X6500 Custom FPGA Miner

[Hawkix]

OK, OK, next time I will explicitly append this: 

[1714004874]

1714004874

[1714004874]

1714004874

[1714004874]

1714004874

[shackleford]

In the last couple weeks, we've been experimenting with different FPGA core voltages (VCCINT). Our measurements have shown that on most rev3 boards, performance can be increased significantly by slightly increasing this voltage. Because of this, all rev3 boards starting today will be shipped with VCCINT set to 1.23V (instead of the original 1.20V). This applies to all boards shipped on or after 3/29/2012.

While this increases the board's power consumption slightly (data here), it also increases the potential limit on clock rate, which is especially useful with the new "overclocker" bitstreams.

If anyone would like to increase the core voltages on an older board, it simply means swapping two resistors (R2 and R8). You can either attempt this yourself or send it back to us and we'll do the work free of charge. We only ask that you pay the return shipping charge.

I am contemplating this… Just to be clear it is taking the resistor from the R2 spot and putting it in the R8 spot (both by the upright voltage regulators?) and vice versa.  I think that is what you are saying but better safe than sorry. I don’t have a magnifying glass but it looks like it says 1911 (or 1161) on both of them and I did not see them in the schematic.  They are pretty small and in a semi tight spot. If they were one of the bigger ones I would not hesitate to do it. I am going to do some research on how to solder this myself because the shipping and downtime sounds like a costly pain. Do you have any tips off hand for working with these tiny resistors? I may look to find a local place to do it if I can’t or just not rock the boat at all.

NM found some good Youtubes on surface mount soldering .. still looks pretty small to me though.

[fizzisist]

In the last couple weeks, we've been experimenting with different FPGA core voltages (VCCINT). Our measurements have shown that on most rev3 boards, performance can be increased significantly by slightly increasing this voltage. Because of this, all rev3 boards starting today will be shipped with VCCINT set to 1.23V (instead of the original 1.20V). This applies to all boards shipped on or after 3/29/2012.

While this increases the board's power consumption slightly (data here), it also increases the potential limit on clock rate, which is especially useful with the new "overclocker" bitstreams.

If anyone would like to increase the core voltages on an older board, it simply means swapping two resistors (R2 and R8). You can either attempt this yourself or send it back to us and we'll do the work free of charge. We only ask that you pay the return shipping charge.

I am contemplating this… Just to be clear it is taking the resistor from the R2 spot and putting it in the R8 spot (both by the upright voltage regulators?) and vice versa.  I think that is what you are saying but better safe than sorry. I don’t have a magnifying glass but it looks like it says 1911 (or 1161) on both of them and I did not see them in the schematic.  They are pretty small and in a semi tight spot. If they were one of the bigger ones I would not hesitate to do it. I am going to do some research on how to solder this myself because the shipping and downtime sounds like a costly pain. Do you have any tips off hand for working with these tiny resistors? I may look to find a local place to do it if I can’t or just not rock the boat at all.

Sorry, "swapping" was a bad word choice. Those two resistors need to be replaced with 1.82k resistors (they are 1.91k). Unless you are getting a lot of invalids at higher clock rates, you should probably not make this change. It will reduce invalids (or increase the clock rate at which invalids start appearing), but if you have 0% invalids now, you can't get any lower than that!

[li_gangyi]

R2 and R8 are 0805 sized resistors. They are not terribly hard to solder, compared to say 0402 sized parts, but replacing R8 might be more of a challenge because of it's location.

A small soldering iron, good light, some soldering (not plumbing) flux, desoldering braid/wick and a pair of tweezers (failing that, a toothpick) should make short work of this.

[shackleford]

R2 and R8 are 0805 sized resistors. They are not terribly hard to solder, compared to say 0402 sized parts, but replacing R8 might be more of a challenge because of it's location.

A small soldering iron, good light, some soldering (not plumbing) flux, desoldering braid/wick and a pair of tweezers (failing that, a toothpick) should make short work of this.

