You don't trust him? I sent 1000 BTC. I'm sure to get it back... right guys? ...guys?

Is there any way to see the net hashrate for SHA and scrypt? The typical "getmininginfo" command throws back an error message.

I had this problem as a result of a driver problem but it wasn't that simple.

I was getting driver crashes because of low quality PCIe risers with weak connections.

If you get a driver crash as soon as your PC is booted, you need to start eliminating cards until it doesn't crash any more.

Bad risers can cause boot/post/driver problems and are extremely hard to diagnose.

Typically if you have recurring problems after successful/clean software installation you need to start looking at your hardware.

I've been attempting to run 4x 290s(non X) on a 1300W PSU and been running into trouble and had to reduce my rig to 3 GPUs. I would not recommend attempting it on a 1000W.

You had too much draw on your board. Do you really have 100cm(3 feet) long risers?

If they are actually that long, your GPUs would pull too hard and that's why you had a burnout.

The only way you could hope to make that work on more than 3 cards would be to undervolt your cards before putting them on risers. If you have them on the stock voltage and software, the card will pull whatever it needs(remember that the power on will momentarily pull max power through every port for POST). If you're on a 100cm riser, you're pulling WAY more than a 1x slot can put out when it's already probably struggling to feed the cards on the 16 pin connectors.

It's a hard lesson but smart people only have to learn it once.

You'll probably have to replace that board but when you do, stick with 3 cards max unless you get shorter powered risers.

Try this:

"intensity" : "13",
"vectors" : "1",
"worksize" : "256",
"kernel" : "scrypt",
"lookup-gap" : "2",
"thread-concurrency" : "8192",
"shaders" : "0",
"gpu-engine" : "925",
"gpu-fan" : "65-65",
"gpu-memclock" : "1550",
"gpu-memdiff" : "0",
"gpu-powertune" : "20",
"gpu-vddc" : "0.000",
"temp-cutoff" : "95",
"temp-overheat" : "85",
"temp-target" : "75",
"api-mcast-port" : "4028",
"api-port" : "4028",
"expiry" : "5",
"gpu-dyninterval" : "7",
"gpu-platform" : "0",
"gpu-threads" : "2",
"hotplug" : "5",
"log" : "5",
"no-pool-disable" : true,
"queue" : "1",
"scan-time" : "5",
"scrypt" : true,
"temp-hysteresis" : "3",
"shares" : "0",
"kernel-path" : "/usr/local/bin"

He might be mining on a card that's in standby mode, too.

Except I don't think any of the HD 7XXX cards ever needed jumpers.

Could be if there's no monitor plugged into the GPU then he just needs to turn off his power control through Windows which will put it on standby if it doesn't detect a monitor.

Having the "shaders" value at 0 means that it will default to whatever the native value would be for the GPU so changing that won't do anything.

OP; What is the exact model number on the sticker on the back of your GPU? It should look something like FX-797A-**** If it has a V1.0, 2.0 or 5.0 list that too.

I would say you might need to flash a new BIOS to your card but getting ~600 is higher than most people(including myself) get without a BIOS flash. This tells me that you have a setting issue. If you tell me the model number I can help tweak it.

Amazon does this too. I had a merchant allow me to complete an order which I paid overnight shipping for. I didn't get a response from them for 2 days when they said that they were cancelling my order and refunding my credit card.

I had to pay exchange fees through my bank for both transaction and the whole thing cost me about $20 and I got nothing. At least my bank refunded the fees after I fought with them about it.

If you're running out of PCIe cables for your GPUs it's a good indicator that you need more PSUs.

If you know that your single PSU is strong enough for the number of cards you are using, you can run the 6+2 to 6+2 connectors directly from the PSU.

The standard 6 pin connector can be hooked up through the 12/24V peripheral chain cable. I wouldn't put more than 2 or 3 on a single PERIF socket.

If you're running 6 cards on a single rig and you have a high-end PSU, you should have 6x 6+2 outlets and at least 2 PERIF outlets for the second adapter chain.

How are you determining your thread concurrency? From my experience, higher concurrency is not better for these cards. Try removing the concurrency line from your .conf file, deleting your bat file and seeing what it defaults to. That will give you a good starting point for where you should be.

Your clock speeds need some tweaking. Your core clock is a bit high and your mem clock is a bit low. Try something like 1040 core clock and 1500 mem clock and see where that gets you. These cards can be massively overclocked with few hardware errors but they do get hot so make sure your ventilation is good.

Try it with:

core clock - 1040
mem clock - 1500
intensity - 13
gpu threads - 2
thread concurrency - 8192
everything else the same as you have currently and report back.

It's not a temp issue. When I start the rig they hit their max hash rates before they hit 60C and cap out at the ~730, ~360, ~360. They never go above.

Ok so I'm going to keep this as short as possible but here goes...

I have 3 of the XFX 7970s that are so well known for their low hash rates for Scrypt mining.

1 of them is fine, but 2 of them(from a separate batch, according to serial numbers) have the low hash issues.

I followed the instructions here: https://litecointalk.org/index.php?topic=8984.0 to flash a new BIOS onto the cards.

I tried the suggested BIOS in that thread on the 2 cards that are having issues and it gave me black screen. I then found another thread that gave information for the UPDATED BIOS directly from XFX for that card. The flash went perfectly, all of my cards are operational and the BIOS on all 3 of them is identical.

Here's where is gets weird... The BIOS, model number and everything is identical on these cards. I tested them individually and they ALL hash at ~730kh. However, when I install them all into the same mining rig, 1 of them hashes at ~730 and the same 2 that had problems before hash at ~360.

What. The. Fuck.

Has anyone seen anything like this? Does anyone have any ideas why this would be happening?

Hey atp, what's the difference between sgminer and cgminer? Just a guess, but is sgminer designed for scrypt and has support following the cgminer 3.7.2 version?

If you're building a GPU mining rig; go for an AMD setup.

If you're building a CPU rig, go for Intel.

I'm glad you got it working! What did you do that solved the hashrate problem?

Also, I may be able to help you with tweaking if you like. What brand/model of 290x do you have?

Just switched my miners over. I love the site dashboard and navigation. Very peppy and clean.

The only complaint I have right now is that it took some doing to figure out how to set my auto payout threshold and add my BTC address to my account.

Definitely going to be spending some time here!

A couple things:

You shouldn't need a dummy plug for 290x so don't bother with that.

The SDK software is of no use to you if you're using Catalyst 13.11 Beta or 13.12.

If you're eliminating all peripheral components(including other video cards) and risers, doing a power drain, wiping your drivers with either Display Driver Uninstaller or a sweeper and reinstalling and still having problems, you might have something corrupted in your OS or a MoBo issue.

Both of those issues would typically give BSODs.

If you're not getting any crashes, just weird hash rates, try a different cgminer version. Don't forget to eliminate cgwatcher for now and don't use GUIminer ever. 

Sounds about right. I'm running a 1300 Watt PSU and was trying to get 4 of them going. Once I dropped to 3, my stability suddenly got 100x better.

They are designed for that orientation. The fans on the outside of GPUs blow down towards the board, not up towards the adjacent cards.

My setup has only about 1.25 to 1.5 inches between cards and I have no heat issues whatsoever.

The exhaust comes out the sides of the assembly, like this: