Carlitos714
FTW Member
- Total Posts : 1126
- Reward points : 0
- Joined: 2009/03/01 02:36:45
- Location: San Diego, CA
- Status: offline
- Ribbons : 3

Re:Some Early GT430 Results
2010/10/17 20:45:42
(permalink)
[HTPC] Zotac Zbox Nano AD12, 4 GB Corsair 7-7-7-20, Intel X25-M 120 GB, 55 inch Samsung plasma [GAMING RIG/DAILY DRIVER] i7-3820 @ 4.85 Ghz, GA-X79-UDR3, Heatkiller 3.0, HWL BI GTX 480, MCP 350 x 2 , XSPC restop, EVGA GTX 680 SLI, HWL BI GTX 360, 16 GB Doms, Crucial M4 256 gb, 120 GB Torqx SSD, DELL PERC 5I W/ HD204UI 2TB x 6 RAID 5, Xonar DX, Corsair AX1200i, WinTV-HVR-1250 XSPC H2 [Peripherals] Ducky Dragon| MX518| Dell 2408WFP| Fiio E9 Amp| BeyerD DT-880 Pro 250 ohms
|
AlHo
Superclocked Member
- Total Posts : 142
- Reward points : 0
- Joined: 2010/06/20 09:11:30
- Location: Whitley Bay, UK
- Status: offline
- Ribbons : 0

Re:Some Early GT430 Results
2010/10/18 04:44:32
(permalink)
Those PPD look great for that price and power draw. I currently get a max of about 6300 PPD with my GTS250 which must draw about 150W. Have finally got enough together to get a GTX460 1GB, so was thinking a GT430 might be great as a physX/folding card. I was thinking of keeping the GTS250 but the 430 wins for PPD/Watt, how would the two compare for physX? I just dont want to take my PSU too close to its limit (also I only have 2 pcie connectors) so a nice 60W card would be great with a GTX460. If I put the physX/folding card in the PCIEx4.0 slot will it loose any performance for either function or does it need to go in the x8.0 slot, thus reducing the bandwith for my main card from 16 to 8. I know its not a real world problem with a GTX460 but I would also get better cooling by having it in the 4.0 slot.
post edited by Al_Ho - 2010/10/18 04:46:15
|
Re:Some Early GT430 Results
2010/10/18 04:51:59
(permalink)
Any PhysX performance questions may be better asked in the 400 series forum. I don't have any info on it since I have only been testing folding. Out of curiosity, what would be the best way to test the PhysX performance? If I get a chance, maybe not until next weekend, I may be able to run a comparison between a couple of different cards that I have.
|
Re:Some Early GT430 Results
2010/10/18 10:10:26
(permalink)
rjbelans Out of curiosity, what would be the best way to test the PhysX performance? If I get a chance, maybe not until next weekend, I may be able to run a comparison between a couple of different cards that I have. Try FluidMark. It's a small PhysX benchmark you can get here. Try running it with 3 emitters, Multi-Core PhysX, and 60000 particles (which should be default). Also, change resolution to 640x480. Just post the results here, and we can do some comparisons. I'll run it on my 470 and 220 in just a sec. I was just surprised that the 430 did so well in F@h, I wouldn't be too surprised if it became the "holy grail" of PhysX cards!
post edited by Nahte27 - 2010/10/18 10:12:12
|
Re:Some Early GT430 Results
2010/10/18 10:18:14
(permalink)
OK. Thanks Nahte27. Like I said though, it will not be until the weekend. I'm in a race to 30 mil with Punchy and can't let up at all.
|
Re:Some Early GT430 Results
2010/10/18 10:26:09
(permalink)
rjbelans OK. Thanks Nahte27. Like I said though, it will not be until the weekend. I'm in a race to 30 mil with Punchy and can't let up at all. That's cool. I appreciate you being willing to try this. Also, I forgot you have to turn on Async mode. So the settings are Post-FX off, Multi-core-PhysX on, Async mode on, Particle Count = 60000, Force PhysX CPU off, # of emitters = 3. Res = 640x480, Time = 60000ms. The benchmark is pretty picky about which settings you run it at.
|
Re:Some Early GT430 Results
2010/10/18 11:07:32
(permalink)
Thanks again. I also just found out from Jacob that the 430 does not support voltage control, so what I have now will be the highest OC I can run on this card.
|
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/18 15:47:11
(permalink)
OP updated with 820 OC SS.
|
Carlitos714
FTW Member
- Total Posts : 1126
- Reward points : 0
- Joined: 2009/03/01 02:36:45
- Location: San Diego, CA
- Status: offline
- Ribbons : 3

Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/18 16:48:09
(permalink)
i though increasing the shaders only help with ppd? or does overclocking all the card help with ppd?