Hi, sorry to be a pain but can you link me to the right resistor? There are some options and I want to be sure I am getting the right one. I checked Radioshacks website and looks like I am going to have to order them. Amazon has this http://www.amazon.com/Thin-Film-Tech-Precision-Continuous/dp/B0015DZIGC . Or would you guys just consider selling a few and shipping usps priority for some btc (would like to attempt this weekend)? Most the other stuff I have but I may need to get a smaller tip for the iron or upgrade.

I am at 0 invalids on the 200mhz firmware with copper and air but would love to push it a little after I get really good cooling so figured I might as well make this change. What kind of head room are you guys seeing?

[fizzisist]

R2 and R8 are 0805 sized resistors. They are not terribly hard to solder, compared to say 0402 sized parts, but replacing R8 might be more of a challenge because of it's location.

A small soldering iron, good light, some soldering (not plumbing) flux, desoldering braid/wick and a pair of tweezers (failing that, a toothpick) should make short work of this.

Hi, sorry to be a pain but can you link me to the right resistor? There are some options and I want to be sure I am getting the right one. I checked Radioshacks website and looks like I am going to have to order them. Amazon has this http://www.amazon.com/Thin-Film-Tech-Precision-Continuous/dp/B0015DZIGC . Or would you guys just consider selling a few and shipping usps priority for some btc (would like to attempt this weekend)? Most the other stuff I have but I may need to get a smaller tip for the iron or upgrade.

I am at 0 invalids on the 200mhz firmware with copper and air but would love to push it a little after I get really good cooling so figured I might as well make this change. What kind of head room are you guys seeing?

No problem! This is the one we're using: http://search.digikey.com/us/en/products/RR1220P-1821-D-M/RR12P1.82KDCT-ND/

The one you linked is fine, too. Even better, actually, with that incredible 0.1% precision!

[TheSeven]

I am at 0 invalids on the 200mhz firmware with copper and air but would love to push it a little after I get really good cooling so figured I might as well make this change. What kind of head room are you guys seeing?

In that case, I'd wait before attempting this until you actually hit invalids. It might well be possible that if you were lucky and got a board with an extraordinary stable power supply and fast silicon, you can get away with >200MH/s at the stock voltage, which is likely to be more energy efficient than a couple MH/s more at a higher voltage.
Remember, increasing the voltage increases the power consumption even at the same hash rate, so there's no point in increasing the voltage as long as you can't push it to the frequency limits anyway.

[shackleford]

I am at 0 invalids on the 200mhz firmware with copper and air but would love to push it a little after I get really good cooling so figured I might as well make this change. What kind of head room are you guys seeing?

In that case, I'd wait before attempting this until you actually hit invalids. It might well be possible that if you were lucky and got a board with an extraordinary stable power supply and fast silicon, you can get away with >200MH/s at the stock voltage, which is likely to be more energy efficient than a couple MH/s more at a higher voltage.
Remember, increasing the voltage increases the power consumption even at the same hash rate, so there's no point in increasing the voltage as long as you can't push it to the frequency limits anyway.

All three of my boards have never gotten an invalid @ 200mhz. I am guessing it is because they are well cooled.

I don't know what they are seeing for speed headroom but I have to guess it must be enough to make the bump to voltage standard. The couple additional watts is not that big of a deal to me and I think it would be best to be the same as the other r3 boards that have the voltage bump.

One big reason I would like to do it sooner rather then later is I will be putting them in oil soon and I don't want to take them out and clean them to do it later. Hopefuly the resistors will be here tomorrow.

[99Percent]

I don't think putting these in oil is of any use. The oil heats up and eventually you will need to take out the heat out of the oil too usually with an air fan which you use directly anyway.

Looking at your setup, I can guess you can use just one fan for all three boards. Why don't you experiment to see if you start getting invalids when you remove some of the fans.

