. So the 2653 is probably very L2/L3 cache dependend or it is also slown down by their FBDIMM's instead of our regular DDR/DDR2/DDR3.

[conzymaher]

Now that they re-assigned the points the stock q6600 in my server is doing a healthy 2400 ppd ( 1200ppd from two instances of the SMP client with the affinity changer service )

Stanford say they have optimized this project for better Quad support, I wonder would running a single SMP client without the affinity changer yield similar or even better results?

Has anyone tried yet? I might try when these WUs finish up....

[MikeB12]

I fnally picked up some 2665's and have a couple turning in ppd with the new points @1920.

some extended data on the pc's n my sig running dual smps. 380x9
Project : 2665
Core : SMP Gromacs
Frames : 100
Credit : 1920


-- X4-A-Vista-Q6600-3.4 ghz --

Min. Time / Frame : 15mn 23s - 1797.27 ppd
Avg. Time / Frame : 16mn 13s - 1704.91 ppd
Cur. Time / Frame : 16mn 56s - 1632.76 ppd
R3F. Time / Frame : 16mn 57s - 1631.15 ppd
Eff. Time / Frame : 16mn 32s - 1672.26 ppd


-- X6-B-XP-Q6600-3.4ghz --

Min. Time / Frame : 15mn 07s - 1828.97 ppd
Avg. Time / Frame : 16mn 03s - 1722.62 ppd
Cur. Time / Frame : 16mn 54s - 1635.98 ppd
R3F. Time / Frame : 16mn 52s - 1639.21 ppd
Eff. Time / Frame : 16mn 21s - 1691.01 ppd


-- X5-B-XP-Q6600 3.4ghz --

Min. Time / Frame : 16mn 15s - 1701.42 ppd
Avg. Time / Frame : 16mn 21s - 1691.01 ppd
Cur. Time / Frame : 16mn 32s - 1672.26 ppd
R3F. Time / Frame : 16mn 26s - 1682.43 ppd
Eff. Time / Frame : 17mn 00s - 1626.35 ppd

[sc00p]

Not so bad for me too (Q6600@fsb_1:1_401X9)...
:)

Project : 2665
Core : SMP Gromacs
Frames : 100
Credit : 1920


-- X_I --

Min. Time / Frame : 14mn 46s - 1872.33 ppd
Avg. Time / Frame : 14mn 46s - 1872.33 ppd
Cur. Time / Frame : 14mn 46s - 1872.33 ppd
R3F. Time / Frame : 14mn 46s - 1872.33 ppd
Eff. Time / Frame : 15mn 40s - 1764.77 ppd


-- X_II --

Min. Time / Frame : 14mn 44s - 1876.56 ppd
Avg. Time / Frame : 14mn 44s - 1876.56 ppd
Cur. Time / Frame : 14mn 45s - 1874.44 ppd
R3F. Time / Frame : 14mn 45s - 1874.44 ppd
Eff. Time / Frame : 15mn 20s - 1803.13 ppd

[SparkyJJO]

I only get ~1000PPD from it on a stock phenom :( The 2653 gets me 1800PPD.

My X2 CPU came so after this weekend I'm going to get that put in my PC and RMA this unstable phenom. Hopefully that won't take too terribly long and I'll have a stable chip that can hopefully OC as well.

[coo-coo-clocker]

I'm going to get that put in my PC and RMA this unstable phenom. Hopefully that won't take too terribly long and I'll have a stable chip that can hopefully OC as well.
If they will still take it back after all this time :rofl: :ROTF:

[The_Barnacle]

I only get ~1000PPD from it on a stock phenom :( The 2653 gets me 1800PPD.

My X2 CPU came so after this weekend I'm going to get that put in my PC and RMA this unstable phenom. Hopefully that won't take too terribly long and I'll have a stable chip that can hopefully OC as well.

Here's hoping for one of the new B3 steppings :up:

[Rmc[AZ]]

I have 4x 2665's going right now at a total of 6544 PPD. :( Even with the 1920 points I am down about 800-1000 PPD from the 2653's.

2x 2665 on a Q6600 @ 3.2
1x 2665 on a Q6600 @ Stock
1x 2665 on a E6600 @ 3.3

[Stijndp]

Imo 2665 are allright, they give enough points for an SMP. 3200ppd with your 3.2GHz is good for a quad. Since they fixed 2665's, I like all SMP core's again.

[MikeB12]

yeah, since they increased the wu points to 1920, these are on par with 30xx's.. so it's pretty fair.
3 Q6600's 3.42ghz, dual smps...
Project : 2665
Core : SMP Gromacs
Frames : 100
Credit : 1920


-- X4-A-Vista-Q6600-3.4 ghz --

Min. Time / Frame : 15mn 17s - 1809.03 ppd
Avg. Time / Frame : 15mn 46s - 1753.57 ppd
Cur. Time / Frame : 15mn 59s - 1729.80 ppd
R3F. Time / Frame : 16mn 04s - 1720.83 ppd
Eff. Time / Frame : 16mn 24s - 1685.85 ppd


-- X4-B-Vista-Q6600-3.4ghz --

Min. Time / Frame : 15mn 59s - 1729.80 ppd
Avg. Time / Frame : 16mn 01s - 1726.20 ppd
Cur. Time / Frame : 17mn 18s - 1598.15 ppd
R3F. Time / Frame : 17mn 00s - 1626.35 ppd
Eff. Time / Frame : 16mn 26s - 1682.43 ppd


-- X5-B-XP-Q6600 3.4ghz --

Min. Time / Frame : 15mn 41s - 1762.89 ppd
Avg. Time / Frame : 16mn 06s - 1717.27 ppd
Cur. Time / Frame : 15mn 50s - 1746.19 ppd
R3F. Time / Frame : 15mn 57s - 1733.42 ppd
Eff. Time / Frame : 16mn 08s - 1713.72 ppd


-- X5-A-XP-Q6600-3.4ghz --

Min. Time / Frame : 15mn 53s - 1740.69 ppd
Avg. Time / Frame : 16mn 06s - 1717.27 ppd
Cur. Time / Frame : 16mn 08s - 1713.72 ppd
R3F. Time / Frame : 16mn 07s - 1715.49 ppd
Eff. Time / Frame : 16mn 14s - 1703.16 ppd


-- X6-B-XP-Q6600-3.4ghz --

Min. Time / Frame : 15mn 19s - 1805.09 ppd
Avg. Time / Frame : 15mn 35s - 1774.20 ppd
Cur. Time / Frame : 15mn 48s - 1749.87 ppd
R3F. Time / Frame : 15mn 48s - 1749.87 ppd
Eff. Time / Frame : 16mn 00s - 1728.00 ppd


-- X6-A-XP-Q6600-3.4ghz --

Min. Time / Frame : 15mn 08s - 1826.96 ppd
Avg. Time / Frame : 15mn 34s - 1776.10 ppd
Cur. Time / Frame : 15mn 12s - 1818.95 ppd
R3F. Time / Frame : 15mn 10s - 1822.95 ppd
Eff. Time / Frame : 15mn 18s - 1807.06 ppd

[coo-coo-clocker]

OK!!!
Finally getting some NON-2665 WU's on my SMP rig!
Turnaround is much better and PPD is up by almost 2000. :clap:

[SparkyJJO]

I've been getting 3065 on my main rig recently. 2900PPD compared to 2000PPD on the 2665.

[MikeB12]

yeah me too. no 2665's atm... only 3062, 3064, 3065... the 3065 is the highest at 2144pointer and seems to be the lowest ppd producer for me.