[HTPC] Zotac Zbox Nano AD12, 4 GB Corsair 7-7-7-20, Intel X25-M 120 GB, 55 inch Samsung plasma [GAMING RIG/DAILY DRIVER] i7-3820 @ 4.85 Ghz, GA-X79-UDR3, Heatkiller 3.0, HWL BI GTX 480, MCP 350 x 2 , XSPC restop, EVGA GTX 680 SLI, HWL BI GTX 360, 16 GB Doms, Crucial M4 256 gb, 120 GB Torqx SSD, DELL PERC 5I W/ HD204UI 2TB x 6 RAID 5, Xonar DX, Corsair AX1200i, WinTV-HVR-1250 XSPC H2 [Peripherals] Ducky Dragon| MX518| Dell 2408WFP| Fiio E9 Amp| BeyerD DT-880 Pro 250 ohms
|
arestavo
CLASSIFIED Member
- Total Posts : 3009
- Reward points : 0
- Joined: 2008/02/06 06:58:57
- Location: Through the Scary Door
- Status: online
- Ribbons : 8

Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/18 16:59:48
(permalink)
With the 400 series Nvidia cards, the GPU clock speed is tied in with the shader speed, and they cannot be overclocked independently.
EVGA affiliate code: 9ZWDWFNW6A (Don't forget to upload your invoice or no credit is given!) FOLD ON
|
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/18 17:03:41
(permalink)
^+1 Also, I tried leaving the memory clock at stock and got artifacts. The highest core/shader I could get was with this particular combination. If I went to 750 or 800 on the memory, I got artifacts, so 775 was the sweet spot.
|
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/18 17:06:15
(permalink)
does overclocking the memory help on the fermi cards? and can you get higher core/shader clocks with lower mem clocks? edit: never mind, answered before i posted
post edited by -ZS-Carpenter - 2010/10/18 17:07:30
Main rig: gaming/folding Z68 FTW 2600k@4.2 980 Ti sc+ Dedicated folders x79 Classy/4820k 3x 970 Q9550/780i 2x 770
|
mramzi
New Member
- Total Posts : 93
- Reward points : 0
- Joined: 2008/05/07 07:11:02
- Location: Far From NoWhere
- Status: offline
- Ribbons : 1

Re:Some Early GT430 Results
2010/10/18 20:32:03
(permalink)
Sound good till now, think to buy one
|
AlHo
Superclocked Member
- Total Posts : 142
- Reward points : 0
- Joined: 2010/06/20 09:11:30
- Location: Whitley Bay, UK
- Status: offline
- Ribbons : 0

Re:Some Early GT430 Results
2010/10/19 04:39:14
(permalink)
Soz for the inappropriate drivel in my previous post, I suspect I was getting a little carried away... Anyway I really appreciate you putting the effort in to test this card for both folding and physX - looking forward to seeing the data. Thanks!
|
Re:Some Early GT430 Results
2010/10/19 07:04:29
(permalink)
_Nite_
rjbelans
OP Updated with info on 2nd wu.
Second wu (10951 [925pts.]), at 40% completion, is showing 9402.35 PPD in HFM.
that is seriously impressive.... I'm gonna snag a couple of those cards when I can.
