HashFast announces specs for new ASIC: 400GH/s

[ninjarobot]

According to MPP they will give you up to 400% more hashing capacity than the Baby Jet you purchased. They calculate the ROI based on network difficulty adjustments, the estimated hashrate, and what hashrate the average Baby Jet achieves.

You purchased a BJ, not 400GH/s. If it turns out the nominal speed of a BJ is 700GH/s then that will be the baseline for the MPP calculations. Personally I think whatever the BJ runs at at 350W at the wall should be the nominal speed.

[fubly]

[DeathAndTaxes]

My updated model based on the next 48 Difficulty changes of 2016 blocks. Now shows percentage of change between diffs as well as the time it should take from one diff change to the next, and the total number of days the model is predictive of based on the shortened diff adjustment periods.

https://docs.google.com/spreadsheet/ccc?key=0AkOUeP-343FQdDBDbjl3Ymw5SGFNbDBxbnFneWJ0TUE

There are 3 sheets showing 500Gh/s BJ + 4 Board MPP, 700Gh/s BJ + 4 Board MPP, and 800Gh/s BJ + 4 board MPP.

You probably want to start you spreadsheet at 1.46 billion difficulty (10.5 PH/s) as we are 4 days from difficulty adjustment so HF batch 1 customers are not going to materially participate in that difficulty period.

It looks like it is starting @ 11 PH/s.  Either I need more coffee or I am losing my mind but I could have swore it showed 9 PH/s earlier.

[bromide]

What's the upshot wth this?  Sorry if it was answered & I missed it.

http://s2.postimg.org/a7nbok8nt/Capture.jpg

*V Thanks, senseless.

[DeathAndTaxes]

One sad fact is that 400 GH/s on Oct 20th is equal to 1,563 GH/s today.  So if HashFast was running 700 GH/s+ per module AND they decided to upgrade all Batch 1 customers to two modules, as amazing (and utterly unlikely) as that would be it would only be roughly as good as 400 GH/s back in Oct.  Kinda shows how deep the hole is now.

PS: If you have nothing else to do, rename the tabs from Sheet1,2,3 to something descriptive.   Nice spreadsheet BTW, I haven't verified the math or anything yet, but I always like using spreadsheets over blackbox website models.

[DeathAndTaxes]

What's the upshot wth this?  Sorry if it was answered & I missed it.

There was no answer to it from them. But I went through and looked at the stats. The problem with the stats wasn't over reporting, or under reporting, but no reporting. There is a gap on every miner at eligius for that period where no stats were recorded. All of the other stats before and after that period are accurate as confirmed by a few eligius miners. It's pretty clear that the BJ was doing 700-800Gh/s. I'm not entirely sure if this 1200Gh/s that is now being mined on that address is a BJ or a sierra. They said they were going to setup a sierra, but did not say if it was going to be on the same address, if they were taking down the BJ, or what.

I wonder if the 1280 GH/s is a 2 module BJ (i.e. testing if power supply and cooling can handle the upgrade module).  That would put the hashrate of a module closer to 600 GH/s?  Meh.  I am going to log off.  Got to much work to do and the lack of info plus stats (of unknown origin) to watch will just burn up the whole day.

[gjpminingco]

hey all I have created a Spreadsheet that should be able to help calculate if ROI is possible

I use it for Calculating Profit on CEX.io but I should be able to mod it to work for you all

[jjiimm_64]

Peaking 1.6Th/s now, has to be either a 2 module bj or sierra. No way is a single chip doing 1.6Th/s.

why cant they just tweet what the frak they are hashing with

[JWU42]

I am thinking they are testing a Sierra now or have added a 2nd Baby Jet.  No chance they are hitting > 1 Th/s with a Baby Jet.

If they were using an appendage to the mining address we could see individual workers (e.g., 1KSZMq1L9ZWEBKEHuW7ACXWZJYdxiTdaVt_BJ and 1KSZMq1L9ZWEBKEHuW7ACXWZJYdxiTdaVt_Sierra)

[Kouye]

why cant they just tweet what the frak they are hashing with

Because they don't wanna admit they finally received their minirigs from BFL.

