[ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs - page 638.

RoadStress

legendary

Activity: 1904

Merit: 1007

Quote from: Grix on May 09, 2014, 04:42:37 AM

Quote from: Guy Corem on May 08, 2014, 05:52:07 AM

Quote from: Grix on May 08, 2014, 05:22:02 AM

So what are the odds of the first tape out having a design flaw that will make us have to wait through a second fab?

All ready discussed at length. We'll ship on time and on spec.

I have an SP30 order from you, and I want to believe.. But if you are so sure, why is this in your Terms of Conditions:

Quote

Delivery dates communicated or acknowledged by Spondoolies in the Order Confirmation are approximate only, and are dependent upon supply chain interruptions and stoppages, development delays, and other factors. Therefore, Purchaser shall not rely upon, and Spondoolies shall not be liable under whatsoever circumstances for any other estimation statement of expectation when delivery will occur.

How did the terms helped customers of: BFL, AMT, TerraHash, HashFail, CoinTerra, scammy VMC/AMC, etc customers? It didn't! Why are you stressing over this? Spondoolies Tech is the only company that funded itself for over 9 months and they launched a product without taking any customer money upfront. They are here to stay and obviously they don't plan on screwing their customers like the mentioned companies above. Do you really think that a company that was funded by 2 VCs firms will screw their customers? Why can't people see the differences over ALL other companies?

Grix

hero member

Activity: 536

Merit: 500

Quote from: Guy Corem on May 08, 2014, 05:52:07 AM

Quote from: Grix on May 08, 2014, 05:22:02 AM

So what are the odds of the first tape out having a design flaw that will make us have to wait through a second fab?

All ready discussed at length. We'll ship on time and on spec.

I have an SP30 order from you, and I want to believe.. But if you are so sure, why is this in your Terms of Conditions:

Quote

Delivery dates communicated or acknowledged by Spondoolies in the Order Confirmation are approximate only, and are dependent upon supply chain interruptions and stoppages, development delays, and other factors. Therefore, Purchaser shall not rely upon, and Spondoolies shall not be liable under whatsoever circumstances for any other estimation statement of expectation when delivery will occur.

Collider

hero member

Activity: 714

Merit: 500

These are server grade PSUs, with more than enough airflow through them even at maximum load.

Spondoolies stated earlier, that they have checked with the manufacturer, receiving green light for 1250W at the PSU.
Typically the PSU only draws 1300W at the wall, this means around 1170W at the PSU (12V).

These PSUs are designed with considerable safety margins in mind. Even desktop PSUs can typically provide around 10% more power.

loshia

legendary

Activity: 1610

Merit: 1000

Quote from: whonesta on May 08, 2014, 08:34:41 PM

Are the SP10's running in turbo mode @ 1.4TH/s using 1325W?
If this is true do these units have an upgraded PSU or standard 1050W?
Does running in "turbo mode" burn-up the PSU?

I am very interested but can not get an answer as to how SP10's are pulling 1300+ watts from a 1050W rated PSU?

Dude,
Every psu rated at x wats pulls x multiplied buy 1.15 depending on efficiency at/in the wall. At full load of course.
These are basics and have nothing to do with sp10 at all
Basically you ca burn any psu when overloading it

whonesta

sr. member

Activity: 379

Merit: 250

Are the SP10's running in turbo mode @ 1.4TH/s using 1325W?
If this is true do these units have an upgraded PSU or standard 1050W?
Does running in "turbo mode" burn-up the PSU?

I am very interested but can not get an answer as to how SP10's are pulling 1300+ watts from a 1050W rated PSU?

zvisha

hero member

Activity: 560

Merit: 500

Quote from: DPoS on May 08, 2014, 12:58:22 PM

Quote from: Guy Corem on May 08, 2014, 12:39:15 PM

Quote from: DPoS on May 08, 2014, 12:33:57 PM

Quote from: Guy Corem on May 08, 2014, 11:52:56 AM

Quote from: stan258 on May 08, 2014, 11:30:07 AM

Spondoolies Tech. This came up in another thread. It is in regards to not solving enough blocks during testing. FYI link below.

https://bitcointalk.org/index.php?topic=49417.new;topicseen#new

Funny. We have less than 5 miners at our possession for SW testing now.
Unfortunately, we didn't sell enough miners to make a difference.