I could easily cram three of these in my folding rig that I have now, stash the 9800 gtx+ away and fold like MAD!
|
mflanaga
CLASSIFIED Member
- Total Posts : 3033
- Reward points : 0
- Joined: 2009/05/13 18:48:11
- Status: offline
- Ribbons : 15
Re:Some Early GT430 Results
2010/10/19 10:19:46
(permalink)
Just bought one off newegg. I plan on putting it in an open x1 slot (with a x1 adapter I already have) next to my 2 450's. I'll post PPD and OC comparison here in a couple of days. (Had a open slot for a single slot card had to put something in it didn't I?  )
|
mflanaga
CLASSIFIED Member
- Total Posts : 3033
- Reward points : 0
- Joined: 2009/05/13 18:48:11
- Status: offline
- Ribbons : 15
Re:Some Early GT430 Results
2010/10/20 15:21:24
(permalink)
Update on the 430.. Seems using it in a X1 slot it no longer possible :-( At least not for Fermi cards. The board won't detect it. Funny how I can put a 9800GT in a x1 slot on a board with 2 GTX275's , no problem. The 430 does work if I put it in a PCI-E slot..but I have nothing open. UPDATE: I DO have an open slot on my X58 board..lol here I go, gonna try to run a 430 along with 2 275's. Have to add the GPU3 client in a new folder, I'll report back later.. this should be interesting..
post edited by mflanaga - 2010/10/20 15:40:38
|
Re:Some Early GT430 Results
2010/10/20 18:41:32
(permalink)
If you have any problems with the mixing of GPU2 and GPU3, just go to GPU Tracker V2 program, it will get everything running smooth together. GL.
|
mflanaga
CLASSIFIED Member
- Total Posts : 3033
- Reward points : 0
- Joined: 2009/05/13 18:48:11
- Status: offline
- Ribbons : 15
Re:Some Early GT430 Results
2010/10/20 19:21:24
(permalink)
Got everything up and running with no "problems"..The only thing is the 430 is a cpu cycle sucker just like the 450's I have in annother rig (the one that doesn't have another PCI-E slot). I'm running 2 275's along with smp2 on an I7 930. Just not enough cpu to go around. The PPD on one of the 275's drops by half when I start up the 430. From 9200-4800PPD. CPU usage in lasso shows the 275 cpu usage goes from 2% to 1%. The 430 pulled a 611 for its first wu, folding it at 3800PPD which is about right for those things. No sense in sacrificing half the points on one card just to make it up on another. Usage in precision shows 99% for both the 275's..NO downclock. Just lack of cpu cycles seems to be the cause. The second I shut off the 430, PPD on the 275 climbs right back to where it belongs in a few minutes. Don't know how to fix what isn't really "broke". These are the 260.89 drivers btw. GPU2 clients are systray (set up a long time ago).. the GPU3 is console. 275's are running core 11 430 is running core 15 E758 board that is dummy overclocked to 3.2.. rock solid stable for months for both cpu and gpu 1000W psu, got plenty to run this hardware. off to bed now.. will brainstorm this in the AM.. If you have any ideas Rjbelans, or anybody, feel free to chime in. This will be good to figure out a fix for..as GPU2 is phased out, more folders will be mixing cards like this as they begin to integrate the Fermis with 200 series like i'm doing now.
post edited by mflanaga - 2010/10/20 19:38:44
|
Re:Some Early GT430 Results
2010/10/20 20:28:43
(permalink)
Only things I can think of is to play around with the number of cores you dedicate to GPUs to maximize your overall PPD and see how the SMP and 275s like playing on the same core. Some have had luck with just changing priorities and getting GPU2 clients to run with SMP without needing to dedicate cores to the GPU (Falcon comes to mind).
|
mflanaga
CLASSIFIED Member
- Total Posts : 3033
- Reward points : 0
- Joined: 2009/05/13 18:48:11
- Status: offline
- Ribbons : 15
Re:Some Early GT430 Results
2010/10/21 04:42:12
(permalink)
rjbelans
Only things I can think of is to play around with the number of cores you dedicate to GPUs to maximize your overall PPD and see how the SMP and 275s like playing on the same core. Some have had luck with just changing priorities and getting GPU2 clients to run with SMP without needing to dedicate cores to the GPU (Falcon comes to mind).
I tried mixing and matching affinities and priorities alot before I stopped for the night. No luck so far. Ironically, when I first built the rig months ago, the 275's use so little CPU cycles that I had no problem running smp on all cores along with the cards with negligible PPD drop from doing so. Depending on the wu's for both of course. I played with cores and affinities assignments, found my best results were when I let everything pick what it wanted. Assigning one core for the gpu's was fine, but led to big TPF increase on the smp2 client. -bigadv did cause problems with PPD no matter how I did the core assignments or priorities, so I settled for smp2. I have virtually no crashes this way, so I don't mind the 'lesser" points. I'll play more with the setup today, and see what, if anything, I can do. I have no other place to put the card, so I have no choice. I think it's just Fermi being Fermi. They eat alot...lol
post edited by mflanaga - 2010/10/21 04:44:29
|
SM-71
New Member
- Total Posts : 48
- Reward points : 0
- Joined: 2010/08/26 09:51:54
- Status: offline
- Ribbons : 0
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/21 09:17:10
(permalink)
rjbelans I was actually hoping I would get at least one 611 pointer for comparison, but it looks like we either chewed them all up or they got removed because of all the failures. You still folding with this card? If you are, then you're getting 611's now... so any updates?
|
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/21 09:36:49
(permalink)
It is folding, but I am at work. I plan to do a similar addition to the OP tonight showing a few 611 wus completed.
|
mflanaga
CLASSIFIED Member
- Total Posts : 3033
- Reward points : 0
- Joined: 2009/05/13 18:48:11
- Status: offline
- Ribbons : 15
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/21 10:01:09
(permalink)
rjbelans
It is folding, but I am at work. I plan to do a similar addition to the OP tonight showing a few 611 wus completed.
Along those lines..while i'm trying to get all my hardware to play nice together, I'm currently running JUST the 430. It's working on P10632 (611) wu at 6067.9 PPD Not bad at all. Just wish that when I start the 275's that they didn't conflict with the 430. I have discovered this AM that it's NOT CPU cycle issues. I shut down the SMP2 client completely, same problem. It's between the cards. The 2 275's, no prob. 430 alone, no problem. All 3 together, ONE of the 275's ppd get cut in half, along with the 430. The other 275 is unaffected. weird
|
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/21 10:06:32
(permalink)
You may have run into a PCI-E lane limitation on your mobo.
|
mflanaga
CLASSIFIED Member
- Total Posts : 3033
- Reward points : 0
- Joined: 2009/05/13 18:48:11
- Status: offline
- Ribbons : 15
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/21 10:14:37
(permalink)
Yeah.. while typing this I'm thinking that you're right. I just now started the 430 first, then started the 275 card that's in the slot right next to it. PPD dropped on both once they got going. So I shut down the client for that 275, then started the 275 that's in the lower slot (it's seperated from the 430 by the old style regular PCI slot) BOTH clients folding at expected PPD! You think that if I moved the 430 to the BOTTOM slot on the mb and put the 275's together that it might help? I hesitate to do so because it would reduce airflow to one of the 275 fans. UPDATE: Well I found out whats going on for sure now, just don't know why. Moved the cards, made no difference. I started the cards one at a time while watching precision. First 275 (GPU1) fired up fine, then I started the 2nd one (GPU2)..good. Then I fired up the 430 and I saw GPU2 drop to 95%, while what should be GPU3 didn't move from zero!!! I have triple checked all my flags and shortcuts, The Fermi client pulled a core 15 like it should have, I just can't get the 430 to fire up as it's OWN card. The macine ID's are all different. I give up for now...need some air..
post edited by mflanaga - 2010/10/21 12:17:26
|
mflanaga
CLASSIFIED Member
- Total Posts : 3033
- Reward points : 0
- Joined: 2009/05/13 18:48:11
- Status: offline
- Ribbons : 15
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/21 14:18:50
(permalink)
FIXED IT!!! I knew I had to get away from the machine for a bit..so clear when I came back.. Simple mistake, I screwed up a flag. Forgot to pont to the folder that contains the GPU3 client in the target line.. stupid stupid stupid.........Putting all the shortcut properties windows side by side did it. I could see it instead of trying to remember it.. So to all who want to know, it's possible to fold a 430 with 2 275's on the same machine. All that along with a I7 930 running SMP2 on all 8 threads . CPU usage seems light for the 430 now that everything is configured correctly..
post edited by mflanaga - 2010/10/21 14:21:18
|
Re:Some GT430 Results - Max OC 820/1640/1550
2010/10/21 18:29:17
(permalink)
Congrats on getting things working. Gotta hate it when it's something simple like that. Anyhow, OP updated with 611pt SS while at max OC. I think I will start a different thread for the FluidMark results (just need to finish race with Punchy to 30mil)
|