Author

Topic: [ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs - page 716. (Read 1260005 times)

legendary
Activity: 896
Merit: 1000
Hi All,

Just to confirm I have received the tracking number for my evaluation unit!

For obvious reasons I wont tell you the tracking number but it has just passed clearance and is about to get put on a plane.

Neil
legendary
Activity: 2576
Merit: 1186
Changing the default firmware will void warranty of course.
Just wondering, but does anyone know if this is recognised by Israel?
At least in the US, companies are not allowed to void warranties for reasons like this (ie, unless you actually caused the damage).
hero member
Activity: 658
Merit: 500
Do you have overheat protection feature in the firmware (shutting miner off automatically, or better yet auto adjust individual chips) ?
Yes, of course, we do DVFS on the ASICs. One of the DVFS parameter is the individual ASIC temperature.
(We have temperature sensors on the ASICs, AC2DC (PSU), DC2DC and three on the PCBs)
If you do, is it configurable ? ( adjusting temp threshold)
No, but we'll have all the source code opened. Changing the default firmware will void warranty of course.
We do allow different mining modes with the default firmware, which results in different power consumption.

Thank you i missed the DVFS part.

However, can you clarify a few things:

DVFS is meant to control the freq of the chips but can it shut off the chip completely to prevent overheating?

What happens when a chip is dead (for whatever reason)? is the FW still able to mine with the rest of the chips (and disabling the dead one) ?

Thanks.
donator
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
Do you have overheat protection feature in the firmware (shutting miner off automatically, or better yet auto adjust individual chips) ?
Yes, of course, we do DVFS on the ASICs. One of the DVFS parameter is the individual ASIC temperature.
(We have temperature sensors on the ASICs, AC2DC (PSU), DC2DC and three on the PCBs)
If you do, is it configurable ? ( adjusting temp threshold)
No, but we'll have all the source code opened. Changing the default firmware will void warranty of course.
We do allow different mining modes with the default firmware, which results in different power consumption.
hero member
Activity: 658
Merit: 500
Do the heat sinks mount directly to the chassis to provide more efficient active cooling in e.g. liquid cooled racks / fully contained cabinets?
Our systems are designed for air flow cooling. It doesn't require any exotic cooling infrastructure, liquid or others.
It can be used in a data center environment and in others environment.
Remote admin is possible for users interested in hosting the machine in a data center.
We'll provide our own competitive hosting plans soon.

Do you have overheat protection feature in the firmware (shutting miner off automatically, or better yet auto adjust individual chips) ? If you do, is it configurable ? ( adjusting temp threshold)

If you dont, can it be done? when will it be done?

I think this is absolutely necessary .

Thanks
donator
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
Do the heat sinks mount directly to the chassis to provide more efficient active cooling in e.g. liquid cooled racks / fully contained cabinets?
Our systems are designed for air flow cooling. It doesn't require any exotic cooling infrastructure, liquid or others.
It can be used in a data center environment and in others environment.
Remote admin is possible for users interested in hosting the machine in a data center.
We'll provide our own competitive hosting plans soon.
hero member
Activity: 686
Merit: 500
FUN > ROI
Since you're providing more technical information that just about any mining hardware company ever and only getting more and more questions in return for it... I guess I'll heap on Wink

Does having fans at the front help?
Additional fans are unnecessary. You won't gain anything from adding them.
Do the heat sinks mount directly to the chassis to provide more efficient active cooling in e.g. liquid cooled racks / fully contained cabinets?
sr. member
Activity: 336
Merit: 250
I just spent 20 minutes searching on that pdf a single 1200W reference. Nothing.
Again, the unit supports up to 1250W. This is after validating it with Emerson and thoroughly testing in our lab.
RoadStress, you're welcome to Israel. We'll do a comprehensive system DR if you're interested.
I just don't understand why you are claiming to be designed for 1200w. Why aren't you using the DS1200 model?
There is no need to use DS1200 which costs more.

Thank you for the invitation but i would like to spend my money on asics than on trips Smiley Why don't you send me one unit and i will pay for it after i receive it (i can do escrow too) and i will get all my answers then. (i know it won't happen so no need to reply about this).

