ZTEX USB-FPGA Modules 1.15x and 1.15y: 215 and 860 MH/s FPGA Boards

[BR0KK]

I'll test that when i have all the other things to build my Mining rig into an IKEA Helmer

For now I'm Hashing with 1,7GH/s@EMC.

[1713485902]

1713485902

[1713485902]

1713485902

[1713485902]

1713485902

[1713485902]

1713485902

[gr0bi42]

Just arrived. The brandnew ZTEX Quad's 

... and a 3,5 inch harddisk for size comparison.

SUPER fast shipping. Thanks ztex.

[BR0KK]

now I'm jealous to

[gr0bi42]

HELP, please. Maybe I'm blind, but I cannot find the firmware download link for the new quad 

I need the file: ztex_ufm1_15y1.ihx.

The link on the BTCMiner page goes only the quad product page.

[BR0KK]

No u dont need to dl that file. Its inside the jar and u call it with:

java -cp place of your .....jar file

Its written in the tut at ztex page. I did the same mustake :d

[gr0bi42]

No u dont need to dl that file. Its inside the jar and u call it with:

java -cp place of your .....jar file

Its written in the tut at ztex page. I did the same mustake :d

Yep. Thanks a lot. First board is running 

[MXRider]

Soooo jealous!! I ordered few for myself but ztex said it would take a month to ship. I think I ordered too many

[BR0KK]

I odered one to play with

Btw i need a new usb hub bcause, that cheapo trust thing doesn't like 10 FPGAs hooked to it. Btc miner allways throws an crc firmware error.... When its hooked op alone or wirh one single everything works fine ..,,
But dont touch even the mouse .....

[MXRider]

Any recommendations for USB hub? Need to connect 25 FPGAs.

Some tips about powering them are also welcome. That ATX Psu way seems little tricky. I think it would be easier to just get those 36€ PSUs from ztex.

[BR0KK]

Nope i can't hook up more than 9 FPGAs to my Mac with that usb hub (TRUST 7 Port Powered USB Hub: Connected to my MAc mini, doesn't matter witch port). Even if its powered separately..... The 10th FPGA gets CRC errors for the Bitstream.

€: its the new single i bought with the quad..... It thinks to be a 1.15y somehow ?
PIC:


what the heck did i do here ? How did this happen?

[Turbor]

Just arrived. The brandnew ZTEX Quad's 

... and a 3,5 inch harddisk for size comparison.

SUPER fast shipping. Thanks ztex.

Wow. Ordered one yesterday.

[gr0bi42]

Any recommendations for USB hub? Need to connect 25 FPGAs.

+1

Some tips about powering them are also welcome. That ATX Psu way seems little tricky. I think it would be easier to just get those 36€ PSUs from ztex.

I will go with an ATX PSU. I'm planning with a Crossair 1200 Gold and PCIe-8 extensions. 2 PCIe-8 will power 5 Quads (2xGND 1x12V per Quad).

[gr0bi42]

2 Quads up and running in cluster mode. Unfortunately I don't see these "high" freqs 

Code:

001-0: ztex_ufm1_15y1-04A3466370-2: Set frequency to 208.00MHz
001-0: ztex_ufm1_15y1-04A3466370-1: f=208.00MHz,  errorRate=0.23%,  maxErrorRate=1.47%,  hashRate=207.5MH/s,  submitted 16 new nonces,  luckFactor=1.06
001-0: ztex_ufm1_15y1-04A3466370-2: f=208.00MHz,  errorRate=0.00%,  maxErrorRate=0.00%,  hashRate=208.0MH/s,  submitted 21 new nonces,  luckFactor=1.12
001-0: ztex_ufm1_15y1-04A3466370-3: f=200.00MHz,  errorRate=0.00%,  maxErrorRate=0.00%,  hashRate=200.0MH/s,  submitted 13 new nonces,  luckFactor=0.94
001-0: ztex_ufm1_15y1-04A3466370-4: f=204.00MHz,  errorRate=0.00%,  maxErrorRate=0.81%,  hashRate=204.0MH/s,  submitted 8 new nonces,  luckFactor=0.82
001-0: ztex_ufm1_15y1-04A346BE51-1: f=208.00MHz,  errorRate=0.02%,  maxErrorRate=0.52%,  hashRate=207.9MH/s,  submitted 10 new nonces,  luckFactor=0.95
001-0: ztex_ufm1_15y1-04A346BE51-2: f=204.00MHz,  errorRate=0.00%,  maxErrorRate=0.57%,  hashRate=204.0MH/s,  submitted 10 new nonces,  luckFactor=1.12
001-0: ztex_ufm1_15y1-04A346BE51-3: f=208.00MHz,  errorRate=0.01%,  maxErrorRate=1.10%,  hashRate=208.0MH/s,  submitted 14 new nonces,  luckFactor=0.97
001-0: ztex_ufm1_15y1-04A346BE51-4: f=212.00MHz,  errorRate=0.13%,  maxErrorRate=0.97%,  hashRate=211.7MH/s,  submitted 12 new nonces,  luckFactor=0.93
001-0: poll loop time: 65ms (USB: 12ms network: 53ms)   getwork time: 160ms  submit time: 152ms
Total hash rate: 1651.2 MH/s
Total submitted hash rate: 1627.5 MH/s

[Turbor]

2 Quads up and running in cluster mode. Unfortunately I don't see these "high" freqs 

Those that order in large numbers get the slow boards 

Not sure if Alex is using the same chips for his license boards. Perhaps he has SG3 spartans ?

