hi guys,
we have contacted this guys from linkedin :
http://www.linkedin.com/profile/view?id=13644568&locale=en_US&trk=tyah2asked him he would be able to built server having the same hashing power than the big one of butterfly:
this is his reply:
Our fastest server has an average rating of 1775 GH/second by our most recent tests.
Liquid Nitrogen Overclocking
http://www.liquidnitrogenoverclocking.com610.818.5063
so i va asked him to send me a screencopy an a video of the hashing power of his server,this is his reply:
Hello again,
That was the average value after only 6 hours of testing. Honestly, I do not know anything about BitCoin, one of my engineers set up the test and ran it for a little while over the weekend.
The power usage is 985 Watts/system. Since we did not design this specifically for BitCoin mining, I have no idea if this is good or bad.
This is the system we used:
http://www.liquidnitrogenoverclocking.com/HSSM_3960X.shtmlThey are a little pricey at $25,000 each, but it comes with 18 processors x 4.8 GHz each = 86.4 GHz worth of computing per system. The cost "per GHz" is low, $289 per GHz, same as a laptop basically.
Liquid Nitrogen Overclocking
http://www.liquidnitrogenoverclocking.com610.818.5063
i ve called him and asked again by email about screenshot and video,his reply:
We are not going to do all of this for you. If you want the hardware, purchase it.
I talked to my engineer again, he said he was getting 1775 GH/second, as I mentioned. But, the power usage was 1.1 Kilowatts, not 975 Watts.
Liquid Nitrogen Overclocking
http://www.liquidnitrogenoverclocking.com610.818.5063
i replied him telling him that we ll buy it if it achieves what you claims otherwise,we stop here.his reply:
The answer is no. We're too busy to do special projects. We already did what you asked. The results are certified. We do about $1,200,000/day in sales for clients who don't ask us to do test after test after test. We simply don't have the time to take on every special request that is asked of us.
Liquid Nitrogen Overclocking
http://www.liquidnitrogenoverclocking.com610.818.5063
i ve called him again and told him that no purchase will be done if there is no proof.after that,he seems to have setup another character in his scenario ,please read below:
Matthew,
Please put your signature on the bottom of your emails. For some reason, "...." thought he was chatting with me regarding the BitCoin test.
I can explain why the results were so high. The BitCoin application software was installed on our Prototype 5.5 GHz Ivy Bridge i7-3770K chip which is about as fast as would be a 6.4 GHz top-of-the-line Sandy Bridge-E processor, and about as fast as a 7.2 GHz Westmere class server.
This Prototype system uses a new Microchannel Heat Pipe implementation which can remove heat far better than even a Single Stage Vapor Phase Change cooler capable of maintaining -40 Celsius in a vacuum chamber across a CPU block.
In other words: The very efficient, and very hot, Ivy Bridge chip can outperform anything under the sun. Also, this was a 1U server x 4 CPU cores/server x 9U = 36 cores = 72 hyperthreads running at (basically) the speed of a 7.2 GHz Westmere server. You had over 500 GHz worth of processing power attacking the BitCoin mining operation.
Why our Chief Techical Officer did not disclose how many cores he was running is unknown to me. Furthermore, the prototype system was to remain offline until our announcement on Monday. I guess he was too eager to see how it could behave in the real world.
The good news is: the average power requirements were even lower than previously stated. The MHP cooling technology uses an inert gas cycling in a closed loop that hits an endothermic phase change as it makes contact with a Copper CPU block cut out as a 6th-order Menger Sponge. The more surface area that this gas is in contact with, the cooler the surface becomes, and a Menger Sponge has nearly infinite surface area while enclosing zero-volume. Google for "Menger Sponge" for more info about this unique geometric construct.
The BitCoin equivalent calculations were at 1768 GH/second based on my calculations. Not all cores were started at the same time, and the log files suggests that Nathan made a rounding error. The power used was 1629 Watts as measured from the wall to the entire 9U rack. The meters are accurate +/- 1.2% for AC power measurements.
With my best regards,
Ed Trice
Liquid Nitrogen Overclocking
http://www.worldsfastestcomputers.comhttp://www.liquidnitrogenoverclocking.com610.818.5063