Topic: H/w Hosting Directory & Reputation - page 44. (Read 113313 times)

Power costs in Seattle, where DF is based, are around $0.068/kWh. The sub-$0.04 stuff is all in central WA, where Titan (Spondoolies's contracted datacenter) and I are. My rates in Grant should be about $0.023/kWh once I get above 200 kW load. Titan's are probably similar. That's probably much of why DF is pricier than Titan.

@All DCs: You need to realize that bitcoin mining doesn't need all the fancy and expensive backup stuff. The diesel generator isn't needed if it brings costs down. You need to see how much will the power be down in 1 year. We miners are ok with 1-2-3% of having the power down. I am sure that all electric problem can be fixed in a matter of hours/days and they don't occur frequently.

Bitcoin mining needs DCs that are like a chicken farm. Minimum investment that provides power and cold air. That's it!

@Digital Fortress

What is the guaranteed / typical cold aisles temperature at digital fortress?

You are also more on the pricy side (charging at least 20.8 cents / kWh), did you consider building out mining capacity without some of the costly redundancies?

I am sure there are some more ways to optimise your pricing, as DCs in WA get 3-4ct/kWh electricity, some of these cost savings can surely be passed on to the customers.

Spondoolies is able to quote 16cts /kWh (115$/kW) in a WA facility (starting from a 2.7kW contract), I am wondering why you are quoting 20% more on a 10kW contract.

Hi guys, quick check-in. We're still building the datacenter in central WA for mining. However, we are no longer working with the ASICSPACE crew, so we'll be using a different name. Policies and prices will remain largely the same, with rates around $70 to $90/kW/month.

Current progress: We our first warehouse leased in Grant County with about 1 MW of available power at the service entrance. Electrical work has begun. At this moment, there are 3 guys from one of the local electrical contracting companies putting in conduit for one of our 500 KVA feeders. (I'll probably release the name of the contracting company later so you guys can call and verify our existence and capacity, but I first need to brief them on what information they can and can't divulge.)

We've had a few delays, so we're not nearly as far along as I wanted to be at this point, but I left enough buffer time in the schedule that I still think we'll be ready to accept customers in September, and possibly even a few in August. It will be tight, though.

More information will follow as we get closer to coming online.

Howdy Forum,
I thought I would introduce ourselves as Seattle Data Center owners. We currently have over 2MW of miners in place and have been deploying a lot of individual machines recently. That said, we have learned a lot and we're happy to share our knowledge from a data center facility standpoint.

•   Market Knowledge – http://dfcolo.com/bitcoin_terraminer_hosting/
o   2 MW of Bitcoin Mining operations currently under management in our datacenters.
o   With significant time and resources  spent learning Bitcoin mining operations  we have become well versed in this segment of our industry. We know how to anticipate and respond to your requirements based on extensive hands on experience in this industry.
o   Currently have hundreds of Spnondoolie 10’s plus many Antminers, Terraminers, KnC boxes and custom deployments in service.
o   Can, manage, deploy setup machines plus configure private VPNs, firewalls and DDoS protection as needed.
•   Infrastructure Seattle–
o   2N 1.5MW Caterpillar diesel generators with 3 days of fuel on site
o   N+1 Cooling on pressurized 24” raised floor
o   VESDA smoke detection and ECARO fire suppression
o   24x7 onsite staff, support and remote hands
o   Low latency redundant fiber backbone
o   Custom realtime RFID temperature and humidity monitoring  
o   True Green State, with Green Hydro Electric power

•   Culture –
o   Profitable and well-funded managing over 70,000 sq. ft.
o   Long history in this community and are building a company positioned for ongoing growth and longevity.
Video Tour http://dfcolo.com/seattle-data-center-video/
Call Kevin 877-720-0483 - 206-681-9732 [email protected]

Even better at night 

I concur:

Temp Front / Back T,B
    32 °C / 74,76 °C

Cheers,
QG

Temps are back to 35 C^o! I can see, that this is being taken care of. Hopefully will have lower figures, after you will be done with ACs.

Thanks MichColo.

Thanks Wish I could do 1.5 TH/s too. Unfortunately, I am not able to achieve that for obvious reasons.


Yes. There are 192 Hammer ASIC chips in one SP10. They are marked from 0-191.

To be honest, most servers and some miners (probably including the sp10) can handle high temperatures, it is variance in temperatures that puts much more stress on a unit.

You will typically see failure of chips / units right after bigger temperature spikes (or when they come down again), but you should be fine afterwards.

Don´t worry too much about 1 broken asic on a 192 chip unit...
I own one sp10 where 9 chips were disconnected from the very first day, and I am still getting my 1.5TH on it.

Interesting. I presume the line with X's is the chip?

Hope the ASICs are used to Middle East high temps, and will survive until we got the temps fixed They are tough Israeli ASICs, I believe they are capable to resist even more extreme

Regarding costs, this temps already cost me. My units are mining ~100 GH/s below my expectations (worse scenario). After fixing the temps, I will be able to figure out, how much hash power I have lost, due to several days of extreme heat.

First off, I want to be clear that my responses here are not intended as combative, i'm just addressing the comments made. I don't want you, or anyone else to think im "attacking" or trying to deviate from the situation.


Right, that's exactly my point. For you to call us unprofessional, or judge our "backup plan" without even knowing what our "backup plan" is, is forming an opinion regarding a situation that you don't have a fair, full group of facts to form said opinion.


Technically, that's exactly what you're doing. It's quite easy to play the role of a keyboard jockey and judge something that you don't know about. (I say "you don't know about" based off the fact that you don't know our cooling, our redundancy situation, or our "backup plans")...


Again. Based off the fact that you don't know what we have in the data center for cooling, or how it's set up, or what our redundancy is, you can't judge what the "perfect storm" is, or how you would plan. There's only so far your department of redundancy department can go. We unfortunately found out.

>> Out of curiosity, have you ever been in a data center?
>> Have you ever been in a data center that has 100 or more mining machines running at the same time?


Would it make the reply different? Possibly. I don't know how you plan to plan, or what your knowledge of data center inner workings are.
I can only tell you this.. Mining machines break out unGodly heat. There's a reason all data centers aren't taking machines in.
IMO (and the reason I asked) if you have been down an aisle with multiple mining machines, you would know how fast the temperatures rise and once they have risen, that it isn't as easy as "flipping a switch" to bring them back to normal.
My reason for asking was, if you had been in a data center, versus a data center with a mining machine section, you get a better grasp on what the area is like.

Our stats from the past have been posted. We run a very cool, climate controlled data center, and as mentioned, that was one of the reasons people have gone with us.
Did we plan on losing AC units? Of course not.
Did we try to combat the rising temps when we had failures? Of course.
Could we plan on losing multiple units all at once? Yes, and we did. It's something that happens on an extremely rare basis, something that is considered the worst case scenario, and we executed the plan based off the threat of losing multiple units.
Did we bring in an emergency air unit to execute an emergency plan? Yes.

You stated that we didn't have a plan in place, when the fact is, we did, we executed, we are living in it right now.

Partial? I can only go by what you have written on this forum.

"It's quite easy to sit back and play Armchair Data Center Operator" ... now that I don't do ... I only comment on things people experience & what I have experienced myself.

"perfect storm" ... planning sorts that one.

>> Out of curiosity, have you ever been in a data center?
>> Have you ever been in a data center that has 100 or more mining machines running at the same time?

No .. but should I? How would that help? If I said yes ... would that make my any reply diffrent? No ... When I choose a hosting solution it is based on cost/reliability/feedback/TEMPS/SLA and an averaged view from others in the forum community who have used them.

If possible can you post a copy of your SLA please?

I don't think you're understanding the current situation. You're basing your opinion on partial information you're getting through an internet forum.

Unit failed.
Redundant unit kicked in to take up some of the load.
Plan for worst case scenario: Bring in emergency air unit. Bring in emergency contractors to repair failed units.
Still have other units working to keep temps from hitting 150 degrees.

We've got multiple overhead units in a 25k+ sq foot data center. We run additional units to ease the workload of each individual unit. When one fails, they all work harder. When 2 fail, they work VERY hard and sometimes, depending on the heat you're experiencing, it's extremely hard to overcome those losses.

It's quite easy to sit back and play Armchair Data Center Operator. You can conjure up whatever you would like to, the fact is, as I mentioned, this was all worst case scenario or better put, a "perfect storm".
We put emergency AC units into play right away, called in emergency mechanics to fix damaged units and moved up project dates on additional unit installs.

Serious question, not a troll...
Out of curiosity, have you ever been in a data center?
Have you ever been in a data center that has 100 or more mining machines running at the same time?

You may have been in the business for years ... which actually makes the current situation worse with your reply.

>> Unfortunately, we had some units fail and in a worst case scenario, our redundant unit also failed.

So, you didn't have proper pre-planning in place for worst case scenario then. Not meaning to shoot you, but if some units fail  ... and at the same time the redundant unit 'also' fails ... there is no robust plan in place.

While I appreciate your opinion, you're well off base.
We've been in the data center business for years. We're not a "fly-by-night mining center".

Unfortunately, we had some units fail and in a worst case scenario, our redundant unit also failed.
As mentioned, we've been in the process of bringing a new unit online, which was a planned upgrade. These unit problems escalated the install of one of two, and had us bring in an emergency AC rental to help overall.

Doesn't seem too professional to me, MichColo should have had redundancy and back-up plans.

If your dead ASICS don't come back to life after a cool down in temps ... they cost you.

Nice, thank you!