Boinc requests tasks based on the estimated finish time of existing work and other things such as your cache size. Boinc has to calculate how long a task will take and get a general idea of how long tasks are for each project. Until your card runs a task Boinc cannot tell how long it will take.

If GPUGrid implemented a strategy of only allocating tasks on the basis of the GFlops peak reported by Boinc then some Fermi's might not get tasks because not all Boinc versions report the performance correctly.

Just to follow up on this, I've logged my results for a while. On analysis it turns our the TONI WUs are short (12-13 Hours on a GTX 260) to be considered LONG_WUs but are claiming credit in line with what LONG_WUs should be getting(approx 35% higher than normal units) while the IBUCH WUs are in line with the quoted time for long WUs (16-19H on a GTX 260) but are claiming the same amount of points as their normal sized counterparts.

Type...............Total time..........Claimed/Total
TONI_AGG.....43,625.77..........0.81
TONI_AB1......46,400.62..........0.83
IBUCH_6........69,739.29..........0.61
IBUCH_7........62,840.91..........0.62
IBUCH_8........56,076.11..........0.63
IBUCH_9........70,151.36..........0.60
IBUCH_10......60,062.49..........0.61

Just thought I'd point this out.

i notice them failing half way thu and some normal wu also.

i notice them failing half way thu and some normal wu also.

Do you have your card overclocked?

What card is it?

What error is it giving you?

NVIDIA GeForce GTS 250 (511MB), Clock rate: 1.84 GHz, driver: 27061
Windows XP Home x86 Edition, Service Pack 3, 2GB RAM
Athlon 64 X2 Dual Core Processor 4400+

SWAN: FATAL : swanBindToTexture1D failed -- texture not found

ERROR: file c:\cygwin\home\speechserver\gpumd2_c\src\pme\CPME_cufft.cpp line 73: cufftExecC2C (gridcalc2.1)
called boinc_finish

SWAN: FATAL : swanMemcpyDtoH sync failed

ERROR: file c:\cygwin\home\speechserver\gpumd2_c\src\pme\CPME_cufft.cpp line 73: cufftExecC2C (gridcalc2.1)
called boinc_finish

ERROR: file c:\cygwin\home\speechserver\gpumd2_c\src\pme\CPME_cufft.cpp line 73: cufftExecC2C (gridcalc2.1)
called boinc_finish

So, lots of cuda fft errors and a few others on a GTS250. This is about power for the course for that GPU.
You probably have a better chance of finishing the shorter tasks, but most GTS250’s are just not good to crunch with here. You might want to plan for a replacement.

How much longer do you expect a gtx 295 to be useful?

One of the researchers might have a better idea, but I would guess it would be useful up until about 2 months after the next generation of NVidia cards. So I would say it will be next year some time before it's obsolete. Obviously this depends on NVidia's Kepler release (i'm guessing it will be in the first half of next year before you can buy one).

To some extent the CC1.3 cards are naturally fading from the project; a GTX 295 is now 2years and 4months old (quite old for a GPU). As you would expect, many have upgraded to newer cards, and this is likely to continue. A couple of months after the next generation of NVidia cards appear, very few users will still be crunching with a CC1.3 card; Kepler should eventually double the performance of comparable Fermi cards, never mind the CC1.3 cards.

I am still running a GTX 285 on a Windows XP machine. It still runs well, after almost 2 years, and is able to finish even long tasks well within 24 hours. It is only about 20% to 30% slower than the GTX 480 running on Windows 7. I am looking to upgrade it, but I do want to wait until the prices for the 500's drop.

You mentioned the Kelper cards. Will they be compatible older machines, 4 to 5 years old? My video cards tend to last about 3 years, but my desk top computers last a lot longer. The last machine, I junked, was 10 years old!

Even if the Kepler cards are up to PCIE3 standards, they should be backward compatible for older PCIE2 and probably PCIE boards, albeit at some loss (gaming, video editing) - 50% bandwidth. For crunching the bandwidth restrictions would be less noticeable.

Different generations are hard to compare, but I would equate a GTX285 with a GTX580, which offers a bit more than a GTX480 (more equivalent to a GTX280). Even now software for Fermi cards is still developing. So the performance should improve again slightly. Win7 is at least 11% slower than XP. When the Kepler cards turn up they will probably triple the performance of their equivalent GTX 200 card. So until that happens a high end GTX 200 card will remain a good card.

I would expect that a GPUs within 1/10 of the power of the top one will be useful. If kepler will have 1024 cores, then anything above 128 cores should be fine for the short workunits (acemd2). For acemd_long workunits 1/3 of the top card is probably acceptable, so over 240-300 cores once kepler is out.

gdf

My latest returned and validated/creditted Long-Run WU took :

Task ID      WU ID       Date and Time sent   and   returned
4015690 	2494307 	23 May 2011 10:23:09 UTC 	24 May 2011 5:58:53 UTC

                                                 Elapsed      CPU          Claimed     Granted
 	Completed and validated 	23,012.34 	2,414.91 	35,067.36 	52,601.04 	Long runs (8-12 hours on fastest card) v6.13 (cuda31) 

Ran on a GTX480 (1.5GHz) and X9650 (3.5GHz) and this is also the highest speed
for GPU and CPU, in order to get valid results.
I like them and search for Alzheimers and what and how to deal with this terrifying disease is badly needed, also finding cause and treatment!

---

Knight Who Says Ni N!

Hi,

I just returned Taskid = 4017232 Wuid = 2495161.

This wu processed more then 40 hrs on a gtx480 card.

This is tooooooooooooooo much?? Correct??

---

Ton (ftpd) Netherlands

SALVE VOLEVO CHIDERE SUL FORUM COME MAI LE ACEMLONG HANNO UNA SCADENZA COSI BREVE NON RIESCO A COMPLETARLE PER UNO SCARTO DI CIRCA 6 ORE VENGONO CONSEGNATE IN RITARDO ORA HO DISABILITATO LE WU LUNGA CORSA LASCIANDO SOLO ACEM 2 SU ELABORAZIONE MIA NON POSSO SUPERARA 600 GFLOP DATI.

Quirino Cuccioli, per i compiti a lungo la GT 220 non è abbastanza veloce.

en. The GT220 is not fast enough for long tasks.

Ton: some other guy got 24.417 ms per timestep (roughly double the speed) on a 470 for a similar task (result 4034521). Don't know what happened but should be transient.

GRAZIE PER LA RISPOSTA INFATTI ERO CONVINTO CHE LA CAUSA ERA QUESTA E MI CHIEDO POICHE' IL COMPUTER IN QUESTIONE NE MONTA 3 DI ENGT 220 CON 1 GB PER OGNI SCHEDA E SONO IN MOD SLY CON PROCESSORE SU SCHEDA MADRE MAXIMU III INTEL I 5 CON 4 GB RAM ALIM CIRCA 1000 W COME MAI MI SVOLGE 3 COMPITI CONTEMPORANEAMENTE INVECE DI DARE SINGOLO LAVORO PER MAGGIORE VELOCITA QUANDO PARTE APPLICAZIONE ACEM 2 VISUALIZZO + 0,08 CPUS
GRAZIE PER ATTENZ.

Buona domanda. E 'possibile, ma solo in laboratorio. Il problema è che BOINC è limitato.

En. Just saying that we cannot run one app over 3 GPU's due to the limitations of Boinc, but it is possible in the labs. Pity, 3 GT220's might do well enough (144 shaders combined), ditto for the GT240 and similar cards.

GRAZIE PER IL CHIARIMENTO RIGUARDO ALLA MODALITA' SLY SU BOINC QUINDI NON SI PUO' SCARICARE NESSUN PROGRAMMA CHE FACCIA CIO',GIUSTO.
OK GRAZIE E CIAO