- If I understand correctly from the original post the price goes down with volume cross-customers. Is there a page with the current applicable price or recent orders by date so that we can know at which price we will be buying ?
Topic: ZTEX USB-FPGA Modules 1.15x and 1.15y: 215 and 860 MH/s FPGA Boards - page 24. (Read 182443 times)
April 24, 2012, 06:03:02 PM
April 24, 2012, 05:47:59 PM
Quote
However, I have once or twice, when I touched the USB port on a running board (not the USB hub), had communication issues with that board. I am not sure if that's because of the hub or my electrical wiring (current flow) or just because of my prototypes. I am going to test this with an alternative USB hub now that you mention it.
Have these com errors to on a 7 port hub from trust (cheap brand) once in a while. But most times all 5 work flawless with it:)
(it could be the hub, my cables, my mac mini witch has a known problem with usb devices or some random event i cause while touching the stuff on my desk?)
April 24, 2012, 04:38:05 PM
Antirack: The 10port USB hubs your using in your photo (the blue one). Have you had any problems with them at all from a reliability standpoint?
So far no problem. But they are cheap and low quality, I'd keep a few spares handy. They cost very little retail here in Hong Kong (I've seen them for $1.5 from a company in China). They do have a good form factor for a cluster in my opinion.
However, I have once or twice, when I touched the USB port on a running board (not the USB hub), had communication issues with that board. I am not sure if that's because of the hub or my electrical wiring (current flow) or just because of my prototypes. I am going to test this with an alternative USB hub now that you mention it.
Did you have any problems or just asking because you are looking for hubs?
April 24, 2012, 12:58:39 PM
No, just mine. And I can only do that because I have access to a manufacturing facility and components (and I am very motivated to build my own mining cluster).
I hate you so much
The hash rates are very impressive. It looks like i lost a bit of speed by changing my fans. Great product. April 24, 2012, 12:45:39 PM
I'll take the the fan of the coolers off and use the sink as passive 
My setup will be built into a IKEA Helmer
locker.
My setup will be built into a IKEA Helmer
locker.
April 24, 2012, 11:45:13 AM
I didn't even notice the 1.15y came out. Gotta start subscribing to this.
April 24, 2012, 11:29:25 AM
I'm not going to change the fans. I won't even use the 40mm fans. I'm just applying new heat sinks and using 120mm fan for 4 or 5 FPGAs. I can sketch a picture if that helps.
April 24, 2012, 11:13:41 AM
Why switching to new fans? I did ist because I wanted to save some height.
Im switching back to the included heatssinks and use them passive.
Im switching back to the included heatssinks and use them passive.
April 24, 2012, 10:36:37 AM
Comments needed about my possible cooling system.
I install aftermarket Northbridge heat sinks to 1.15y FPGAs. As far as I understand, every Northbridge heat sink fits there or are there some differences with the mounting holes? How much pressure can the FPGA take? Can I use a heat sink that weighs 150g?
I use 1850rpm Gentle Typhoon fans to blow cold air to the heat sinks. FPGAs are placed in line inside of wooden tunnel so the air flows through the heat sinks.
Pros? Cons?
I install aftermarket Northbridge heat sinks to 1.15y FPGAs. As far as I understand, every Northbridge heat sink fits there or are there some differences with the mounting holes? How much pressure can the FPGA take? Can I use a heat sink that weighs 150g?
I use 1850rpm Gentle Typhoon fans to blow cold air to the heat sinks. FPGAs are placed in line inside of wooden tunnel so the air flows through the heat sinks.
Pros? Cons?
April 24, 2012, 09:04:49 AM
Antirack: The 10port USB hubs your using in your photo (the blue one). Have you had any problems with them at all from a reliability standpoint?
April 24, 2012, 08:36:50 AM
Ah ok
April 24, 2012, 08:22:44 AM
Ahhhh nice to know, so you are producing boards for Ztex
No, just mine. And I can only do that because I have access to a manufacturing facility and components (and I am very motivated to build my own mining cluster).
April 24, 2012, 08:19:14 AM
Ahhhh nice to know, so you are producing boards for Ztex
April 24, 2012, 08:16:33 AM
antirack,
what about putting a single big fan over the four low-profile heatsinks?
spiccioli
what about putting a single big fan over the four low-profile heatsinks?
spiccioli
I am sure that would work, but i am going to put them in a case so I don't need any individual fans on the units. I'll just use (case) fans and establish air flow through the case, forcing the air through the heat sinks (similar to servers, channeling the air with styrofoam if I have to).