All i have to do now is to wait for dogie's review. Hope to see a power measurement too.
oh you know he is gonna 'sniff' all over this!! Cheesy


ps. had they taken credit cards, I would have pulled the trigger already.....hmmm venture capital money, 8 months or more building/designing a SUPER miner, but couldn't figure out how to take credit cards?? Looking out for the customer means giving the customer everyone option, including protection on his purchase/investment with PAYPAL/C.C......wire transfer and btc payment could be like entering a black hole. just saying.
legendary
Activity: 1904
Merit: 1007
I just spent 20 minutes searching on that pdf a single 1200W reference. Nothing.
Again, the unit supports up to 1250W. This is after validating it with Emerson and thoroughly testing in our lab.
RoadStress, you're welcome to Israel. We'll do a comprehensive system DR if you're interested.
I just don't understand why you are claiming to be designed for 1200w. Why aren't you using the DS1200 model?
There is no need to use DS1200 which costs more.

Thank you for the invitation but i would like to spend my money on asics than on trips Smiley Why don't you send me one unit and i will pay for it after i receive it (i can do escrow too) and i will get all my answers then. (i know it won't happen so no need to reply about this).

All i have to do now is to wait for dogie's review. Hope to see a power measurement too.
donator
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
I just spent 20 minutes searching on that pdf a single 1200W reference. Nothing.
Again, the unit supports up to 1250W. This is after validating it with Emerson and thoroughly testing in our lab.
RoadStress, you're welcome to Israel. We'll do a comprehensive system DR if you're interested.
I just don't understand why you are claiming to be designed for 1200w. Why aren't you using the DS1200 model?
There is no need to use DS1200 which costs more.
donator
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
The temp is the chip temp?
Yes
Many of them is quite high compared to what i'm used to.... 113c?
The ASICs are designed to work up to 125C. The SW is limiting them to 119C
Whats your ambient temp?
I'm testing the unit at my room, which is quite hot now. I guess above 28C
Whats your recommendation for hosting environment?
It depends on whether the environment support free cooling. In Israel it doesn't, so the recommendation is 22C
Max ambient temperature is 35C
Does having fans at the front help?
Additional fans are unnecessary. You won't gain anything from adding them.
legendary
Activity: 1904
Merit: 1007
What is the temperature of SP10 chips and what temperatures do you estimate to have for SP30 chips?

In both cases the max Tj=125c. we design both system to allow the die working temperature to be up to 125c
To reduce leakage, we keep the temperature by sw control below 119

In the picture that you posted regarding the PSU there are no technical specs. There are only Environmental Specifications. Nothing about Ampers and Watts. You are the first company in the bitcoin mining field that draws more power than the PSU can handle. I just don't understand it and while on SP30 the overhead is low, in the SP10 the overhead is 30% which i find it pretty big.

The official DS1050 rated power is 1050W but the actual value the PSU was designed is 1200w.
After consulting with Emerson, to achieve max performance, we wallow to draw almost 1250w form the PSU.
We tested it and the systems in our lab runs under those conditions.
The PSU have built-in over current and overheat shutdown mechanism that protect itself and the system.

You can find the spec here: http://www.emersonnetworkpower.com/documents/en-us/products/embeddedpower/documents/documentationarchive/technicaldocumentation/01/trn_ac-dc_ds1050-3_release1_0_1312516765_techref.pdf

I just spent 20 minutes searching on that pdf a single 1200W reference. Nothing. I just don't understand why you are claiming to be designed for 1200w. Why aren't you using the DS1200 model?
hero member
Activity: 658
Merit: 500

This is the ASICs status at the moment I took the UI snapshot:

We're using DVFS (http://en.wikipedia.org/wiki/Dynamic_frequency_scaling) to squeeze the most of the ASICs and system.
We'll fine tune the algorithm to give better results in the near future.


The temp is the chip temp?

Many of them is quite high compared to what i'm used to.... 113c?

