Maybe it has something to do with the fact that these new boards have UEFI and fancy chips like PLX PEX8608, NF200, Lucid Hydra and these chips add extra latency to the routing and the extenders also do that and the driver cuts off any cards that do not respond within a certain ms limit or something ?
UEFI; no not likely a factor. But you may have a point, if the motherboard has a PCIe switch onboard, it may give you less headroom to use extenders. Extenders do not conform to the PCIe spec, it just exploits some headroom, but there might not be enough on complex boards with long traces or perhaps slightly more crappy PCIe controllers. I really am going mad over this issue. Is it the P67 chipset with only 16 PCIe lanes from the CPU and the rest have to be duplicated ? IRQ conflicts ? UEFI problems ? Technically the extender is just a cable extender and it is dumb ( has no extra circuitry ) so should work as long as it is within the 19 cm or so spec limit. I can't even get 11cm and 7cm lengths to work on my motherboards. Should I stick to old hardware from 5 years ago from now on ? I really want to resolve this issue and maybe jjshabadoo can cooperate so we can solve it since it seems that he also faces the same problem as me. Thanks ! |
|
|
|
Exactly right guys, the extenders are killing me. I just whacked in 4 5870's on the asus p67 board and BANG, no issues. Even able to take a card in and out. WTF is wrong with extenders and why do they cause so many issues? It's killing us on my other rigs too. I've realized the cards do fine even in close quarters on the mobo as long as you blow some extra fans through them and screw the overclocks. Get what you can and just f'in leave it alone.
I do have those 5870's at 950/160 and they are fine. worst temp is 71c. These are stock 5870's from ben's outlet.
I do have one box with one card on an extender and its been fine. I'm going to move toward always putting the first card on the board in slot 1 and then a card in the last slot. Open cases of course, basically mobo trays nailed to boards or whatever and then maybe put one more card on an extender in slots 2 or 3 or 4.
I've also learned we are getting killed by trying to use two psu's. We have two rigs with 3 5970's connected to two seasonic x750 gold psu's connected with a lian li PSU connectors and those bitches fry pci extenders like they're going out of style, powered or not. Nevermind they are NEVER stable.
So I'm going to either unload the 750's or actually, probably cheaper to just buy one more motherboard, cheap ram and cheap cpu. Just a shame to have boards like MSI 890FXA-GD70's with only two cards, but F it. We can mine for a while and then maybe to decide to screw with one rig at a time.
Sorry for the long post, but thanks for the help and this noob has learned his lesson. Don't get too fancy maxing shet out when you're just getting started even if you think you know hardware, because mining is its own animal.
I can build and overclock the S out of a gaming machine, but so what.
Please don't tell me that the Asus board which did not work with extenders is the P8P67 WS Revolution motherboard. I just bought one hoping it will solve my extender problems. I pray that S works because I am trying since June 2011 to solve this extender problem. 4 cards plugged straight into the mobo works flawlessly. As soon as I use just 1 extender and the rest plugged in normally the whole thing does not work anymore and it is not xorg.conf or any other such BS ( Linux headless so not even SSH can solve my issues without any monitor ). I'm no noob but this issue has been killing me ever since. The PSU ( AX1200 ) is just fine etc. If you are having the same issue maybe let us cooperate so we can solve it and use extenders for better cooling and less fan noise / wear ! Maybe it has something to do with the fact that these new boards have UEFI and fancy chips like PLX PEX8608, NF200, Lucid Hydra and these chips add extra latency to the routing and the extenders also do that and the driver cuts off any cards that do not respond within a certain ms limit or something ? I have tried PCI latency timer but still the same BS. I really am desperate LOL. Is this your issue ? I have an old Asus P5ND2-SLI motherboard which works flawlessly with the same extenders and is just perfect, old BIOS based and no nonsense crap chips like Lucid or PLX BS and it just works with 3 extenders and 1 plugged into the board directly and Linuxcoin never fails to start. On the new UEFI rig I managed to get Linuxcoin / BAMT to start while using the cards with the extenders but it really is hit and miss. One time it works but the next 100 times it does not for some reason. I have dmesg logs if you need them etc. all using Linuxcoin / BAMT through SSH without any monitor etc. Thanks ! |
|
|
|
Nice to see that finally in February people should be getting their products. Hopefully BFL delivers so all the naysayers can stop. I doubt they will deliver though. 4-6 weeks anyone  ? |
|
|
|
Seriously stop using deepbit  No. Deepbit is awesome. It is the biggest pool for a reason. They should do a 51% attack to wake up the folks thinking "don't fix the 51% vulnerability because it will not happen anyway". |
|
|
|
Just install the version you prefer, same as you would do on any linux box. So the two would not conflict like on Winblows, right ? |
|
|
|
Any idea about how I might replace that with 2.1 because 5870s + 2.1 = rocks while 5870s + 2.4 = meh. Thanks ! |
|
|
|
Thanks, I have another 480 I can use, so might as well I guess, but was wondering if I then should just go double loop. I have an extra radiator pump and reservoir exactly like the ones below.