We're hearing other rumors about concentrated block withholds attacks on the public pools.

you only need one to test with correct? can you explain your test cases?

Incorrect. We're doing long tests with various firmwares versions under various power consumption scenarios.
One miner is definitely not enough.

yes more units are best for environmental and stability testing

what about your functionality test cases? using a test pool with test work? surely you are not just going to wait until you solo mine a block to check that off your list

No, we didn't validate solo mining. However, you can see in our code in GIT that we support solo-mining by using lower, tested leading zeroes and letting cgminer clean the unneeded results.
However, the pool attack is happening in non-solo mining, with regular pool difficulty of leading zeroes between 38 and 42. That scenario was tested.

DPoS

sr. member

Activity: 462

Merit: 250

Quote from: Guy Corem on May 08, 2014, 12:39:15 PM

Quote from: DPoS on May 08, 2014, 12:33:57 PM

Quote from: Guy Corem on May 08, 2014, 11:52:56 AM

Quote from: stan258 on May 08, 2014, 11:30:07 AM

Spondoolies Tech. This came up in another thread. It is in regards to not solving enough blocks during testing. FYI link below.

https://bitcointalk.org/index.php?topic=49417.new;topicseen#new

Funny. We have less than 5 miners at our possession for SW testing now.
Unfortunately, we didn't sell enough miners to make a difference.

We're hearing other rumors about concentrated block withholds attacks on the public pools.

you only need one to test with correct? can you explain your test cases?

Incorrect. We're doing long tests with various firmwares versions under various power consumption scenarios.
One miner is definitely not enough.

yes more units are best for environmental and stability testing

what about your functionality test cases? using a test pool with test work? surely you are not just going to wait until you solo mine a block to check that off your list

Guy Corem

donator

Activity: 1414

Merit: 1051

Spondoolies, Beam & DAGlabs

Quote from: DPoS on May 08, 2014, 12:33:57 PM

Quote from: Guy Corem on May 08, 2014, 11:52:56 AM

Quote from: stan258 on May 08, 2014, 11:30:07 AM

Spondoolies Tech. This came up in another thread. It is in regards to not solving enough blocks during testing. FYI link below.

https://bitcointalk.org/index.php?topic=49417.new;topicseen#new

Funny. We have less than 5 miners at our possession for SW testing now.
Unfortunately, we didn't sell enough miners to make a difference.

We're hearing other rumors about concentrated block withholds attacks on the public pools.

you only need one to test with correct? can you explain your test cases?

Incorrect. We're doing long tests with various firmwares versions under various power consumption scenarios.
One miner is definitely not enough.

DPoS

sr. member

Activity: 462

Merit: 250

Quote from: Guy Corem on May 08, 2014, 11:52:56 AM

Quote from: stan258 on May 08, 2014, 11:30:07 AM

Spondoolies Tech. This came up in another thread. It is in regards to not solving enough blocks during testing. FYI link below.

https://bitcointalk.org/index.php?topic=49417.new;topicseen#new

Funny. We have less than 5 miners at our possession for SW testing now.
Unfortunately, we didn't sell enough miners to make a difference.

We're hearing other rumors about concentrated block withholds attacks on the public pools.

you only need one to test with correct? can you explain your test cases?

Guy Corem

donator

Activity: 1414

Merit: 1051

Spondoolies, Beam & DAGlabs

Quote from: stan258 on May 08, 2014, 11:30:07 AM

Spondoolies Tech. This came up in another thread. It is in regards to not solving enough blocks during testing. FYI link below.

https://bitcointalk.org/index.php?topic=49417.new;topicseen#new

Funny. We have less than 5 miners at our possession for SW testing now.
Unfortunately, we didn't sell enough miners to make a difference.

We're hearing other rumors about concentrated block withholds attacks on the public pools.

stan258

sr. member

Activity: 440

Merit: 250

Spondoolies Tech. This came up in another thread. It is in regards to not solving enough blocks during testing. FYI link below.

https://bitcointalk.org/index.php?topic=49417.new;topicseen#new

hardhouseinc

sr. member

Activity: 672

Merit: 250

Most Advanced Crypto Exchange on the Blockchain

Quote from: calichomp on May 07, 2014, 10:07:40 PM

Quote from: zvisha on May 07, 2014, 04:41:09 PM

Quote from: DevonMiner on May 07, 2014, 04:15:21 PM

Quote from: zvisha on May 07, 2014, 04:00:42 AM

Quote from: seriouscoin on May 06, 2014, 08:54:33 PM

Spondoolies, does the SP10 firmware check fans activity?

