Hi Folks,

let's have a different sort of comparison - credit races are so old hat ;o)

What's your best (lowest) DCF? You can find out by going to your host's details and scrolling down in the "Computers on this account" page linked in "Your Account" above.

My overall best DCF: 0.235826
MacMini G4 using Alex Kan's excellent optimized science app

My best DCF on PC: 0.361601
Pentium-D 805 @ 3.85 GHz using my P4-SSE3 optimized client (still getting lower slowly)

Obviously, this is heavily related to raw clock speed (i.e. a faster computer will have a lower DCF). Still, I'd like to see some comparisons from people using optimized and standard clients and differing hosts :o)

Regards,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Hi Folks,

let's have a different sort of comparison - credit races are so old hat ;o)

What's your best (lowest) DCF? You can find out by going to your host's details and scrolling down in the "Computers on this account" page linked in "Your Account" above.

My overall best DCF: 0.235826
MacMini G4 using Alex Kan's excellent optimized science app

My best DCF on PC: 0.361601
Pentium-D 805 @ 3.85 GHz using my P4-SSE3 optimized client (still getting lower slowly)

Obviously, this is heavily related to raw clock speed (i.e. a faster computer will have a lower DCF). Still, I'd like to see some comparisons from people using optimized and standard clients and differing hosts :o)

Regards,
Simon.

Worst = 38.63801
(My daughters CPU)

Best = 0.452821


What does this information tell us?

Ken

---

Hi Folks,

let's have a different sort of comparison - credit races are so old hat ;o)

What's your best (lowest) DCF? You can find out by going to your host's details and scrolling down in the "Computers on this account" page linked in "Your Account" above.

My overall best DCF: 0.235826
MacMini G4 using Alex Kan's excellent optimized science app

My best DCF on PC: 0.361601
Pentium-D 805 @ 3.85 GHz using my P4-SSE3 optimized client (still getting lower slowly)

Obviously, this is heavily related to raw clock speed (i.e. a faster computer will have a lower DCF). Still, I'd like to see some comparisons from people using optimized and standard clients and differing hosts :o)

Regards,
Simon.

Here's all the CPU's that I have that list it. Some are running older clients

0.641183
0.536872
0.590055
0.523544
0.452821
0.540374
0.486078
1.048731
0.750295
0.999379
1.651578
1.340696
0.869975
1.02241
0.870178
0.698181
1.097006
0.615779
37.610995
0.786634
1.272358
1.259785
0.800684
0.977459
1.002171
38.63801
0.538293
0.964803
0.633295
0.811982

---

I've only got one, for a 2.6GHz P4 HT PC running your P4 SSE2 v1.3 client. The DCF is not really stable (yet). I just did a manual update to see how much it would change, and it changed from 0.485556 to 0.56794 (just from one update).

I just had a few very short WUs (done in an hour and a half). And weirdly enough, for example this one seems to have claimed a different amount of credit than another cruncher (maybe due to a different boinc core client 4.45 vs 5.49: does the 4.45 still claim credit according to computation time and benchmark results?). I've just installed BoincLogX, so I should be able to keep better track in the future.

---

Hi Folks,

let's have a different sort of comparison - credit races are so old hat ;o)

What's your best (lowest) DCF? You can find out by going to your host's details and scrolling down in the "Computers on this account" page linked in "Your Account" above.

My overall best DCF: 0.235826
MacMini G4 using Alex Kan's excellent optimized science app

My best DCF on PC: 0.361601
Pentium-D 805 @ 3.85 GHz using my P4-SSE3 optimized client (still getting lower slowly)

Obviously, this is heavily related to raw clock speed (i.e. a faster computer will have a lower DCF). Still, I'd like to see some comparisons from people using optimized and standard clients and differing hosts :o)

Regards,
Simon.

My 1.4 GHz Pentium-m is currently at 0.368411 using your Windows P-m build. It's probably still declining, too,

Raw clock speed probably has little to do with the value, it is basically a ratio between the Whetstone benchmark and real crunching. L1 and L2 cache sizes and memory bandwidth are probably more significant.

These WUs from 2005 seem mostly in the "normal" range so making these comparisons now is OK. But as soon as some VLAR or VHAR work comes along DCF will probably become irrational again; the splitter estimate curve doesn't match reality for any system I've seen.

                                                          Joe

Yes Bart,

there are still lots of people out there with 4.xx BOINC clients, and those claim credit based on benchmarks, not flops.

Therefore, it can happen that there are two using old clients who claim a wholly different amount of credit, but make up 2/3 of the quorum. So you might end up getting less credit for a WU than originally intended, but that will change over time as more people update.

The DCF can go upwards very quickly, and downwards only slowly. What the number tells you is this:

The reference computer has a DCF of 1. When your computer is faster than the reference computer in processing a WU, its DCF gets lowered. This takes a while to stabilize, 20-30 WUs is probably a minimum figure.

Shorter WUs have less speedup with optimized apps(meaning the default app is comparatively faster on those than on ones with longer processing times), so the DCF may go up a bit with those, but will continue its downward trend with longer WUs you will eventually get.