To Art's comments about the kaze fans, I got 20 of them for $80 as part of all this other water cooling stuff used, so what the hell, I'll use them until they burn up and then go delta or panaflos. They should at least go well with these rads, because the research I did showed these radiators are great with high RPM fans.
Here's the system I'm building:
ASUS P8P67 WS motherboard
silverstone 1500 watt psu
Intel celeron G530 from newegg
Crucial 2GB 240 pin RAM
4 x HD 5970's
4 x EK acetal/nickel full cover water blocks
4 x EK RAM backplates
EK fc quad parallel bridge
EK fc link r5970
koolance quad 120mm rads HX-CU
Swiftech MCRES-MICRO™ Rev. 2
UltraKaze 120x38mm fans
Swiftech mcp 655 stock
1/2" fittings
1/2" ID 3/4" OD tubing
Operating System
16gb flash drive with linuxcoin running cgminer
Of course all unecessary motherboard features turned off, USB 3.0, etc. etc. whatever I can find to lower the wattage.
I'd also like to get all of this stuff in a case because I will be keeping it at one of my offices(no I'm not stealing the electric, I own the place!)
I just don't want to set it up all home rigged and then have an employee attempt to move it or something and screw it up. Sucks though, because any case that can hold two 480 rads is NOT cheap. Even if you're willing to gut some of the bigger cases. It looks like the cheapest one is the Zigmatek Elysium. I think you still have to remove all the HD cages to get two 480's to fit.
Sick configuration. You certainly get the rep. Good value for money, too, I might add. 4*5970s is a max of 8 GPUs so very good utilization of density right there. |
|
|
|
|
What SDK does this use ? 2.1 or 2.4 ?
Thank you !
|
|
|
|
Looking at the circuit board and not investing a whole bunch of time, it looks like this is the likely chipset: $487Xilinx XC6VCX75T-1FFG784CFPGA Virtex®-6 CXT Family 74496 Cells 40nm (CMOS) Technology 1V 784-Pin FCBGA Package 784FCBGA
Family Name Virtex®-6 CXT
Device Logic Units 74496
Number of Registers 93120
Typical Operating Supply Voltage 1 V
Maximum Number of User I/Os 360
RAM Bits 5750784The spec sheet for current draw also might make one think it would draw around ~22w... And they make money using this? Can't be this because it has to be 2 of them chips I think. |
|
|
|
MiningBuddy's awesome enclosure and directed air paths setup. WOW. That is really NEAT. I wish I had your DIY skills Sir ! I can barely put together a ghetto shoe rack setup  |
|
|
|
Another thing worth noting is that you have yet to disclose any personal information, which by no means is necessary, but can be good for credibility.
You can look up who registered the domains btclending.com and gpumax.com. I'll go ahead and call BS on that. http://www.google.com/search?q="Don+Shrents" He is not Don Shrents. I know his name, birthdate, current city, and a previous city he lived in. All of this information is publicly available to anyone who cares enough to thoroughly research it. I did this, and asked him a few questions before sending him a single BTC. But thanks for the warning Mr. Scammertag. Care to share that information for the rest of us mortals without your killer internet research "skillz" ? Thanks ! |
|
|
|
Well someone is trying to beat mybitcoin.com's record I see IMHO if you put your money in these schemes you deserve to lose them or be found out ( I doubt pirate's business is 100% legal ) |
|
|
|
It is only the trolls with too much time on their hands and not enough money in their pockets that seems to create all of the FUD. Sad.
