1GH/s, 20w, $700 (was $500) — Butterflylabs, is it for real? (Part 2)

[CubedRoot]

What i don't understand is how they came up with those figures of 20W and ~1GH/s...You have to test to find these figures...

they're basically using all this forum attention to sell their mediocre FPGAs, which by the time they ship it (if ever), would probably have the same performance as existing ones. even if everyone canceled their preorders because the specs don't match, they essentially got 2+ months of interest free loan.

I don't understand all the bashing on BFL.  Sure they missed the mark on their initial numbers. Its a first design. Even at ~850 MH/sec and $599 with the current specs you still are getting one of the best MH/$ thats currently availble at nearly 1.42  MH/$.  Thats substantially more than any other FPGA out there.  Not to mention, this will be a fully finished piece with a stackable enclosure, fan and power supply.

BFL = 1.42 MH/$  --- Ready to go, with enclosure, fan, and power supply.

Icarus = 0.6 MH/$ --- Ships from China, does not include enclosure. Does ship with power supply and fans (although some say fans are loud)

ZTEX = 0.47 MH/$ --- No enclosure, No heatsink included (does need one), no power supply included.

X6500 = 0.38 MH/$ --- No enclosure,  includes heatsink, but developer said it *probably* will need airflow, no power supply included

So, lets get real here, the BFL unit nearly triples the MH/$ returns even on the diminished claims of 850 MH/s.

And if you want to discuss the larger power draw, the BFL unit is currently doing about 80 watts, but BFL claims that could be do to runaway, and may run lower wattages once released.  But still, we know it runs 80 watts.  The nearest competitor, Icarus, is running aboutn 20 - 23 watts real world.  IT would take about 3 of these units to equal a BFL.  When you consider the efficiency of the wall warts in the additional Icarus units needed to make up the performance difference, the power differential is pretty much null.  Or so small, that it would take YEARS of mining to make up the cost differential between the BFL and Icarus.

So, I dont want to take anything away from the FPGA home brew guys, but BFL's product is substantially better in my opinion. Mind you, this is their FIRST iteration of the design.  If they do indeed deliver their product to market, they will be the ones to beat, and that will be a hard battle to fight for the home brew guys.

[1714972342]

1714972342

[1714972342]

1714972342

[1714972342]

1714972342

[1714972342]

1714972342

[worldinacoin]

What i don't understand is how they came up with those figures of 20W and ~1GH/s...You have to test to find these figures...

they're basically using all this forum attention to sell their mediocre FPGAs, which by the time they ship it (if ever), would probably have the same performance as existing ones. even if everyone canceled their preorders because the specs don't match, they essentially got 2+ months of interest free loan.

I don't understand all the bashing on BFL.  Sure they missed the mark on their initial numbers. Its a first design. Even at ~850 MH/sec and $599 with the current specs you still are getting one of the best MH/$ thats currently availble at nearly 1.42  MH/$.  Thats substantially more than any other FPGA out there.  Not to mention, this will be a fully finished piece with a stackable enclosure, fan and power supply.

BFL = 1.42 MH/$  --- Ready to go, with enclosure, fan, and power supply.

Icarus = 0.6 MH/$ --- Ships from China, does not include enclosure. Does ship with power supply and fans (although some say fans are loud)

ZTEX = 0.47 MH/$ --- No enclosure, No heatsink included (does need one), no power supply included.

X6500 = 0.38 MH/$ --- No enclosure,  includes heatsink, but developer said it *probably* will need airflow, no power supply included

So, lets get real here, the BFL unit nearly triples the MH/$ returns even on the diminished claims of 850 MH/s.

And if you want to discuss the larger power draw, the BFL unit is currently doing about 80 watts, but BFL claims that could be do to runaway, and may run lower wattages once released.  But still, we know it runs 80 watts.  The nearest competitor, Icarus, is running aboutn 20 - 23 watts real world.  IT would take about 3 of these units to equal a BFL.  When you consider the efficiency of the wall warts in the additional Icarus units needed to make up the performance difference, the power differential is pretty much null.  Or so small, that it would take YEARS of mining to make up the cost differential between the BFL and Icarus.

So, I dont want to take anything away from the FPGA home brew guys, but BFL's product is substantially better in my opinion. Mind you, this is their FIRST iteration of the design.  If they do indeed deliver their product to market, they will be the ones to beat, and that will be a hard battle to fight for the home brew guys.