I'm about to test my air chamber and i'm ready to remove the fans (PSU fan is still intact)

Let me know if its a problem with the FW.

Thanks

No, we monitor the temperature of ASICs/on board, not the fan activity.
Good luck!

Is this something you might look at in the future? I have read the fans run fulltime 24/7 ... can you update the firmware so they are PWM regulated? I'm sure this was mentioned way back in the thread, sorry if I missed something ... two SP10's arriving soon, so I'm interested if the noise can be 'contained'.

Do you mean control the FANs based on system temperature?

From tests we did, it is very hard to even define wanted behaviour to control the fans based on temperature.
The fan speed effects the hash-rate. It's not binary "too cold/too hot" system. The more you turn on the fans, the colder the ASICs, the less leakage you have, the bigger the hash-rate. Even if we set a desired temperature "threshold", it's really annoying when the system starts changing the fans. It's easier to set "quiet/noisy" based on location of the miner in your apartment and then have no change.

You can write to PWM values once the system is stable (just be careful of course or it will thermal shut-down) - it's in this thread (and in the miner-gate code) how to do that.

I plan to unplug the fans entirely. Call it coarse PWM control. But, it has worked well for my Terraminers and Baby Jets.

Explain coarse PWM control?
What does the cooling?
If it can keep it at below 25 degrees and makes it quieter Id
like to understand the concept??

Guy Corem

donator

Activity: 1414

Merit: 1051

Spondoolies, Beam & DAGlabs

Quote from: Grix on May 08, 2014, 05:22:02 AM

So what are the odds of the first tape out having a design flaw that will make us have to wait through a second fab?

All ready discussed at length. We'll ship on time and on spec.

Grix

hero member

Activity: 536

Merit: 500

So what are the odds of the first tape out having a design flaw that will make us have to wait through a second fab?

calichomp

newbie

Activity: 17

Merit: 0

Quote from: zvisha on May 07, 2014, 04:41:09 PM

Quote from: DevonMiner on May 07, 2014, 04:15:21 PM

Quote from: zvisha on May 07, 2014, 04:00:42 AM

Quote from: seriouscoin on May 06, 2014, 08:54:33 PM

Spondoolies, does the SP10 firmware check fans activity?

I'm about to test my air chamber and i'm ready to remove the fans (PSU fan is still intact)

Let me know if its a problem with the FW.

Thanks

No, we monitor the temperature of ASICs/on board, not the fan activity.
Good luck!

Is this something you might look at in the future? I have read the fans run fulltime 24/7 ... can you update the firmware so they are PWM regulated? I'm sure this was mentioned way back in the thread, sorry if I missed something ... two SP10's arriving soon, so I'm interested if the noise can be 'contained'.

Do you mean control the FANs based on system temperature?

From tests we did, it is very hard to even define wanted behaviour to control the fans based on temperature.
The fan speed effects the hash-rate. It's not binary "too cold/too hot" system. The more you turn on the fans, the colder the ASICs, the less leakage you have, the bigger the hash-rate. Even if we set a desired temperature "threshold", it's really annoying when the system starts changing the fans. It's easier to set "quiet/noisy" based on location of the miner in your apartment and then have no change.

You can write to PWM values once the system is stable (just be careful of course or it will thermal shut-down) - it's in this thread (and in the miner-gate code) how to do that.

I plan to unplug the fans entirely. Call it coarse PWM control. But, it has worked well for my Terraminers and Baby Jets.

dogie

legendary

Activity: 1666

Merit: 1185

dogiecoin.com

Quote from: wanna4fun on May 07, 2014, 09:02:47 AM

Quote from: Biodom on May 06, 2014, 07:49:27 PM

Quote from: wanna4fun on May 06, 2014, 02:54:38 AM

I can't imagine letting one SP10 running around me. If you put it on your balcony, the neighborhood will hate you.

can anybody post decibel reading of SP-10 at 1.4-1.5 th/s within a feet or so? There are free android and apple apps that do it.
I am just trying to see how noisy (or not) it is in comparison with S1.