[ImI]

Peaking 1.6Th/s now, has to be either a 2 module bj or sierra. No way is a single chip doing 1.6Th/s.

why cant they just tweet what the frak they are hashing with

they did

HashFast ‏@HashFast 8h

We so need to get a test Sierra into a mining pool too and see how it does as well.

[jelin1984]

What happen
With terra hash

[cedivad]

I have to admit that 800GH/chip, if stable, would be impressive to say at least. I wonder how the sierra is supposed to dissipiate 3kW of heat in its 4U. Not even recent and ultra high density datacenters can handle 30kW/rack.

The Sierra is sold as 1.2 TH/s unit not a 3 chip unit.   IF (and I haven't seen any convincing evidence) they are doing 600 GH/s+ they could drop the Sierra to two modules and just keep the extra hashing power for themselves.  Does anyone think that is unlikely given all the other ways they have screwed over customers so far?

From my proforma invoice:

"Numer of Sierras"
"Sierra contains 3 HashFast Golden Nonce (GN) ASICs, Performance: 1,200 Ghash/s at nominal clock speed"

[jspielberg]

Have we gotten an estimate of the power yet?

The eligius reporting 1200GH/s is freaking me out a little ...  
I may have to backtrack on my negativism with HF over the last couple of days.

I will update my spreadsheet to include the suggestions of moving from daily... to per retarget.

[ImI]

Have we gotten an estimate of the power yet.

The eligius reporting 1200GH/s is freaking me out a little ... 
I may have to backtrack on my negativism with HF over the last couple of days.

that will DEFINITELY wont be one BJ

maybe two hashing along with 600 or one sierra with 1200

i wouldnt hang my expectations to high at the moment until they officially declare whats going on

[cedivad]

Right. After all they said that it's only one BJ, so it can't be a sierra, right?

What if they ship those amazing 1PH/s BJ after the 31 of Dec and you accept the delivery anyway, refusing the refund, and you discover that they only go at 600GH top?
I think i could should expect such a move from them.

[jspielberg]

Right. After all they said that it's only one BJ, so it can't be a sierra, right?

Its possible they underestimated the performance of the GN1.  A single GN pulling 1.2TH/s seems just too good to be true...

It would be great to get some clarity on the exact hardware and power is being used for http://eligius.st/~wizkid057/newstats/userstats.php/1KSZMq1L9ZWEBKEHuW7ACXWZJYdxiTdaVt
And if possible some screen shots of the same hardware against btcguild or some other pool.

[perezoso]

I have to admit that 800GH/chip, if stable, would be impressive to say at least. I wonder how the sierra is supposed to dissipiate 3kW of heat in its 4U. Not even recent and ultra high density datacenters can handle 30kW/rack.

The Sierra is sold as 1.2 TH/s unit not a 3 chip unit.   IF (and I haven't seen any convincing evidence) they are doing 600 GH/s+ they could drop the Sierra to two modules and just keep the extra hashing power for themselves.  Does anyone think that is unlikely given all the other ways they have screwed over customers so far?

It's this kind of question that causes me to despair at the situation.  Hashfast's customers, having been offered the opportunity to work together on issues like this (since we now know who all the B1 customers are), instead seem 300% determined to behave like a herd of cats, each off doing their own crazy thing.

A more united front would probably force them to include the full module set, but crazy wingnuts that we are, off living our own fantasies, it leaves the door open for them to continue abuses.

[perezoso]

Have we gotten an estimate of the power yet.

The eligius reporting 1200GH/s is freaking me out a little ...  
I may have to backtrack on my negativism with HF over the last couple of days.

that will DEFINITELY wont be one BJ

Outside chance it's a BJ with two modules.

In fact, thinking, that makes sense to me.  They've gotta be sure the upgrade kit will work before shipping the BJ box. So test the BJ with two boards and the shipping power supplies, etc.

[cedivad]

Spoiler Alert: this picture has nothing to do with this hashfast.