[antirack]

I am using the same FPGA. The passive components are different, but this surely can't be the reason why.

I know this is very far fetched, and it's just a theory, but I am at sea level (actually I am sitting in the 43rd floor right now with my quads buzzing away next to me). At higher altitudes, the air density, fan efficiencies, and heat transfer efficiencies all decrease. I read somewhere that a ballpark figure is 3% for 1000 feet (300 meters). So you guys in the alps are definitely at a disadvantage. The good news is that the new home for my cluster is on the 3rd floor so I should get even higher numbers

Send me one of your boards and I will put it right next to mine in order to confirm this theory.

The quads have been running for 8 days straight now:

Code:

001-0: ztex_ufm1_15y1-0000004-01-1: f=228.00MHz,  errorRate=0.06%,  maxErrorRate=1.32%,  hashRate=227.9MH/s,  submitted 24 new nonces,  luckFactor=1.00
001-0: ztex_ufm1_15y1-0000004-01-2: f=232.00MHz,  errorRate=0.03%,  maxErrorRate=0.95%,  hashRate=231.9MH/s,  submitted 12 new nonces,  luckFactor=0.98
001-0: ztex_ufm1_15y1-0000004-01-3: f=232.00MHz,  errorRate=0.01%,  maxErrorRate=0.99%,  hashRate=232.0MH/s,  submitted 16 new nonces,  luckFactor=1.05
001-0: ztex_ufm1_15y1-0000004-01-4: f=224.00MHz,  errorRate=0.22%,  maxErrorRate=0.85%,  hashRate=223.5MH/s,  submitted 13 new nonces,  luckFactor=0.98

001-0: ztex_ufm1_15y1-0000004-02-1: f=224.00MHz,  errorRate=0.00%,  maxErrorRate=0.71%,  hashRate=224.0MH/s,  submitted 14 new nonces,  luckFactor=1.02
001-0: ztex_ufm1_15y1-0000004-02-2: f=228.00MHz,  errorRate=0.00%,  maxErrorRate=1.84%,  hashRate=228.0MH/s,  submitted 12 new nonces,  luckFactor=0.98
001-0: ztex_ufm1_15y1-0000004-02-3: f=232.00MHz,  errorRate=1.23%,  maxErrorRate=2.27%,  hashRate=229.1MH/s,  submitted 12 new nonces,  luckFactor=1.02
001-0: ztex_ufm1_15y1-0000004-02-4: f=220.00MHz,  errorRate=0.00%,  maxErrorRate=0.00%,  hashRate=220.0MH/s,  submitted 16 new nonces,  luckFactor=1.02

001-0: poll loop time: 141ms (USB: 12ms network: 129ms)   getwork time: 262ms  submit time: 258ms
Total hash rate: 2460.3 MH/s
Total submitted hash rate: 2467.0 MH/s

#1. 915.3 MH/s
#2: 901.1 MH/s

Ambient temperature around 25C right now, no air conditioning, relative humidity at 84% though :/

[Hpman]

I am using the same FPGA. The passive components are different, but this surely can't be the reason why.

I guess you've got a better charge of Spartan6 Devices. Same Speedgrad labeled but inside better.

[Chefnet]

or better cooling, my single went up from 204 to 220 with artic mx-2 under the cooler

[gr0bi42]

or better cooling, my single went up from 204 to 220 with artic mx-2 under the cooler

I think, this could be the reason. The thermal compound and/or the way how you apply it.

I've used Arctic Silver 5. And later today I will post some photos of "how" I apply it.

[ummas]

i`m wodnring with waterblock put on it
i should have 2 or 3 NB waterblocks... still need one or 2 :|

I have lost AS5 and used somme simmilar shit, witch brought me f=196-208
Is it ok to put on it thermapad 0.5mm 7W/mK ??

any sollution to cool VRM section ?

[antirack]

or better cooling, my single went up from 204 to 220 with artic mx-2 under the cooler

I doubt that this is the secret why I get high hash rates.

In my case I have been using Arctic MX-4 thermal compound for my initial installation of the heatsinks on both of my quad boards. See my post on April 24 with picture. Pay special attention how my quad #2 (on the right) does 901 MH/s and which heatsinks I am using.

In the meantime I changed my #2 to stock heatsinks (Xilence/Deep Cool) and used the heatsink compound that came with the coolers. And today, in my previous post, the hashrate on #2 was still 901.1 MH/s. Conclusion: no difference if Arctic MX-4 or the one included in the small bag with the stock heat sinks (at least in my case). Maybe in your case the cooler just didn't sit right (air bubbles?).