As a potential customer, it makes sense for me to investigate their product. There's plenty of FUD going around, but that's directly caused by BFL's major problems. We Fear the shipped product won't meet their claims because BFL keeps changing the claims. We are Uncertain when they will ship because BFL keeps changing the ship date. We Doubt they know what they're doing because BFL launched products solely based on a simulations. The FUD is all BFLs fault. Indeed, they are trolling us and they are trolling themselves A little more transparency about what is going on under that heatsink and in their company wouldn't hurt at all right now IMHO. |
|
|
|
930 core. under 70C.
430Mh/s at 930 core on a 58*50*? That'd be 335Mh/s stock. On a 5850. Or 436Mh/s on a stock 5870. So... why is everyone else reporting numbers *way* below that? ArtForz, how do you calculate these formulas and stuff ? I look and don't understand anything. LOL How do you get the mhash/s value from the core and shaders the card has ? |
|
|
|
Can't remember who it was but some dude got them at around 720 or 700 mhash/s !
But then the power consumed makes your MH/W ratio suck !
1-upd! max @ stock V: 1070 core, 695Mh/s @ 1.20V: 1170, 759Mh/s and for shits and giggles... @ 1.25V: 1240 core, 805Mh/s Tip: try phoenix with -k poclbm AGGRESSION=10 WORKSIZE=256 So what is the consensus ? Are CGN shaders more or less or the same in terms of mining compared to VLIW4 or VLIW5 shaders from 6970 and 5870 ? |
|
|
|
Without AMD support it is unlikely any Open Source driver that takes advantage of shaders will ever exist.
Err, what? Take a look at the Xorg radeon driver and how it does several parts of 2D acceleration on 5xxx and 69xx  Yeah, I pointed that out above but D&T is under the impression that Intel and open source ATI drivers don't support OpenCL when in fact they do ( somewhat ). If someone could let us use the damn open source drivers that would be magical. No need to have X server running anymore ! |
|
|
|
Looking to buy a 5830/50/70 to replace a card tha just died in my miner. Good condition preferred!
Where are you located ? |
|
|
|
From Intersango-
"The current lock on our account is due to a technical glitch which should be resolved soon"
It took them 3 days to respond to a support ticket via their website regarding this situation. An auto-response would have sufficed.
Also, I find their reason, 'a technical glitch' to smell of bs. I have two merchant accounts, one Lloyds and the other HSBC, and in 5 plus years have never had, or heard of, an account being frozen for this length of time because of 'a technical glitch'. Usually if there is an IT problem, which itself is very rare, all merchants are effected. Not just one.
So, what precisely is going on here? Now Intersango have publicly stated the problem is 'a technical glitch', we won't get a straight answer from them (no company in their right mind would correct previous statements unless forced to). I therefore can only guess as to the real reason for them holding on to peoples money for 4+ days. Assuming it hasn't gone walkabouts with a rouge staff member.
Also, can anyone recommend an exchange that allows GBP deposits via uk bank transfers or Ukash? Preferably one that won't keep your money from you for 96+ hours with only a lame excuse to justify their actions?
Heh, and I was fooling enough to believe they could be trusted ! |
|
|
|
If AMD ever changed their drivers to allow 16 (or better unlimited) GPU per system well then it might make more sense.
How much worse are the open source ATI drivers for linux than the proprietary ones for mining? It seems that it should be possible to modify them to remove a limit, but if they perform badly then there may be no point. Yeah open source drivers for mining would really be HEAVEN for me. Screw AMD and their crappy drivers and bugs. Somebody with the skill ( like ArtForz ) needs to try and sort this out for us. Look here to start : http://dri.freedesktop.org/wiki/GalliumComputehttp://www.x.org/wiki/RadeonFeature |
|
|
|
|
Show Posts
) 