What did u do to become a beta tester ? Would love to test to!
? Would love to test to!I have paid Ztex for the license and made them in "my own" assembly line here in Hong Kong. The two boards you see are from the test production.
April 24, 2012, 07:53:52 AM
What did u do to become a beta tester ? Would love to test to!
? Would love to test to! April 24, 2012, 06:48:22 AM
antirack,
what about putting a single big fan over the four low-profile heatsinks?
spiccioli
what about putting a single big fan over the four low-profile heatsinks?
spiccioli
April 24, 2012, 06:38:04 AM
Regarding miner support, I should have that on cgminer soon, but since I don't own any 1.15y I'll depend on external testing to assert it does, in fact, work.
April 24, 2012, 06:14:41 AM
2 questions :
- If I understand correctly from the original post the price goes down with volume cross-customers. Is there a page with the current applicable price or recent orders by date so that we can know at which price we will be buying ?
- Which miners do support the 1.15y today and which others are working on its support ?
April 24, 2012, 06:04:36 AM
I have been running two test boards of the 1.15y (and three 1.15x) for a few days now (licensed production).
Total Device #01: 913 MH/s
Total Device #02: 901 MH/s
The last 3 units are 1.15x.
Room temperature is 25C, I am using the stock heatsinks with fans. They are very stable and I have seen even more impressive hash rates (had them running on <20C AC for a while).
Connected to a low cost Atom board with 2GB memory running Ubuntu, using 0% CPU load average 0.00.
Currently all (incl. Atom) powered by a very inefficient 350W Mini ATX PSU. My kill-a-watt device shows 199 Watt at the wall. The Atom board alone is 27 Watt.
It will be interesting to see what happens if I move to a better PSU and passive cooling without the 11 small fans.
In the photo above I was experimenting with different heatsinks. While the ones on the right are easier to mount (the mounts are attached to the heatsink) they were too noisy and so I replaced them with the stock heatsinks after two days (can't take a new photo, logged in remotely via RD then on to VNC to the Ubuntu box).
Code:
001-0: ztex_ufm1_15y1-0000004-01-1: f=228.00MHz, errorRate=0.00%, maxErrorRate=1.07%, hashRate=228.0MH/s, submitted 17 new nonces, luckFactor=0.96
001-0: ztex_ufm1_15y1-0000004-01-2: f=232.00MHz, errorRate=0.00%, maxErrorRate=0.50%, hashRate=232.0MH/s, submitted 13 new nonces, luckFactor=1.02
001-0: ztex_ufm1_15y1-0000004-01-3: f=232.00MHz, errorRate=0.54%, maxErrorRate=0.99%, hashRate=230.7MH/s, submitted 12 new nonces, luckFactor=0.96
001-0: ztex_ufm1_15y1-0000004-01-4: f=224.00MHz, errorRate=0.25%, maxErrorRate=0.97%, hashRate=223.4MH/s, submitted 18 new nonces, luckFactor=0.98
001-0: ztex_ufm1_15y1-0000004-02-1: f=224.00MHz, errorRate=0.00%, maxErrorRate=0.65%, hashRate=224.0MH/s, submitted 12 new nonces, luckFactor=1.00
001-0: ztex_ufm1_15y1-0000004-02-2: f=228.00MHz, errorRate=0.33%, maxErrorRate=1.38%, hashRate=227.2MH/s, submitted 10 new nonces, luckFactor=1.05
001-0: ztex_ufm1_15y1-0000004-02-3: f=232.00MHz, errorRate=0.69%, maxErrorRate=2.62%, hashRate=230.4MH/s, submitted 14 new nonces, luckFactor=1.00
001-0: ztex_ufm1_15y1-0000004-02-4: f=220.00MHz, errorRate=0.00%, maxErrorRate=0.00%, hashRate=220.0MH/s, submitted 7 new nonces, luckFactor=1.00
001-0: ztex_ufm1_15d4-04A3469722-1: f=216.00MHz, errorRate=0.00%, maxErrorRate=0.66%, hashRate=216.0MH/s, submitted 13 new nonces, luckFactor=0.97
