I was perusing the Rosetta web site yesterday and was looking at the computers that some of the top performers were using to crunch. Similar hardware to mine, but they are getting WAY more points per day average than I am on similar hardware, three times as much in some instances. I saw one box this guy had, an Athlon 64 3000, that was getting over 600 points per day!

HOW ARE THEY DOING THAT?

Are there any tuning tips availabe that anyone has seen for tuning the client? I would love to triple my production without tripling my electricity bill!!!

I have included a link to show what I was talking about. This is not meant as an accusation of cheating, the results are just so totally different from what I am getting that I am puzzled big time.

https://boinc.bakerlab.org/rosetta/hosts_user.php?userid=43013

Look at computer ID 169797 or 169025.

This dude has a Sempron 2800 running Windows XP averaging 644 ppd and an Athlon 64 3000 averaging over 700 ppd!!!!

---

Are you running yours 24/7?

---

Are you running yours 24/7?

Yes, I am. All of them run 24/7. The one with the worst comparison ratio is my Athlon 64 3000. It is a custom built gaming machine with 2 GB of ram and 2 300 GB raided drives for 600 GB disk space. It has a high end ATI 256 MB Video Card. It does nothing but Rosetta for prob 22 hours a day and I use it for a couple of hours most evenings.

I was looking at Recent Average Credit, he had s similar computer (not as much stuff in it) that was getting over 700 points a day! I am getting between 200 and 250 a day on mine. We are both running Windows XP SP2.

Is there an optimized client that could provide this much difference?

---

I was perusing the Rosetta web site yesterday and was looking at the computers that some of the top performers were using to crunch. Similar hardware to mine, but they are getting WAY more points per day average than I am on similar hardware, three times as much in some instances. I saw one box this guy had, an Athlon 64 3000, that was getting over 600 points per day!

HOW ARE THEY DOING THAT?

The Rosetta project gives you whatever amount of credit your boinc client asks for as long as the WU completes successfully. There are boinc clients (called "optimized" clients) that will ask for several times as much credit as the recomended client. Using such a client has no effect on how fast a machine crunches a WU, it simply asks for more credit for that work. And with rosetta, it gets whatever it asks for.

The software does'nt always report the CPU correctly. I have an AMD64 @2.6Ghz and it is reported as a Sempron 3100+ which can be either 1.8Ghz or 2.0Ghz if I remember it right. So if you are seeing a Sempron 2800+, it might actually be something better. My intels report fine though.

---

I have seen this also, but to have that much discrepancy it would have to be a Dual Core X2 4800 (or similar) to generate 700+ points per day.

---

I have seen this also, but to have that much discrepancy it would have to be a Dual Core X2 4800 (or similar) to generate 700+ points per day.

You're mixing things up... not suprisingly, because RAC is a very strange way of indicating "something" about a pc's output.

If you upload your results say once a week and not connect with the server inbetween your RAC keeps on growing skyhigh... if you connect a few times a day with the same pc, your RAC grows slowly and not that high.

So don't be puzzled... there isn't a trick, he is probablity just uploading a lot's of wu's at one time and never connecting inbetween.

More info and discussion about the pro's and con's of RAC click

And if you're interessed the real average daily output of those pc's click

---

So don't be puzzled... there isn't a trick, he is just uploading a lot's of wu's at one time and never connecting inbetween.

There's a bit more to it than just that. Just as an example, one of the systems mentioned (this one) is reporting its work every day, and is getting 60+ credits for 2 hours of work. The data being returned does not show any of the common optimized clients is being used (but he certainly could have optimized his own client). What grabs me as out of place, though, is the benchmarks:

Measured floating point speed 3705.26 million ops/sec
Measured integer speed 10957.09 million ops/sec

and it is identified as:

AuthenticAMD
AMD Sempron(tm) Processor 2800+

Now, admittedly, I don't know a lot about what the normal benchmarks for the various AMD processors should be, but come on - integer speed of almost 11000.

---

So don't be puzzled... there isn't a trick, he is just uploading a lot's of wu's at one time and never connecting inbetween.

There's a bit more to it than just that. Just as an example, one of the systems mentioned (this one) is reporting its work every day, and is getting 60+ credits for 2 hours of work. The data being returned does not show any of the common optimized clients is being used (but he certainly could have optimized his own client). What grabs me as out of place, though, is the benchmarks:

Measured floating point speed 3705.26 million ops/sec
Measured integer speed 10957.09 million ops/sec

and it is identified as:

AuthenticAMD
AMD Sempron(tm) Processor 2800+

Now, admittedly, I don't know a lot about what the normal benchmarks for the various AMD processors should be, but come on - integer speed of almost 11000.

It's an awful lot...
but he is a member of XtremeSystem and when you visit their forum you'll see that their AMD Sempron(tm) Processor 2800+ aren't the same as mine. Just a quote:

My sempy 2800+ @ 2.7ghz is running:

3747 Whetstone
11355 Drystone

---

I think that I am beginning to understand because of how RAC works. It is a bit confusing, but it is still a good measure of a longer period. Mine does not seem to fluctuate nearly so much, prob because of how my prefs have been set.

However following the threads on their forum has led me to seriuosly wonder if someone has found a way to cheat. They have some very interesting threads in their public forum that were moved to their private forum just because of this topic.

---

Can someone explain to me how this guy gets around 120 pints per work unit completed and I am getting 40-60?

https://boinc.bakerlab.org/rosetta/results.php?hostid=97215&offset=20

I can not find anyway to set the prefs to get higher point work units, but he must have. He is getting 2500 or so points per day from ONE computer and has been for weeks.

---

He even has some work units that he got 500 points for! I want all of that kind!

---

As I continue to research, I find more and more stuff that pisses me off. Thsi cruncher is an example:

https://boinc.bakerlab.org/rosetta/results.php?hostid=160017&offset=140

On March at 11:38 his computer turned in results for a HBLR_1.0_1dtj_348_5529_0 and claimed 56 points for that work unit with a little over 7000 seconds CPU time to complete.

Almost exactly 6 hours later his computer turned in results for a HBLR_1.0_1mky_348_5534_0 and claimed 228 points for that work unit with a little over 29000 seconds CPU time to complete.

Every work unit since then has been in the 200+ point category. Something smells fishy. How did he suddenly start getting only 200+ point work units?

How are points awarded? This seems to be the major issue.

---

As I continue to research, I find more and more stuff that pisses me off. Thsi cruncher is an example:

https://boinc.bakerlab.org/rosetta/results.php?hostid=160017&offset=140

On March at 11:38 his computer turned in results for a HBLR_1.0_1dtj_348_5529_0 and claimed 56 points for that work unit with a little over 7000 seconds CPU time to complete.

Almost exactly 6 hours later his computer turned in results for a HBLR_1.0_1mky_348_5534_0 and claimed 228 points for that work unit with a little over 29000 seconds CPU time to complete.

Every work unit since then has been in the 200+ point category. Something smells fishy. How did he suddenly start getting only 200+ point work units?

How are points awarded? This seems to be the major issue.

CPU time . benchmark = points claimed (granted)

I think it have to do with optimized clients. I just installed Mikkyo's superbench client on a dual G5 and the benchmark almost doubled to 5284.7 / 17614.79 million ops/sec. That means that I will get twice as much points for the same CPU time.

(Scrolling thru the Number crunching forum shows a lot of topics about the use of optimized clients / cheating / inflating the benchmarks).

---

On March at 11:38 his computer turned in results for a HBLR_1.0_1dtj_348_5529_0 and claimed 56 points for that work unit with a little over 7000 seconds CPU time to complete.

Almost exactly 6 hours later his computer turned in results for a HBLR_1.0_1mky_348_5534_0 and claimed 228 points for that work unit with a little over 29000 seconds CPU time to complete.

Every work unit since then has been in the 200+ point category. Something smells fishy. How did he suddenly start getting only 200+ point work units?
How are points awarded? This seems to be the major issue.

I don't see anything wrong with it.
Your credits are calculated as:
Credits = CPUbenchmark * number_of_hours

The second job, you mention, last 4 times more than the first one. So it's correct that the 2nd job gets 4 times more credits than the first one...

---

Member of Dutch Power Cows

However following the threads on their forum has led me to seriuosly wonder if someone has found a way to cheat. They have some very interesting threads in their public forum that were moved to their private forum just because of this topic.

I read a lot of their public forums yesterday and thought some of that was very strange, too. They seem to have a lot of threads that get made private so people won't see them and think they're cheating. Well, logic sort of says that's backwards - hiding them makes it look like they are cheating, whereas if they really weren't cheating and left them public, people would see it all and know there was no cheating.

I also thought it was interesting that they don't seem to advocate participating in projects that require multiple crunchers to form a quorum (not that they don't, but I couldn't find any mention of it if they do), thus doing away with any ability to compare their crunchers side by side with other crunchers.

Since rosetta doesn't require multiple crunchers to form a quorum, they really need to come up with some kind of a way to introduce a sanity check on the participating computers. Perhaps set a maximum deviation from the average benchmark for each cpu. Anyone who understands the code and compiles their own boinc software can screw with the way the benchmarks are determined as well as the amount of credit the computer asks for, but on other projects the quorum method of granting credit kills off any real benefit of doing that. Since that's not logical on rosetta, maybe they need to devise a script that runs every so often looking for out-of-whack benchmarks or credit claims based on the machine's cpu and then resets those benchmarks to some determined maximum allowable deviation for that particular cpu and recalculates the credit granted to those hosts with some real-world benchmarks applied.

---

As I continue to research, I find more and more stuff that ****. Thsi cruncher is an example:

https://boinc.bakerlab.org/rosetta/results.php?hostid=160017&offset=140

On March at 11:38 his computer turned in results for a HBLR_1.0_1dtj_348_5529_0 and claimed 56 points for that work unit with a little over 7000 seconds CPU time to complete.

Almost exactly 6 hours later his computer turned in results for a HBLR_1.0_1mky_348_5534_0 and claimed 228 points for that work unit with a little over 29000 seconds CPU time to complete.

Every work unit since then has been in the 200+ point category. Something smells fishy. How did he suddenly start getting only 200+ point work units?

I think you need to look at his results a bit more carefully.

The 200 point units took 4 times longer than the 50 point units - Simple arithmetic.

There are clearly many machines that are dirty, but I happen to know this machine is clean and probably one of the highest (if not the highest) producer. Be careful whom you tar with that brush my friend.

---

As I continue to research, I find more and more stuff that ****. Thsi cruncher is an example:

https://boinc.bakerlab.org/rosetta/results.php?hostid=160017&offset=140

On March at 11:38 his computer turned in results for a HBLR_1.0_1dtj_348_5529_0 and claimed 56 points for that work unit with a little over 7000 seconds CPU time to complete.

Almost exactly 6 hours later his computer turned in results for a HBLR_1.0_1mky_348_5534_0 and claimed 228 points for that work unit with a little over 29000 seconds CPU time to complete.

Every work unit since then has been in the 200+ point category. Something smells fishy. How did he suddenly start getting only 200+ point work units?

I think you need to look at his results a bit more carefully.

The 200 point units took 4 times longer than the 50 point units - Simple arithmetic.

There are clearly many machines that are dirty, but I happen to know this machine is clean and probably one of the highest (if not the highest) producer. Be careful whom you tar with that brush my friend.

that happens to be my machine. It is a dual socket Opteron 280 which makes it a 4 way box. Now when a 2.8 p4 reports a higher RAC than a 64 bit AMD with hypertransport memory access then I would through stones at the P4. Now do you think that 8 hours of CPU time on this machine is not worth 220 points then you just have the admins say the word and I will be gone! This machine is clean.

---

that happens to be my machine. It is a dual socket Opteron 280 which makes it a 4 way box. Now when a 2.8 p4 reports a higher RAC than a 64 bit AMD with hypertransport memory access then I would through stones at the P4. Now do you think that 8 hours of CPU time on this machine is not worth 220 points then you just have the admins say the word and I will be gone! This machine is clean.

Not to speak for him, but I think maybe the original poster just overlooked the changes in run time when he was looking at your system.

As for me, I don't know much about AMD processors (or any processor for that matter) so I admit I go out on a limb when I decide to throw stones at them. But since I've jumped in and been pretty vocal in this thread, I will take a moment to elaborate before I wind up getting attacked. When your system first got mentioned, I looked up the processor and didn't think anything about it claiming that much. It's a $1000 quad processor - no doubt it can do a lot of work in a very short time. But that still brings me back to one of the systems mentioned in the original post. That one is a Sempron 2800+. That's a $75 processor and it's claiming higher benchmarks than your $1000 processor.

That's the kind of stuff that causes me to throw stones.

---

that happens to be my machine. It is a dual socket Opteron 280 which makes it a 4 way box. Now when a 2.8 p4 reports a higher RAC than a 64 bit AMD with hypertransport memory access then I would through stones at the P4. Now do you think that 8 hours of CPU time on this machine is not worth 220 points then you just have the admins say the word and I will be gone! This machine is clean.

Not to speak for him, but I think maybe the original poster just overlooked the changes in run time when he was looking at your system.

As for me, I don't know much about AMD processors (or any processor for that matter) so I admit I go out on a limb when I decide to throw stones at them. But since I've jumped in and been pretty vocal in this thread, I will take a moment to elaborate before I wind up getting attacked. When your system first got mentioned, I looked up the processor and didn't think anything about it claiming that much. It's a $1000 quad processor - no doubt it can do a lot of work in a very short time. But that still brings me back to one of the systems mentioned in the original post. That one is a Sempron 2800+. That's a $75 processor and it's claiming higher benchmarks than your $1000 processor.

That's the kind of stuff that causes me to throw stones.

I have to agree with that for sure. and the quad did cost me a bit more than a grand! ;) But I like them so well I have 3, and the TOC is where that comes into play, cost lots less to operate a 4 way than 4 single CPU machines. I think the sweet spot today migh be the Opteron 170's but to each his own.

---