What ambient temperature would Sp-10 tolerate?
i have an extra room, but when i put 3 S1 there and closed the door, temp rose to ~35 degree celsius, so I had to open the door. obviously, if SP10 is like a rock band, it would be difficult.
I figure that heat-wise 1 SP10=3 S1 (with much higher productivity for Sp-10, of course)

it's max 35°C ans as said by SP-Tech it can run with ambiant air @45°C (they said it worked for them). Don't go to hot or you will lose a lot of hashrate. Plus, I want to say that if air go in @ 22-23°C it comes out @45-46°C. You must evacuate all of this hot air. I bet you can do something fancy with this hot air like powering a Stirling engine Cheesy

The noise is about 70 dB within 3 feets. It's like 3/4 vacuum cleaner at the same time and I underestimate the reality. It's a high pitch fan noise.

I did this in my thread (see sig)

DevonMiner

sr. member

Activity: 471

Merit: 250

Quote from: RoadStress on May 07, 2014, 05:14:21 PM

Quote from: DevonMiner on May 07, 2014, 05:04:17 PM

Any shipping dates for May delivery of SP10's available yet? Will be interesting to play with the settings once they arrive.

If you join my GB the shipping starts on 16th May. (or maybe a bit earlier for the early birds)

Lol, I did .. for the SP30 ... SP10 ... 16th it is hopefully ...

RoadStress

legendary

Activity: 1904

Merit: 1007

Quote from: DevonMiner on May 07, 2014, 05:04:17 PM

Any shipping dates for May delivery of SP10's available yet? Will be interesting to play with the settings once they arrive.

If you join my GB the shipping starts on 16th May. (or maybe a bit earlier for the early birds)

DevonMiner

sr. member

Activity: 471

Merit: 250

Quote

Do you mean control the FANs based on system temperature?

From tests we did, it is very hard to even define wanted behaviour to control the fans based on temperature.
The fan speed effects the hash-rate. It's not binary "too cold/too hot" system. The more you turn on the fans, the colder the ASICs, the less leakage you have, the bigger the hash-rate. Even if we set a desired temperature "threshold", it's really annoying when the system starts changing the fans. It's easier to set "quiet/noisy" based on location of the miner in your apartment and then have no change.

You can write to PWM values once the system is stable (just be careful of course or it will thermal shut-down) - it's in this thread (and in the miner-gate code) how to do that.

Thanks zvisha,

I get the fact that the faster the fans run, the cooler the system runs ... so higher the hash rate, I'm coming at this from a none DC angle, so hosting the units my end. Guess I'm used to running a load of Antminer S1's that run 'loud and fast' and then 'run quiet' as and when they are required to do so. SP10 is a different beast I know.

Any shipping dates for May delivery of SP10's available yet? Will be interesting to play with the settings once they arrive.

zvisha

hero member

Activity: 560

Merit: 500

Quote from: DevonMiner on May 07, 2014, 04:15:21 PM

Quote from: zvisha on May 07, 2014, 04:00:42 AM

Quote from: seriouscoin on May 06, 2014, 08:54:33 PM

Spondoolies, does the SP10 firmware check fans activity?

I'm about to test my air chamber and i'm ready to remove the fans (PSU fan is still intact)

Let me know if its a problem with the FW.

Thanks

No, we monitor the temperature of ASICs/on board, not the fan activity.
Good luck!

Is this something you might look at in the future? I have read the fans run fulltime 24/7 ... can you update the firmware so they are PWM regulated? I'm sure this was mentioned way back in the thread, sorry if I missed something ... two SP10's arriving soon, so I'm interested if the noise can be 'contained'.

Do you mean control the FANs based on system temperature?

From tests we did, it is very hard to even define wanted behaviour to control the fans based on temperature.
The fan speed effects the hash-rate. It's not binary "too cold/too hot" system. The more you turn on the fans, the colder the ASICs, the less leakage you have, the bigger the hash-rate. Even if we set a desired temperature "threshold", it's really annoying when the system starts changing the fans. It's easier to set "quiet/noisy" based on location of the miner in your apartment and then have no change.

You can write to PWM values once the system is stable (just be careful of course or it will thermal shut-down) - it's in this thread (and in the miner-gate code) how to do that.

Topic: [ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs - page 638. (Read 1260389 times)