Topic: HashFast announces specs for new ASIC: 400GH/s - page 445. (Read 880485 times)

@HashFast

Can we have a production update? It's been a week since the last one, maybe you can make Thursday night production update night?

   

Only people that really can't read can seriously think that those are the objection that where risen in this thread.
Maybe you where trying to catch them? Well, fail.

IKR?  Those digs about hands-free devices, "American" smiles, and suits were completely uncalled for.

God forbid the CEO of a tech start-up wear a suit, use a headset, and smile when seeing the fruits of his labor finally delivered.   

Personal attacks = ages olds technique. Doesn't really even fit in.

There.  Fixed.

I Agree.

That (non-American) man is smiling because he is holding the result of thousands of hours and bitcoins worth of effort, in the form of the world's first true 3rd gen ASICs.

His pride and joy is completely deserved.

I'm glad he's wearing a hands-free device because he's obviously extremely busy coordinating all the aspects of building our mining hardware ASAP.

We don't want him driving around holding a phone to his ear, or missing important calls while fumbling for his cell.

Haven't you ever been proud of an outstanding accomplishment?

Your hatred and contempt for the rational exuberance associated with demonstrated productive capacity makes me think you are perhaps what Ayn Rand called moochers, looters, and second-handers.

Chip efficiency is often not the whole story.  You are looking at 10% inefficiency (12%-13% probably more realistic under real world conditions) in DC regulator and another 10% for the ATX PSU.

So if chip is 0.65 J/GH @ 1VDC then the entire board is ~0.74 J/GH @ 12VDC.   @ 400 GH/s nominal that is 295W.  Now add on two probably 6W fans per radiator and you are looking at 307W per board.   On sierra we know there are two exhaust fans.   So 307W*3 + 12W = 933W @ 12VDC.  Throw in 90% efficient PSU and that is 933W/0.9 = 1036W at the wall.   1013/1200 = 0.87 J/GH.  

HashFast has stated they believe the boards can be faster possibly 500 GH/s.   To push the clock that high may require a higher nominal voltage.   Power usage for silicon scales linear with the clock (no change in J/GH) but scales by the square of the voltage.  So say it requires 10% higher voltage that would be 21% higher power consumption.  0.87 * 1.21 ~ 1.1 J/GH.

Since HashFast doesn't yet know what the final clock and voltage will be the 1 W per GH/s makes sense.  My total guess is that if the chips end up closer to 400 GH/s without much headroom they will run it at a lower nominal voltage and higher efficiency.  If the chips show more potential performance they will run them at a higher nominal voltage and lower efficiency. 

Just my opinion got no inside info or anything like that.  

1W/GH is the number that John throw at me long time ago. 0.65*1.1*1.1 = 0.8, round that to the excess and you are almost there.

Just saw this in the product spec of the sierra and BJ:

At the wall, this unit consumes 1 watt per gigahash +/- 20%

Is that new? The chips where supposed to achieve 0.65GH/W. From there to 1GH/W at the wall seems a big difference to account for inefficiencies and cooling..

Yeah its a really bad way to do business, but its like that 

By the way, they will tell you that everything is under schedule and as planned as they did to us back at October when they did not even expect the wafers to be shipped until a week ago.
Do not trust on HF words. They will tell you all you wanted to listen until you are in. Once you are in, say bye bye to any kind of contact or suggestion. These days they did not even answer to the emails and the same for Josh at skype.

Run away now that you can !

If you wanted a real estimated shipping date look at their terms of sale. Until that limit date they will take your money and do whatever they want. We will tell you what happens when that date comes close what is their next brilliant movement

Well they promised us October 20th-30th deliveries..
said that that everything was on track...
Waited until Ocotober 23rd to announce delays..
they blamed it on the substrate and said everything else was on track which turns out is a lie since they are only now getting the wafers..
said the new estimated ship date was mid-november..
and then just recently announced that it was more like mid-December.

They have lied to us and I see no reason why they wouldn't lie to you.
Do not pay for your order and run away now while you still can.

It seems that ASIC manufacturers make much more money that the miners!!
Why should they ALL ask for our money in front?! To do business with our money? That's the worst way to develop a business!
I am waiting for their telephone to confirm the delivery, otherwise will not pay for the order placed.
Let see what happens...

Sorry but the first batch is expected to be shipping in mid December IF everything goes well. They have not said a word of the next batches , but I personally think that we could consider ourselfs "lucky" if we got our first rigs in late December

Is there any topic about hashfast 1200GH/s ?
I have ordered one today and would like to find out if there are others who did so. January delivery; hope they will be able to keep their word.

KnC would have had their substrates ready to go. Also, they had way more time between when they started taking orders and their deadline, which they missed by deadline by 4-15 days (which caused me to lose a bitbet, although my Saturn has covered that so far - a Saturn that was delivered about 15 days earlier then they'd promised I'd get it  

Oh, and they weren't scumbag scammers. That helped.

ohh god, this guy is obviously so fucking important. american smile, non-stop wearing of his mighty blue-tooth headset, suit and in hand some strange package in grey plastic blister.

love this kind of updates.)

Ok , lets forget the ponies picture , which was kind of funny..
Can somebody explain why that tridimensional package looks like an elephant stomped on those ?

Bitcoin doesn't change human interaction.  There will be favorites and they will be supported without any logical reasons for it.  For some companies every little thing will be inferred to have sinister intent and for others every mistake will be an act of God unpredictable by all.