001-0: ztex_ufm1_15d4-04A346CEC7-1: f=216.00MHz, errorRate=0.08%, maxErrorRate=1.19%, hashRate=215.8MH/s, submitted 11 new nonces, luckFactor=0.94
001-1: ztex_ufm1_15d4-04A32E00E9-1: f=212.00MHz, errorRate=0.00%, maxErrorRate=1.35%, hashRate=212.0MH/s, submitted 15 new nonces, luckFactor=1.00
001-0: poll loop time: 123ms (USB: 11ms network: 112ms) getwork time: 262ms submit time: 260ms
001-1: poll loop time: 14ms (USB: 0ms network: 14ms) getwork time: 391ms submit time: 337ms
Total hash rate: 2459.6 MH/s
Total submitted hash rate: 2421.3 MH/s
--------
001-0: ztex_ufm1_15y1-0000004-01-2: f=232.00MHz, errorRate=0.00%, maxErrorRate=0.50%, hashRate=232.0MH/s, submitted 13 new nonces, luckFactor=1.02
001-0: ztex_ufm1_15y1-0000004-01-3: f=232.00MHz, errorRate=0.54%, maxErrorRate=0.99%, hashRate=230.7MH/s, submitted 12 new nonces, luckFactor=0.96
001-0: ztex_ufm1_15y1-0000004-01-4: f=224.00MHz, errorRate=0.25%, maxErrorRate=0.97%, hashRate=223.4MH/s, submitted 18 new nonces, luckFactor=0.98
001-0: ztex_ufm1_15y1-0000004-02-1: f=224.00MHz, errorRate=0.00%, maxErrorRate=0.65%, hashRate=224.0MH/s, submitted 12 new nonces, luckFactor=1.00
001-0: ztex_ufm1_15y1-0000004-02-2: f=228.00MHz, errorRate=0.33%, maxErrorRate=1.38%, hashRate=227.2MH/s, submitted 10 new nonces, luckFactor=1.05
001-0: ztex_ufm1_15y1-0000004-02-3: f=232.00MHz, errorRate=0.69%, maxErrorRate=2.62%, hashRate=230.4MH/s, submitted 14 new nonces, luckFactor=1.00
001-0: ztex_ufm1_15y1-0000004-02-4: f=220.00MHz, errorRate=0.00%, maxErrorRate=0.00%, hashRate=220.0MH/s, submitted 7 new nonces, luckFactor=1.00
001-0: ztex_ufm1_15d4-04A3469722-1: f=216.00MHz, errorRate=0.00%, maxErrorRate=0.66%, hashRate=216.0MH/s, submitted 13 new nonces, luckFactor=0.97
001-0: ztex_ufm1_15d4-04A346CEC7-1: f=216.00MHz, errorRate=0.08%, maxErrorRate=1.19%, hashRate=215.8MH/s, submitted 11 new nonces, luckFactor=0.94
001-1: ztex_ufm1_15d4-04A32E00E9-1: f=212.00MHz, errorRate=0.00%, maxErrorRate=1.35%, hashRate=212.0MH/s, submitted 15 new nonces, luckFactor=1.00
001-0: poll loop time: 123ms (USB: 11ms network: 112ms) getwork time: 262ms submit time: 260ms
001-1: poll loop time: 14ms (USB: 0ms network: 14ms) getwork time: 391ms submit time: 337ms
Total hash rate: 2459.6 MH/s
Total submitted hash rate: 2421.3 MH/s
--------
Total Device #01: 913 MH/s
Total Device #02: 901 MH/s
The last 3 units are 1.15x.
Room temperature is 25C, I am using the stock heatsinks with fans. They are very stable and I have seen even more impressive hash rates (had them running on <20C AC for a while).
Connected to a low cost Atom board with 2GB memory running Ubuntu, using 0% CPU load average 0.00.
Currently all (incl. Atom) powered by a very inefficient 350W Mini ATX PSU. My kill-a-watt device shows 199 Watt at the wall. The Atom board alone is 27 Watt.
It will be interesting to see what happens if I move to a better PSU and passive cooling without the 11 small fans.
In the photo above I was experimenting with different heatsinks. While the ones on the right are easier to mount (the mounts are attached to the heatsink) they were too noisy and so I replaced them with the stock heatsinks after two days (can't take a new photo, logged in remotely via RD then on to VNC to the Ubuntu box).
April 24, 2012, 04:54:37 AM
The initial posting (https://bitcointalksearch.org/topic/ztex-usb-fpga-modules-115x-and-115y-215-and-860-mhs-fpga-boards-49180) has been updated: the new quad was added. Its now available in the shop at http://shop.ztex.de/product_info.php?cPath=21&products_id=74
Jump to: