Discontinued: Lancelot (Heavy Duty 400 Mega hashes, 2 x FPGA Mining Board)

[Xialla]

Yours is probably in the air.  Mine hasn't even been scanned.

yes, otherwise, DHL is working during weekend but delivery is possible (here) only during working days, so I will have to wait until Monday. If you have got already tracking #, I'm pretty sure, that you will receive your package also on Monday.)

[blackarrow]

Those brick power supplies are #$%@

I disagree. I have 9 pcs running here and no problems (apart from melting the plastic if they're too close to each other)

[blackarrow]

At least an hour.
The speed is of course an estimate, the A over a day or 2 will tell you if it is accurate.
However, 3 HW with 12 A is pretty bad too (20%)

Apart from bogus speeds reported by cgminer (900Megahashes), it actually started yielding better results... It had quite a lot of dropped packets but more accepted packages than the other devices.

Code:

--------------------------------------------------------------------------------
 [P]ool management [S]ettings [D]isplay options [Q]uit
 ICA 0:                | 529.1M/760.8Mh/s | A:58 R:0 HW: 4 U: 2.11/m
 ICA 1:                | 436.6M/425.9Mh/s | A:53 R:0 HW: 2 U: 1.93/m
 ICA 2:                | 391.7M/394.7Mh/s | A:50 R:0 HW: 0 U: 1.82/m
 ICA 3:                | 979.8M/1.001Gh/s | A:70 R:0 HW:14 U: 2.55/m
 ICA 4:                | 398.1M/411.5Mh/s | A:55 R:0 HW: 2 U: 2.00/m
--------------------------------------------------------------------------------

Guess which one is running v42 firmware ?

[bitpop]

Yeah mine says it left hk. But i have 6 boxes I think !

[bitpop]

5    Departed Facility in HONG KONG - HONG KONG    HONG KONG - HONG KONG    15:07    
 

[bitpop]

I'm not risking them going bad and delivery mains voltage to my miners.

Those brick power supplies are #$%@

I disagree. I have 9 pcs running here and no problems (apart from melting the plastic if they're too close to each other)

[FlappySocks]

OK, thanks. I guess there is something wrong with my tracking number. Hopefully it's on it's way.

[bitpop]

Don't worry, customs just picked yours up for xray.

[elchorizo]

I received the remaining 5 Lancelots in my order today. They seem to run just a tad faster then the previous 5 I got... I'm not complaining They are all working great. New ones: 399 MH/s, old ones, 379 MH/s.

Here are some pictures.

[HTML5 Gordon]

For new orders, any idea on current turnaround times?

[jimmy3dita]

At least an hour.
The speed is of course an estimate, the A over a day or 2 will tell you if it is accurate.
However, 3 HW with 12 A is pretty bad too (20%)

Apart from bogus speeds reported by cgminer (900Megahashes), it actually started yielding better results... It had quite a lot of dropped packets but more accepted packages than the other devices.

Code:

--------------------------------------------------------------------------------
 [P]ool management [S]ettings [D]isplay options [Q]uit
 ICA 0:                | 529.1M/760.8Mh/s | A:58 R:0 HW: 4 U: 2.11/m
 ICA 1:                | 436.6M/425.9Mh/s | A:53 R:0 HW: 2 U: 1.93/m
 ICA 2:                | 391.7M/394.7Mh/s | A:50 R:0 HW: 0 U: 1.82/m
 ICA 3:                | 979.8M/1.001Gh/s | A:70 R:0 HW:14 U: 2.55/m
 ICA 4:                | 398.1M/411.5Mh/s | A:55 R:0 HW: 2 U: 2.00/m
--------------------------------------------------------------------------------

Guess which one is running v42 firmware ?

Looks like a Xilinx Jtag adapter is needed to upgrade the firmware a do some tests, any hints? Wich one are you using?

[kano]

At least an hour.
The speed is of course an estimate, the A over a day or 2 will tell you if it is accurate.
However, 3 HW with 12 A is pretty bad too (20%)

Apart from bogus speeds reported by cgminer (900Megahashes), it actually started yielding better results... It had quite a lot of dropped packets but more accepted packages than the other devices.

Code:

--------------------------------------------------------------------------------
 [P]ool management [S]ettings [D]isplay options [Q]uit
 ICA 0:                | 529.1M/760.8Mh/s | A:58 R:0 HW: 4 U: 2.11/m
 ICA 1:                | 436.6M/425.9Mh/s | A:53 R:0 HW: 2 U: 1.93/m
 ICA 2:                | 391.7M/394.7Mh/s | A:50 R:0 HW: 0 U: 1.82/m
 ICA 3:                | 979.8M/1.001Gh/s | A:70 R:0 HW:14 U: 2.55/m
 ICA 4:                | 398.1M/411.5Mh/s | A:55 R:0 HW: 2 U: 2.00/m
--------------------------------------------------------------------------------

Guess which one is running v42 firmware ?

However, if you look at the HW errors, it is indeed very high - almost 17%
I'd think that might mean it might shorten it's life?

Is there a way to clock that firmware down a little to get that HW number a lot lower?

Also the incorrect MH/s could be directly related to the firmware - the standard Icarus firmware divides the work in 2 on 2 FPGAs and runs at a constant MHz
I'm of course assuming that's what the standard Lancelot firmware does also and it seems to match that by reporting the MH/s correctly.
Do you know what that firmware does (and also do you have details about how to upload it? I have a jTag cable also that I have never used ...)

[Pitlord]

How is your group manufacturing plan going, blackarrow? I'm interested in 5 pcs.

[blackarrow]

I received the remaining 5 Lancelots in my order today. They seem to run just a tad faster then the previous 5 I got... I'm not complaining They are all working great. New ones: 399 MH/s, old ones, 379 MH/s.

You must have the 190 mhz firmware on the first ones. Buy a xilinx jtag cable and write the 200 mhz firmware and all will run at 400 mhashes. 

[LaserHorse]

You must have the 190 mhz firmware on the first ones. Buy a xilinx jtag cable and write the 200 mhz firmware and all will run at 400 mhashes.  

Will all new orders use the latest firmware?

EDIT: blackarrow - sent you a PM

[trebor_trebor]

At least an hour.
The speed is of course an estimate, the A over a day or 2 will tell you if it is accurate.
However, 3 HW with 12 A is pretty bad too (20%)

Apart from bogus speeds reported by cgminer (900Megahashes), it actually started yielding better results... It had quite a lot of dropped packets but more accepted packages than the other devices.

Code:

--------------------------------------------------------------------------------
 [P]ool management [S]ettings [D]isplay options [Q]uit
 ICA 0:                | 529.1M/760.8Mh/s | A:58 R:0 HW: 4 U: 2.11/m
 ICA 1:                | 436.6M/425.9Mh/s | A:53 R:0 HW: 2 U: 1.93/m
 ICA 2:                | 391.7M/394.7Mh/s | A:50 R:0 HW: 0 U: 1.82/m
 ICA 3:                | 979.8M/1.001Gh/s | A:70 R:0 HW:14 U: 2.55/m
 ICA 4:                | 398.1M/411.5Mh/s | A:55 R:0 HW: 2 U: 2.00/m
--------------------------------------------------------------------------------

Guess which one is running v42 firmware ?

However, if you look at the HW errors, it is indeed very high - almost 17%
I'd think that might mean it might shorten it's life?

Is there a way to clock that firmware down a little to get that HW number a lot lower?

Also the incorrect MH/s could be directly related to the firmware - the standard Icarus firmware divides the work in 2 on 2 FPGAs and runs at a constant MHz
I'm of course assuming that's what the standard Lancelot firmware does also and it seems to match that by reporting the MH/s correctly.
Do you know what that firmware does (and also do you have details about how to upload it? I have a jTag cable also that I have never used ...)

v42 firmware issues like HW (10% acceptable), miscalculated hashrates, jtag usage etc.:
https://github.com/ngzhang/Lancelot/tree/master/Bitsteam

@blackarrow
ICA3 is v42 ?

[FCTaiChi]

So without a gox to play with how can we order?

[Xialla]

blackarrow,

could you please explain me, why you describe the package as "lancelot" on not as "sample of motherboard" as before? because officers know what is motherboard but don't know what is lancelot, i have got clearance event and must pay VAT+CLO and other fees for extra storing package, etc.

next thing is, that will have to wait until tuesday because of this.

[blackarrow]

Okay, been mining for about 30 mins now, with power from the PSU and going strong, updated my earlier posts too as It was my fault that I didn't notice the not working green light on the power brick

Use  --icarus-timing short for 400Mhash performance.

[kano]

Regarding v42
So ... I had a look at the github and indeed it does dynamically adjust the MHz clock
So --icarus-timing will not fare well with that.
(as I mentioned it expects a constant MHz)

Since it can't even know what the MHz is it would be better to instead use a clock figure slower than it's slowest speed and a abort time of around 100 to ensure it's under the speed it always runs at.
So something like:
--icarus-timing 2.5=100
It will show the MH/s incorrectly, but it will still hash at the highest rate it can.

Anyone know if you can get more information out of the Lancelot other than just send it work and wait for the reply?
I guess once I've finished the code - (almost done but haven't had a lot of time today to do that last bit: sort out the chip initialisation) - I'll look at the github and see if there are other commands you can send to it ...
The random unknown MHz is actually problematic