► ► ►HashFast Endorsement

[DyslexicZombei]

Hehehe, we're talking marketing speak here AussieHash. It's like a whole new lingo where you make up your own meanings and take small (or giant) liberties:

"New & Improved!"
"4 out of 5 dentists prefer our gum!"
"More clowns prefer our Big red noses over our competitors lousy red noses!"

The whole thing reeks of PR fluff instead of an article.

[itod]

The whole thing reeks of PR fluff instead of an article.

It's true that this is a PR text, but there is an important fact in it:

The design simulation met or exceeded the design targets, performing more than 400 billion double-SHA256 hashing operations per second (GHash/s), while consuming less than 0.6 W/GHash/s at nominal clock rate, setting a new industry benchmark.

God to know this step turned out better than expected.

[creativex]

It's just simulated mumbo jumbo. Pre-order vendors have been issuing overly optimistic simulation results since forever. Until they have test chips in hand and some solid measured results everything will remain up in the air.

[iCEBREAKER]

It's true that this is a PR text, but there is an important fact in it:

The design simulation met or exceeded the design targets, performing more than 400 billion double-SHA256 hashing operations per second (GHash/s), while consuming less than 0.6 W/GHash/s at nominal clock rate, setting a new industry benchmark.

God to know this step turned out better than expected.

One more important new fact: TSMC.

Established in 1987, TSMC is the world's largest dedicated semiconductor foundry. As the founder and leader of the dedicated foundry business, TSMC has built its reputation on offering advanced wafer production processes and unparalleled manufacturing efficiency.

Nobody has a better chance of meeting or exceeding the "simulated mumbo-jumbo" than TSMC.

HashFast just dropped the hammer on their competition.  It doesn't get any better than TSMC, the 500lb gorilla of the chip biz!   

[regular]

It's true that this is a PR text, but there is an important fact in it:

The design simulation met or exceeded the design targets, performing more than 400 billion double-SHA256 hashing operations per second (GHash/s), while consuming less than 0.6 W/GHash/s at nominal clock rate, setting a new industry benchmark.

God to know this step turned out better than expected.

One more important new fact: TSMC.

Established in 1987, TSMC is the world's largest dedicated semiconductor foundry. As the founder and leader of the dedicated foundry business, TSMC has built its reputation on offering advanced wafer production processes and unparalleled manufacturing efficiency.

Nobody has a better chance of meeting or exceeding the "simulated mumbo-jumbo" than TSMC.

HashFast just dropped the hammer on their competition.  It doesn't get any better than TSMC, the 500lb gorilla of the chip biz!   

Agreed.  TSMC was responsible for the first ever shipped ASIC as well (Avalon) so they have a good track record with putting out working chips.

[Melbustus]

So, I think where I'm looking for more clarity is:


1) Cost of the board/rig required to run additional chips if the MPP gets executed. Do we have a parts manifest for this yet, or a price to outright buy a pre-assembled rig from Hashfast?

2) If anyone from Batch-1 gets additional chips from the MPP, everyone should. Otherwise, there can be situations in which people who are ahead in the shipment queue actually do worse. For example, let's say that the 1st BabyJet from batch-1 goes out on October 25th and has made a slight profit by Jan 23rd; under the current MPP verbiage as I read it, that means no new chips for Owner #1. Let's further say that the 500th BabyJet from batch-1 goes out November 25th, but has just barely not covered costs by Feb 22nd. In this case, the owner of the 500th gets new chips for free, but the owner of the 1st still doesn't. Given the low power consumption of BabyJets, they'll likely be worth running for quite some time, so over that lifetime, the owner of the 500th probably does better. It depends on the exact nature of how the overall network hash-rate curve develops, of course, but it seems to me that the fairest thing to do is to execute the MPP across the entire Batch; ie, if anyone from the batch gets more chips, everyone should.


Can we get some clarity/comments on the above?


In any event, HF seems to be doing great work. Tape-out ~10 days ago, plus the TSMC announcement...good stuff.

[jjiimm_64]

It is going to be a horserace.....  I hope I picked the right horse!

[HashFast]

HashFast has launched sales of it's second production batch. We are now selling our commercial grade unit, the Sierra, which contains 3 Golden Nonce (GN) ASICs, and all Baby Jet and Sierra machines can now be purchased with or without the Miner Protection Program™.

John
HashFast.com

[cypherdoc]

everyone, please pay attention:

"The Baby Jet motherboard has room for an additional GN chip.**"

this applies to the Batch 1 units as well.

[Visesrion]

everyone, please pay attention:

"The Baby Jet motherboard has room for an additional GN chip.**"

this applies to the Batch 1 units as well.

So this means that with the MPP if we receive 3 bare chips we could install one by improving the power supply which is ok. But price for aditional motherboard ?

Any plans of hosting ?

I also think that executing the MPP will be really dificult depending on international shipping delays as in my case and depending on when the diferent units in the batch gets shipped, so you should find a solution to that.

[bobsag3]

everyone, please pay attention:

"The Baby Jet motherboard has room for an additional GN chip.**"

this applies to the Batch 1 units as well.

So this means that with the MPP if we receive 3 bare chips we could install one by improving the power supply which is ok. But price for aditional motherboard ?

Any plans of hosting ?

I also think that executing the MPP will be really dificult depending on international shipping delays as in my case and depending on when the diferent units in the batch gets shipped, so you should find a solution to that.

A bunch of us are already offering hosting, and working on a project to develop DIY boards. I think the thread is in this section or in GB's.

[cypherdoc]

FYI.

I am no longer being paid by Hashfast. That concluded with batch one.