[freshzive]

What heatsinks are those?

[shackleford]

What heatsinks are those?

The heatsinks are these http://www.enzotechnology.com/cnb_s1.htm

I don't think putting these in oil is of any use. The oil heats up and eventually you will need to take out the heat out of the oil too usually with an air fan which you use directly anyway.

Looking at your setup, I can guess you can use just one fan for all three boards. Why don't you experiment to see if you start getting invalids when you remove some of the fans.

We were discussing changing the resistor to bump the voltage to the now standard level for these boards. I mentioned mineral oil only as a reason why I wanted to do it now rather than later. There are many things you are not considering with the oil. I don’t really care to argue about it, I have done a lot of research and planning and I am doing it. Hopefully next week I can share results.

[fpgaminer]

Hey guys. I finally put together an installation package of sorts for Windows:

X6500 Easy Package for Windows
Download


This should make it simple and easy take an X6500 and go from 0 to Bitcoin in no time (on Windows). It includes a driver installation program for the libusb driver, MPBM with compiled EXE, and all the current mining firmware. I tested it out on a clean VM install of Windows 7 64-bit, Windows 7 32-bit, and Windows XP 32-bit.

If that's too convenient for you, the individual pieces can also be downloaded:

• FPGA Mining Driver Install
• Modular Python Bitcoin Miner
• Firmware

Let me know if there are any problems. Though I did take a lot of time testing it, this is still fresh off the presses

[99Percent]

We were discussing changing the resistor to bump the voltage to the now standard level for these boards. I mentioned mineral oil only as a reason why I wanted to do it now rather than later. There are many things you are not considering with the oil. I don’t really care to argue about it, I have done a lot of research and planning and I am doing it. Hopefully next week I can share results.

I am looking forward to your results!

[BR0KK]

Hm on cablesaurus the new batch is sold out (not on stock). How long does it take to get one or two units (to Germany)?

[hashking]

Hm on cablesaurus the new batch is sold out (not on stock). How long does it take to get one or two units (to Germany)?

Anyone know how to get the x6500 to mine on GPUMAX.  Having no luck.

[Pipesnake]

Hey guys. I finally put together an installation package of sorts for Windows:

X6500 Easy Package for Windows
Download


This should make it simple and easy take an X6500 and go from 0 to Bitcoin in no time (on Windows). It includes a driver installation program for the libusb driver, MPBM with compiled EXE, and all the current mining firmware. I tested it out on a clean VM install of Windows 7 64-bit, Windows 7 32-bit, and Windows XP 32-bit.

If that's too convenient for you, the individual pieces can also be downloaded:

• FPGA Mining Driver Install
• Modular Python Bitcoin Miner
• Firmware

Let me know if there are any problems. Though I did take a lot of time testing it, this is still fresh off the presses

Very nice!  Will test it out asap and post up any issues.

[fpgaminer]

Anyone know how to get the x6500 to mine on GPUMAX.  Having no luck.

Using x6500-miner software or MPBM? If you're using MPBM, TheSeven can help address that issue, though I don't know if he has access to GPUMAX so it could be difficult until he does. I recall the issue being brought up before, and it may be related to the way MPBM handles Long Poll address changes.

[tgmarks]

Easy package for windows is awesome and simple.  Much appreciated.

[freshzive]

I've mined on GPUMAX since its inception using MBPM with no issues. What kind of errors are you seeing?

[TheSeven]

Anyone know how to get the x6500 to mine on GPUMAX.  Having no luck.

Using x6500-miner software or MPBM? If you're using MPBM, TheSeven can help address that issue, though I don't know if he has access to GPUMAX so it could be difficult until he does. I recall the issue being brought up before, and it may be related to the way MPBM handles Long Poll address changes.

It was another long polling issue (the address change thing was with some other proxys), but that should have been fixed some time ago, and I think it has been pushed to master by now (not sure though).