So - the lower your DCF is, the quicker your computer can crunch WUs, basically.

HTH,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Two very different types:

Pentium M 1.9GHz = 0.170871
iMac G4 1GHz = 0.287163

Both run optimized workers.

QS

---

All computers start, or after reset, with DCF of 1. It is primarily used to correct the predicted crunch times, which are initially worked out on the benchmarks.
The benchmarks for most computers only measure the clock speed and efficiency of the core of the cpu. The dcf helps to correct differnces caused by all the memory differences, larger/smaller cache memory, main memory bandwidth and speed.

The DCF goes down by 10% of the difference between predicted and actual crunch time, but up to the actual crunch time immediately if it is longer than predicted. It take about 20 to 25 units to be crunched for the DCF to stabalise when calculated with similar work units.

On Seti, at the moment, with the non-linear ar/time graph, which is also affecting credits, the predicted estimates are not accurate so it can suddenly rise if you get a unit that crunches slower than expected on the computer.

Oh, and at the moment my Pent M DCF is at 0.45 using the appropriate optimised app, thanks KWSN. Before enhanced using crunch3r's app it was stable at 0.13 +/- 0.01. My P3 using the standard app on Beta is at 0.87nnn.

Andy

Pentium-M based processors seem pretty good at Seti@Home (probably due to their large caches, in part). Core Solo/Duo and especially Core 2 will be even quicker.

The reason I asked for DCFs is because I want to have a relatively comparable performance metric for different PCs - so if you post your DCFs, please also post some details about the machines :o) Oh, and please only post your quickest (one or few).

Thanks for the explanation about how DCF works, WinterKnight :o)

Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Pentium-M based processors seem pretty good at Seti@Home (probably due to their large caches, in part). Core Solo/Duo and especially Core 2 will be even quicker.

The reason I asked for DCFs is because I want to have a relatively comparable performance metric for different PCs - so if you post your DCFs, please also post some details about the machines :o) Oh, and please only post your quickest (one or few).

Thanks for the explanation about how DCF works, WinterKnight :o)

Simon.

My Opteron 175 has Result duration correction factor 0.455042

---

Official Abuser of Boinc Buttons...
And no good credit hound!

Pentium-M based processors seem pretty good at Seti@Home (probably due to their large caches, in part). Core Solo/Duo and especially Core 2 will be even quicker.

The reason I asked for DCFs is because I want to have a relatively comparable performance metric for different PCs - so if you post your DCFs, please also post some details about the machines :o) Oh, and please only post your quickest (one or few).

Thanks for the explanation about how DCF works, WinterKnight :o)

Simon.

My Pentium 4 (530) has a DCF of 0.337335. Using optimized Enhanced.

Both running Optimized versions.

AMD 64 3200+ = 0.438932

Intel P4 D 3.0ghz = 0.731988

Cannot get model number on the intel because CPU-Z keep erroring.

---

Theres a Fine Line between Hobby and Mental Illness.

2.6 HT 5.12 0.54896
2.6 HT 5.15 0.695942
797 P3 5.12 0.642602
1.7 P4 5.12 0.394486
1.7 P4 5.12 0.421082
1.5 P4 5.12 0.49797

---

Celeron(R) CPU 2.80GHz - (SSE3 client - 1GB) 0.565627
Pentium(R) D CPU 2.66GHz - (non optimized - 512MB) 0.832361
Athlon(tm) XP 2600+ - (SSE2 client - 256MB) 1.022724

All running 5.4.9 BOINC manager

---

Don't know my best but my current DCF is 0.399591, and dropping.
Pentium 4 HT Prescott @3.30 Ghz, using da chicken's SSE3 app. ;)
Boinc mgr 5.4.9.

---

/Mav

We have lingered long enough on the shores of the cosmic ocean.
We are ready at last to set sail for the stars.
(Carl Sagan)

iMac Intel Core Duo 2.00G. Stock 5.13 App. DCF=0.319593

Simon just switched to your app from crunch3r's the other day.

Good work!

Average CPU efficiency 0.999532
Result duration correction factor 0.706283

---

My result duration correction factors:

Joker2 Opteron 165 @ 2.63GHz [MS6702E v1.0] XP x32 1.037531
Joker3 Opteron 270 @ 2.36GHz [K8N-DL v1.03] XP x64 0.926109


So far this is It, As I only have two cpus right now.

5.4.9 Boinc install, 5.5.0 client and 5.15 SSE2 kwsn optimized app on both PCs.

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

using chickens opt client on both

amd 4200 x2, 0.518
intel 3GHz HT, 0.333


interesting that the amd is alot higher...also, the benchs for the amd are alot lower on chicken app, than previously..


Cheers.

---

using chickens opt client on both

amd 4200 x2, 0.518
intel 3GHz HT, 0.333


interesting that the amd is alot higher...also, the benchs for the amd are alot lower on chicken app, than previously..


Cheers.

Benchmarks are run by BOINC not application. So make sure min other programs are running and do them manually. If they are still down something has changed on your 'puter and needs investigating/fixing.

Andy