Whats your ambient temp? Whats your recommendation for hosting environment? Does having fans at the front help?
donator
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
When do you plan to tape-out the RockerBox?
End of April, beginning of May. Samples in beginning of July. Same team that did the hammer.

I see that you are avoiding the temperatures questions. Will try again. What is the temperature of the Hammer ASIC chip while running? What is your estimation on the temperature of the RockerBox?

Already replied, but I wanted to give you screenshots.
I'm testing Meni's unit before hand delivering it later tonight.

This is how the UI is looking (based on the amazing MinePeon work by Neil Fincham. Enjoy your soon to arrive unit Neil. Big thank you):

Again, we still have some bugs to root out, but the basics is working very good. Including OTA firmware upgrade developed by us.
(SP30 will use the same TI Processor, so all the SW except the ASICs specific code will remain the same. Big plus)

This is the ASICs status at the moment I took the UI snapshot:

We're using DVFS (http://en.wikipedia.org/wiki/Dynamic_frequency_scaling) to squeeze the most of the ASICs and system.
We'll fine tune the algorithm to give better results in the near future.
staff
Activity: 4172
Merit: 8419
I now have a tracking number for the review/test unit being sent to me and it looks like it's on its way.
hero member
Activity: 756
Merit: 502
Hi SyRenity,
are there any plans for accepting payment cards?

Unfortunately, not at the moment.
donator
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
What is the temperature of SP10 chips and what temperatures do you estimate to have for SP30 chips?

In both cases the max Tj=125c. we design both system to allow the die working temperature to be up to 125c
To reduce leakage, we keep the temperature by sw control below 119

In the picture that you posted regarding the PSU there are no technical specs. There are only Environmental Specifications. Nothing about Ampers and Watts. You are the first company in the bitcoin mining field that draws more power than the PSU can handle. I just don't understand it and while on SP30 the overhead is low, in the SP10 the overhead is 30% which i find it pretty big.

The official DS1050 rated power is 1050W but the actual value the PSU was designed is 1200w.
After consulting with Emerson, to achieve max performance, we wallow to draw almost 1250w form the PSU.
We tested it and the systems in our lab runs under those conditions.
The PSU have built-in over current and overheat shutdown mechanism that protect itself and the system.

You can find the spec here: http://www.emersonnetworkpower.com/documents/en-us/products/embeddedpower/documents/documentationarchive/technicaldocumentation/01/trn_ac-dc_ds1050-3_release1_0_1312516765_techref.pdf
legendary
Activity: 1904
Merit: 1007
Sounds like ...NO running engineering sample for SP30...right...?
Correct.
We're deep into the design cycle of RockerBox, the ASIC inside SP30 Yukon. By running simulations, we know it's parameters and we're able to design and simulate the miner.
In SP10, we've aimed for 1.3 TH/s with margins and we got 1.4 - 1.6 TH/s. SP30 parameters will be according to the spec we've published.

When do you plan to tape-out the RockerBox?

I see that you are avoiding the temperatures questions. Will try again. What is the temperature of the Hammer ASIC chip while running? What is your estimation on the temperature of the RockerBox?
ZiG
sr. member
Activity: 406
Merit: 250
Sounds like ...NO running engineering sample for SP30...right...?
Correct.
We're deep into the design cycle of RockerBox, the ASIC inside SP30 Yukon. By running simulations, we know it's parameters and we're able to design and simulate the miner.
In SP10, we've aimed for 1.3 TH/s with margins and we got 1.4 - 1.6 TH/s. SP30 parameters will be according to the spec we've published.

Thank you for this clarification...

On the other hand you are taking pre-orders for that "deep design cycle"...based on simulations...Just trying to get the "whole picture"... Wink
donator
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
Clearly Meni has never seen the hardware. Where is Adam's review of the hardware?
My point still stands. Why were you so desperate to open orders before delivering any actual sample hardware?
Meni will get his evaluation unit tonight. Adam will get his evaluation unit next week.
Desperate ? Definitely not. We are proud of our product. We opened orders the moment the web site was ready.
We didn't ship until yesterday because we preferred to do the necessary burn-in testings. In addition the UI wasn't ready.
Jump to: