[ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs

[jtoomim]

From our chats with datacenter owners,  the over 3KW of heat that the SP35 puts out often is too much.  

Only if you're not designed for it. We've got a 48U rack running 73.5 kW at our datacenter on pure air cooling with no trouble. It has 15 SP35s and 7 SP30/SP31s. The idea that you have to go to water cooling in order to achieve high densities is both expensive and false.

[1713997533]

1713997533

[Biffa]

From our chats with datacenter owners,  the over 3KW of heat that the SP35 puts out often is too much.  

Only if you're not designed for it. We've got a 48U rack running 73.5 kW at our datacenter on pure air cooling with no trouble. It has 15 SP35s and 7 SP30/SP31s. The idea that you have to go to water cooling in order to achieve high densities is both expensive and false.

To be fair, depends on the datacenter owners they have been talking to, if they are traditional hot/cold aisle airconned DC's then they have a point, yours, awesome as it is, is pretty specifically designed and build for mining hardware.

[djroadking]

My next bit of gear will be from SPtech or MinersEdge... Whoever launches next gen of hardware first and depends on power usage and stuff.

[MinerEdge]

From our chats with datacenter owners,  the over 3KW of heat that the SP35 puts out often is too much.  

Only if you're not designed for it. We've got a 48U rack running 73.5 kW at our datacenter on pure air cooling with no trouble. It has 15 SP35s and 7 SP30/SP31s. The idea that you have to go to water cooling in order to achieve high densities is both expensive and false.

You are correct.   It's not that we couldn't do a max density air cooled solution.  Spondoolies already does this.   It's that our architecture is designed to be upgradeable.  Some design concessions had to be made.

Water cooling also opens up the hottest parts of the world to mining.  We can easily handle 40c+ ambient air temps.   And have tested to 70c exiting water temps if so desired for industrial waste heat usage.  Higher temps may be possible.   With a serial power design this will easily be achievable.

-Nathan

[TuffToad]

Hello All.

This is the Miner Edge LLC official Bitcointalk user.

Mr. Ho

Thank you Sir and I thank Spondoolies-Tech (Official Chat forum) for giving this community a heads up about your upcoming products.

Why don't we move this discussion to a "Miner Edge" forum and copy/links that are relevant.   

I look forward to what MinerEdge, LLC. is bringing to the table.


P.S. Wish the best in Huston this week @ Blockchain Innovation  ( http://Texasbitcoinconference.com)


  /s/ TuffToad
Louisville, Ky U.S.A

[dmwardjr]

Thank you, Sir!  Yes, it was a good run for him.

Thanks for the kind words David,

My condolences to your grandfather.  97 is a good run!  

-Nathan

Like what you are doing, Sir.  Wish I could be in Austin.  Unfortunately, I have my grandfather's funeral to attend Saturday.  He lived 1.5 months shy of his 97th birthday.  He was a great man!

I like the idea and the architecture!  I like the idea of swappable modules for upgrading!  Keep up the good work and thank you for your contribution, Sir!

Kind regards,

David

[Guy Corem]

Guy,  Will you do the honors of confirming that this is the real Miner Edge.    We emailed earlier about shipping a broken SP35 for serial power testing.

Confirmed.

[HerbPean]

LOOP[2] ON TO:0 (w:79)
 4: DC2DC/1/:[vlt1:660 vlt2:664(DCl:794 Tl:664 Ul:727) 12W  18A  87c] ASIC:[ 85c (125c) 770hz(BL: 770)   35 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:724 vlt2:727(DCl:794 Tl:794 Ul:727) 106W 146A  60c] ASIC:[ 85c (125c) 1050hz(BL:1050)   44 (E:193) F:0 L:0]

What is wrong ? That ASIC was all good before tonight ... :-(

then after like 15 minutes ..

LOOP[0] OFF TO:0 (test serial failed)
0: disabled (serial problem)
1: disabled (serial problem)

same behaviour after 2-3 reboot (power off/on)

[RoadStress]

MinerEdge rack mounted system based on RockerBox pictures can be found here http://imgur.com/a/5g0cq

Hello All.

This is the Miner Edge LLC official Bitcointalk user.

Welcome to the forums. Very beautiful and clean design and good idea with the upgradable parts!

[Searing]

Hello All.

This is the Miner Edge LLC official Bitcointalk user.

Mr. Ho

Thank you Sir and I thank Spondoolies-Tech (Official Chat forum) for giving this community a heads up about your upcoming products.

Why don't we move this discussion to a "Miner Edge" forum and copy/links that are relevant.   

I look forward to what MinerEdge, LLC. is bringing to the table.


P.S. Wish the best in Huston this week @ Blockchain Innovation  ( http://Texasbitcoinconference.com)


  /s/ TuffToad
Louisville, Ky U.S.A

is there a "miners edge' forum here on bitcointalk yet as a thread (I can't seem to find them but then again....I can rarely find my way home)

shoot us a link (or start one up)

[bomberb17]

Anyone know how to install no-ip (or some other free ddns service) on a SPxx for remote administration?

Don't you need to set NoIP on your router and do some port forwarding from there to you SPxx? I recently setup my home web-server (not on port 80) with no-ip and all I had to do was add a game / application on the router and ensure that the server was running on a static IP (or set the always use this IP if DHCP on router ... though may cause conflicts). I also turned off upnp.

The problem is that my router doesn't support no-ip..

Then you'll need to install the No-IP app either on a PC that you keep on all the time you want access to your SPxx or on the SPxx itself (not sure whether they have one, but you can always look on the web for 3rd party solutions, even code!). If you go the former route, aka the official app, then NoIP have a very good starter guide which should make the process trivial.

That's what I'm looking for, to install no-ip on the SPxx itself. It would be an easy task on a linux distro like ubuntu or debian, but not sure how to do it on this embedded linux SPT uses..

build instructions - no support.
http://docdroid.net/v1k4

Ok done these steps on my Ubuntu VM, what would be the next step after ""make deploy"" for installing no-ip to my machine?

Very nice you now have the dev environment for the SPxx, now all that is left is the routine to update the NoIP servers with your ip. NoIP have an api for that here. All you need to do now (in code to include on the SPxx) is obtain the SPxx gateway IP (i.e your internet IP NOT the SPxx IP) and periodically update their servers on where to find you. Theres also an arduino forum discussion on implementing NoIP that may interest you here.

PS. I have to say you can implement that routine above in pearl (or any scripting language the SPxx OS supports) rather than the arduino way, so just provided for the interfacing with NoIP.

So what's next after "make deploy"? Can someone help on using the dev environment for SPxx?

[RealMalatesta]

Just wanted to share the efficiency of my SP20 here:



Gee, wish I could get more of them....

[Biffa]

From our chats with datacenter owners,  the over 3KW of heat that the SP35 puts out often is too much.  

Only if you're not designed for it. We've got a 48U rack running 73.5 kW at our datacenter on pure air cooling with no trouble. It has 15 SP35s and 7 SP30/SP31s. The idea that you have to go to water cooling in order to achieve high densities is both expensive and false.

You are correct.   It's not that we couldn't do a max density air cooled solution.  Spondoolies already does this.   It's that our architecture is designed to be upgradeable.  Some design concessions had to be made.

Water cooling also opens up the hottest parts of the world to mining.  We can easily handle 40c+ ambient air temps.   And have tested to 70c exiting water temps if so desired for industrial waste heat usage.  Higher temps may be possible.   With a serial power design this will easily be achievable.

-Nathan

Can you outline where/how the radiators fit into the setup? I imagine they are external, so is the unit really a 1U unit or is it 1U plus radiator?

Also could you outline what your expected performance output will be per unit? An idea of price would be good too

As suggested already, best to open up a new thread with your details/announcement.

[HerbPean]

LOOP[2] ON TO:0 (w:79)
 4: DC2DC/1/:[vlt1:660 vlt2:664(DCl:794 Tl:664 Ul:727) 12W  18A  87c] ASIC:[ 85c (125c) 770hz(BL: 770)   35 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:724 vlt2:727(DCl:794 Tl:794 Ul:727) 106W 146A  60c] ASIC:[ 85c (125c) 1050hz(BL:1050)   44 (E:193) F:0 L:0]

What is wrong ? That ASIC was all good before tonight ... :-(

then after like 15 minutes ..

LOOP[0] OFF TO:0 (test serial failed)
0: disabled (serial problem)
1: disabled (serial problem)

same behaviour after 2-3 reboot (power off/on)

Allright ... so now I made a few tests with another power supply.

As soon as I put power in connector #4 ... 3 and 4 doesn't startup.

If I put power only in the connector 3 ... 3 is working.

Smells ... really bad (smells actually nothing ... just that I have a bad feeling)

EDIT: well in fact the PSU do not start if there a connector plugged in the connector 4. Tested with one PSU and two PSU... I think i have DC2DC chip issue ... Like ASIC 4 ... i think the DC2DC chip is overheating.

[RealMalatesta]

One of my SP31 is not behaving very nicely. Actually, it is hashing with a maximum of about 1 Th. I have access via ssh, and looking in the logs, I see the following - well, it asks me to disable "one by one". How can this be done through ssh? Or has anybody an advice what I should do?

Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 15: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 16: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 17: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 6 !^[[0;0m
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 18: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 19: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 20: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 7 !^[[0;0m
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 21: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 22: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 23: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 8 !^[[0;0m
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 24: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 25: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 26: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 9 !^[[0;0m
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 27: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 28: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 29: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Testing LOOPs done = ret=20
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Data Timeout on read fffe:70 (1), (problem:20)
Mar 28 00:06:22 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:5^[[0;0mMar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:3^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:4^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:5^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:6^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:7^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:8^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:9^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:10^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:11^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:12^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:13^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mdc2dc i2c error ASIC:14^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 0 !^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 1 !^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 2 !^[[0;0m
Mar 28 00:06:23 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 3 !^[[0;0m
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 4 !^[[0;0m
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 5 !^[[0;0m
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 6 !^[[0;0m
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 7 !^[[0;0m
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 8 !^[[0;0m
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: ^[[31mERROR: Loop failed 9 !^[[0;0m
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: Testing LOOPs done = ret=25
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: ^[[31mBad loop 0 = trying power cycle^[[0;0m
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: PSU restart
Mar 28 00:06:24 miner local1.warn minergate[1737]: Critical: PSU restart
Mar 28 00:06:28 miner local1.warn minergate[1737]: Critical: PSU restart
Mar 28 00:06:28 miner local1.warn minergate[1737]: Critical: PSU restart

[s1gs3gv]

Hello All.

Hi. Will you be showing one of your new machines down town tomorrow here in Austin ?

[MinerEdge]

Yes. 

-Nathan

Hello All.

Hi. Will you be showing one of your new machines down town tomorrow here in Austin ?

[klondike_bar]

One of my SP31 is not behaving very nicely. Actually, it is hashing with a maximum of about 1 Th. I have access via ssh, and looking in the logs, I see the following - well, it asks me to disable "one by one". How can this be done through ssh? Or has anybody an advice what I should do?

Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 15: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 16: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 17: Loop RT loss - disable us one-by-one to find the bad one

I have one like that where it seems that its only stable once i manually disable at least one asic from the 'trouble loop', and it generally disables 2-3 more for failing BIST, then becomes reasonably stable.

go to settings, and one of the lower categories allows you to DISABLE specific chips. disable 1-2 at a time fromt he trouble loop, as the issue is sometimes a chip and sometimes poor power delivery

[RealMalatesta]

One of my SP31 is not behaving very nicely. Actually, it is hashing with a maximum of about 1 Th. I have access via ssh, and looking in the logs, I see the following - well, it asks me to disable "one by one". How can this be done through ssh? Or has anybody an advice what I should do?

Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 15: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 16: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 17: Loop RT loss - disable us one-by-one to find the bad one

I have one like that where it seems that its only stable once i manually disable at least one asic from the 'trouble loop', and it generally disables 2-3 more for failing BIST, then becomes reasonably stable.

go to settings, and one of the lower categories allows you to DISABLE specific chips. disable 1-2 at a time fromt he trouble loop, as the issue is sometimes a chip and sometimes poor power delivery

I _guess_ it is poor power delivery.  Right now I see it running with 4.2 th/s.

Nevertheless, the issue is that I have only ssh-access to the miner at the moment. In /etc/mnt I see the file "mg_disabled_asics" with the following content:

0:0 1:0 2:0
3:0 4:0 5:0
6:0 7:0 8:0
9:0 10:0 11:0
12:0 13:0 14:0
15:0 16:0 17:0
18:0 19:0 20:0
21:0 22:0 23:0
24:0 25:0 26:0
27:0 28:0 29:0

I assume that it should be possible to disable the ASICs here. Or am I totally wrong? And IF it is possible to disable asics, would "1" mean disabled and "2" removed?

[Biodom]

One of my SP31 is not behaving very nicely. Actually, it is hashing with a maximum of about 1 Th. I have access via ssh, and looking in the logs, I see the following - well, it asks me to disable "one by one". How can this be done through ssh? Or has anybody an advice what I should do?

Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 15: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 16: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 17: Loop RT loss - disable us one-by-one to find the bad one

I have one like that where it seems that its only stable once i manually disable at least one asic from the 'trouble loop', and it generally disables 2-3 more for failing BIST, then becomes reasonably stable.

go to settings, and one of the lower categories allows you to DISABLE specific chips. disable 1-2 at a time fromt he trouble loop, as the issue is sometimes a chip and sometimes poor power delivery

I _guess_ it is poor power delivery.  Right now I see it running with 4.2 th/s.

Nevertheless, the issue is that I have only ssh-access to the miner at the moment. In /etc/mnt I see the file "mg_disabled_asics" with the following content:

0:0 1:0 2:0
3:0 4:0 5:0
6:0 7:0 8:0
9:0 10:0 11:0
12:0 13:0 14:0
15:0 16:0 17:0
18:0 19:0 20:0
21:0 22:0 23:0
24:0 25:0 26:0
27:0 28:0 29:0

I assume that it should be possible to disable the ASICs here. Or am I totally wrong? And IF it is possible to disable asics, would "1" mean disabled and "2" removed?

you can disable individual asics in gui