Do you have a unit of BFL with you to test out?

[grue]

they're basically using all this forum attention to sell their mediocre FPGAs, which by the time they ship it (if ever), would probably have the same performance as existing ones. even if everyone canceled their preorders because the specs don't match, they essentially got 2+ months of interest free loan.

First of all, I don't see any competing FPGAs with the same performance envelope.  Second, if by some chance someone does have a competing FPGA around the same envelope by that time, then nothing was really lost as far as "bait-and-switch" goes, since currently, there is nothing released that is even close to the neutered version that was demonstrated... and if they manage to get something that performs better than what was demonstrated, that's just icing on the cake.

>implying they won't lower the specs again
>implying it's not bait and switch if there isn't anything better
>implying they haven't overestimated before

I don't understand all the bashing on BFL.  Sure they missed the mark on their initial numbers. Its a first design. Even at ~850 MH/sec and $599 with the current specs you still are getting one of the best MH/$ thats currently availble at nearly 1.42  MH/$.  Thats substantially more than any other FPGA out there.  Not to mention, this will be a fully finished piece with a stackable enclosure, fan and power supply.

it's clear that they didn't even have a first design. it's obvious since the specs were WAY off. They just made a website, gathered a ton of money from preorders, THEN they started designing. And just because there still the best (assuming they don't lower specs again), doesn't mean it's right to overestimate by a factor of 3.

[CubedRoot]

What i don't understand is how they came up with those figures of 20W and ~1GH/s...You have to test to find these figures...

they're basically using all this forum attention to sell their mediocre FPGAs, which by the time they ship it (if ever), would probably have the same performance as existing ones. even if everyone canceled their preorders because the specs don't match, they essentially got 2+ months of interest free loan.

I don't understand all the bashing on BFL.  Sure they missed the mark on their initial numbers. Its a first design. Even at ~850 MH/sec and $599 with the current specs you still are getting one of the best MH/$ thats currently availble at nearly 1.42  MH/$.  Thats substantially more than any other FPGA out there.  Not to mention, this will be a fully finished piece with a stackable enclosure, fan and power supply.

BFL = 1.42 MH/$  --- Ready to go, with enclosure, fan, and power supply.

Icarus = 0.6 MH/$ --- Ships from China, does not include enclosure. Does ship with power supply and fans (although some say fans are loud)

ZTEX = 0.47 MH/$ --- No enclosure, No heatsink included (does need one), no power supply included.

X6500 = 0.38 MH/$ --- No enclosure,  includes heatsink, but developer said it *probably* will need airflow, no power supply included

So, lets get real here, the BFL unit nearly triples the MH/$ returns even on the diminished claims of 850 MH/s.

And if you want to discuss the larger power draw, the BFL unit is currently doing about 80 watts, but BFL claims that could be do to runaway, and may run lower wattages once released.  But still, we know it runs 80 watts.  The nearest competitor, Icarus, is running aboutn 20 - 23 watts real world.  IT would take about 3 of these units to equal a BFL.  When you consider the efficiency of the wall warts in the additional Icarus units needed to make up the performance difference, the power differential is pretty much null.  Or so small, that it would take YEARS of mining to make up the cost differential between the BFL and Icarus.

So, I dont want to take anything away from the FPGA home brew guys, but BFL's product is substantially better in my opinion. Mind you, this is their FIRST iteration of the design.  If they do indeed deliver their product to market, they will be the ones to beat, and that will be a hard battle to fight for the home brew guys.

Do you have a unit of BFL with you to test out?

Of course not, but several of the forum members have indeed confirmed the claims of around 850 MH/s.  Inaba has saw the unit in person on 2 occasions, and witnessed it doing said performance.  I am also no on any pre-order list either. I am however excited to see these things when they get out in the wild, and we start seeing real world mining with them.  I do plan on investing in FPGA mining in the near future, but I want to see the BFLs in real life before I do.  I am optimistically skeptical, but so far, BFL hasn't given me any reason to be turned off enough to completely bash them.

[grue]

Of course not, but several of the forum members have indeed confirmed the claims of around 850 MH/s.  Inaba has saw the unit in person on 2 occasions, and witnessed it doing said performance.[...]

i'm still interested why only 1 person has ever seen the unit. and that person seems to be biased towards BF.

[fred0]

Of course not, but several of the forum members have indeed confirmed the claims of around 850 MH/s.  Inaba has saw the unit in person on 2 occasions, and witnessed it doing said performance.[...]

i'm still interested why only 1 person has ever seen the unit. and that person seems to be biased towards BF.

Maybe he is biased because he saw the unit.

[CubedRoot]

they're basically using all this forum attention to sell their mediocre FPGAs, which by the time they ship it (if ever), would probably have the same performance as existing ones. even if everyone canceled their preorders because the specs don't match, they essentially got 2+ months of interest free loan.

First of all, I don't see any competing FPGAs with the same performance envelope.  Second, if by some chance someone does have a competing FPGA around the same envelope by that time, then nothing was really lost as far as "bait-and-switch" goes, since currently, there is nothing released that is even close to the neutered version that was demonstrated... and if they manage to get something that performs better than what was demonstrated, that's just icing on the cake.

>implying they won't lower the specs again
>implying it's not bait and switch if there isn't anything better
>implying they haven't overestimated before

I don't understand all the bashing on BFL.  Sure they missed the mark on their initial numbers. Its a first design. Even at ~850 MH/sec and $599 with the current specs you still are getting one of the best MH/$ thats currently availble at nearly 1.42  MH/$.  Thats substantially more than any other FPGA out there.  Not to mention, this will be a fully finished piece with a stackable enclosure, fan and power supply.

it's clear that they didn't even have a first design. it's obvious since the specs were WAY off. They just made a website, gathered a ton of money from preorders, THEN they started designing. And just because there still the best (assuming they don't lower specs again), doesn't mean it's right to overestimate by a factor of 3.

I clearly remember BFL stating that they were discovered before they even had the website up and running, and before any initial work had been done.  It sounded like someone leaked to the presses before they were ready to go public (after having testing a design).  So, like any company, they went with it.  They had tons of emails asking them to take pre-orders and what not, so they said they would.
I personally heard about BFL on a Linux Users group I am a member of, and not here.  There was alot of chatter about them elsewhere as well.

[RandyFolds]

they're basically using all this forum attention to sell their mediocre FPGAs, which by the time they ship it (if ever), would probably have the same performance as existing ones. even if everyone canceled their preorders because the specs don't match, they essentially got 2+ months of interest free loan.

First of all, I don't see any competing FPGAs with the same performance envelope.  Second, if by some chance someone does have a competing FPGA around the same envelope by that time, then nothing was really lost as far as "bait-and-switch" goes, since currently, there is nothing released that is even close to the neutered version that was demonstrated... and if they manage to get something that performs better than what was demonstrated, that's just icing on the cake.

>implying they won't lower the specs again
>implying it's not bait and switch if there isn't anything better
>implying they haven't overestimated before

I don't understand all the bashing on BFL.  Sure they missed the mark on their initial numbers. Its a first design. Even at ~850 MH/sec and $599 with the current specs you still are getting one of the best MH/$ thats currently availble at nearly 1.42  MH/$.  Thats substantially more than any other FPGA out there.  Not to mention, this will be a fully finished piece with a stackable enclosure, fan and power supply.

it's clear that they didn't even have a first design. it's obvious since the specs were WAY off. They just made a website, gathered a ton of money from preorders, THEN they started designing. And just because there still the best (assuming they don't lower specs again), doesn't mean it's right to overestimate by a factor of 3.

I clearly remember BFL stating that they were discovered before they even had the website up and running, and before any initial work had been done.  It sounded like someone leaked to the presses before they were ready to go public (after having testing a design).  So, like any company, they went with it.  They had tons of emails asking them to take pre-orders and what not, so they said they would.
I personally heard about BFL on a Linux Users group I am a member of, and not here.  There was alot of chatter about them elsewhere as well.

You are making quite a few assumptions here. How exactly could they 'be discovered' without a website or public office or advertising?

Only one person has seen a unit in action, and that was under BFL's conditions, not his own. They obviously didn't even have a working prototype or development board, because they would have seen the power issues a mile away. 10 years experience my ass...

[fred0]

they're basically using all this forum attention to sell their mediocre FPGAs, which by the time they ship it (if ever), would probably have the same performance as existing ones. even if everyone canceled their preorders because the specs don't match, they essentially got 2+ months of interest free loan.

First of all, I don't see any competing FPGAs with the same performance envelope.  

There is something being worked on that is in this range/better. It not public yet because there is no need to pre-sale... It will be at least 2 months before its made public.

Can't wait for the BFL widget, but willing to wait for this?  Surely you jest. Tick-tock, tick-tock, I think I hear the block reward halving.  Two months before an announcement.

[DeathAndTaxes]

What i don't understand is how they came up with those figures of 20W and ~1GH/s...You have to test to find these figures...

they're basically using all this forum attention to sell their mediocre FPGAs, which by the time they ship it (if ever), would probably have the same performance as existing ones. even if everyone canceled their preorders because the specs don't match, they essentially got 2+ months of interest free loan.

I don't understand all the bashing on BFL.  Sure they missed the mark on their initial numbers. Its a first design. Even at ~850 MH/sec and $599 with the current specs you still are getting one of the best MH/$ thats currently availble at nearly 1.42  MH/$.  Thats substantially more than any other FPGA out there.  Not to mention, this will be a fully finished piece with a stackable enclosure, fan and power supply.

BFL = 1.42 MH/$  --- Not ready to go, not available for shipping, only operated for limtied periods of time at extreme thermal and electrical stress.  Product isn't available for sale at this time.

Icarus = 0.6 MH/$ --- Ships from China, does not include enclosure. Does ship with power supply and fans (although some say fans are loud).  Product shipping now.

ZTEX = 0.47 MH/$ --- No enclosure, Heatsink included , no power supply included.  Performance per $ rises w/ bulk purchases reaching a peak of up to 1.1 MH/W.  It has double the performance per watt. Product Shipping now.

X6500 = 0.38 MH/$ --- No enclosure,  includes heatsink, but developer said it *probably* will need airflow, no power supply included. Product shipping now.

And if you want to discuss the larger power draw, the BFL unit is currently doing about 80 watts, but BFL claims that could be do to runaway, and may run lower wattages once released.  But still, we know it runs 80 watts.  The best competitor, ztek has 2x the performance per watt.

Fixed your post.

So, I dont want to take anything away from the FPGA home brew guys, but BFL's product is substantially better in my opinion. Mind you, this is their FIRST iteration of the design.  If they do indeed deliver their product to market, they will be the ones to beat, and that will be a hard battle to fight for the home brew guys.

They have no product ... yes.  Under the thermal and electrical load they have that board under it would be dead in a week.  So lets talk when they have a product which can run more than a 1 hour long limited test and is shipping?  Lets compare the product out then (4 weeks, 6 weeks, 8 weeks from now) to the performance under 24/7/365 load in the real world.  Remember this is the company which promised delivery in 4-6 weeks 10 weeks ago. 

Yes someone reputable shipping a product which gets <$1 per MH and >10 MH/W is impressive.  Of course that assumes it can stand up to 24/7/365 mining for 2+ years and you trust the company enough to make that kind of capital commitment.  Everything BFL has done so far indicates they are good at overpromising and under delivering.

[RandyFolds]

Why are people so intent on defending such a clusterfuck of a company?

[fred0]

Why are people so intent on defending such a clusterfuck of a company?

Because the criticism has been so hyperbolic. What's wrong with objectivity?

[ursa]

I clearly remember BFL stating that they were discovered before they even had the website up and running, and before any initial work had been done.  It sounded like someone leaked to the presses before they were ready to go public (after having testing a design).  So, like any company, they went with it.  They had tons of emails asking them to take pre-orders and what not, so they said they would.
I personally heard about BFL on a Linux Users group I am a member of, and not here.  There was alot of chatter about them elsewhere as well.

IMHO something may leak from INTEL,APPLE,SAMSUNG,ATI....ETC...it comes very hard for me to believe that something can leak from a mysterious one man-one cell phone company. Noone except Inaba knows this guy(s).  Where can the leakage be ? Maybe at the sink,in the kitchen ) or in the bathroom...

[RandyFolds]

Why are people so intent on defending such a clusterfuck of a company?

Because the criticism has been so hyperbolic. What's wrong with objectivity?

The valid criticism based on real world events, as opposed to the defense based purely on...I'm not sure what...maybe faith in humanity? Hope for a better tomorrow? The resilience of the human spirit?

[fred0]

Why are people so intent on defending such a clusterfuck of a company?

Because the criticism has been so hyperbolic. What's wrong with objectivity?

The valid criticism based on real world events, as opposed to the defense based purely on...I'm not sure what...maybe faith in humanity? Hope for a better tomorrow? The resilience of the human spirit?

I just find the emotional attacks less valuable than the objective attacks. For example, is it possible to say that the widgets work together seamlessly, when apparently, only two have been made?

I find it less valuable to say that BFL is a clusterfuck of a company, than to say that the ROI/payback of the widget is twice as fast as any of their competitors.

I find it hard to take seriously attacks of incompetence, when no one else has produced a widget with comparable hashing capacity.  I guess only incompetent people make technical advances.

I know that this sounds Spock-like, but the hyper-emotional posts are not very productive.

Live long and prosper.

[ngzhang]

Icarus = 0.6 MH/$ --- Ships from China, does not include enclosure. Does ship with power supply and fans (although some say fans are loud)

Hi, bro.

Ships from China: sorry, i'm proud of Icarus as a MADE IN CHINA product, and i put my name on it. i think of Icarus as my son.

does not include enclosure: 1st batch shipped with a full dev kit, include a platform cable, software in a 8GB usb stick, cables, adapter and standard heatsink and fan. the fan is made by japanese manufacturers and is ball-bearing.

(although some say fans are loud): i can sleep with 2 boards mining in my bedroom. why not try to sleep with 2 mining HD5850s.

[Inaba]

I sleep with 8 5870's and a 6990 mining just fine. 

Although mine are in the basement and I'm two floors up.

[rjk]

I sleep with 8 5870's and a 6990 mining just fine. 

Although mine are in the basement and I'm two floors up.

I love the sound of fans whirring away <3
If I could live in a datacenter, I would

[CubedRoot]

Icarus = 0.6 MH/$ --- Ships from China, does not include enclosure. Does ship with power supply and fans (although some say fans are loud)

Hi, bro.

Ships from China: sorry, i'm proud of Icarus as a MADE IN CHINA product, and i put my name on it. i think of Icarus as my son.

does not include enclosure: 1st batch shipped with a full dev kit, include a platform cable, software in a 8GB usb stick, cables, adapter and standard heatsink and fan. the fan is made by japanese manufacturers and is ball-bearing.

(although some say fans are loud): i can sleep with 2 boards mining in my bedroom. why not try to sleep with 2 mining HD5850s.

I meant no offense on being from China, it sounds as if you took offense to that. For me, shipping products from China to the US has always been a nightmare, and so I usually avoid it. Thanks for the other info, but you never confirmed (or I may have missed it) that ALL future batches will also ship with the 8gb USB stick and other cables.

As for all the other FPGA's I mentioned in my previous post, I wanted to do an apples-to-apples PERFORMANCE / PRICE comparison using the data we have confirmed to date. I do not make purchases or accusations based on emotions as it seems others have done in these crazy threads.  The more people scream SCAM!!! SCAM!! and WOLF!!! WOLF!!! without first having hard evidence, the less attractive bitcoin becomes to folks that are on the fence about it as a legitimate currency.  

Can someone provide me with solid proof, and not opinions, as to why Inaba's observations were not indeed true?  It also seems BFL has been running a legit operation so far. Granted, they have not come through with the expectations, but they were definately not trying to hide anything.  They also have been refunding all the money back to the skeptics that cancel their pre-orders.
Does anyone have any physical proof that BFL has indeed frauded anyone to date? I mean, have they taken money and ran? Have they not refunded the funds to the first group of people that wanted out due to the missed expectations?

I am asking these questions legitimately, since I am looking at investing a good bit of cash with an FPGA maker soon.

Also, what do you think BFL's competition will do once these units hit the market and perform great? Will they drop their prices to get more inline with the MH/$ that the Bitforce unit does? Will they try to develop a better FPGA?

BFL, have you guys copyrighted and/or patented your designs to keep other makers from copying your units once they are in the public?

[CubedRoot]

I sleep with 8 5870's and a 6990 mining just fine. 

Although mine are in the basement and I'm two floors up.

I love the sound of fans whirring away <3
If I could live in a datacenter, I would

I work in a datacenter many many hours per day.  We are running IBM, HP, and Cisco UCS cabinets. I love my job, and I love my datacenters (we have 4 total in different locations).  There is no way in hell I would want to put up with that whirring for more than I already have to. hehe.  Its standard procedure for us to mandate ear plugs when entering our datacenters.   Besides, with hot aisle / cold aisle HVAC, its impossible to find a comfortable temp. LOL.   60F with ice cold air blowing up from the floor plates on one aisle to 85F and dry air on the hot aisle.
Oh, and the lights on the front of the Freaking EMC Symmetrix VMAX racks are BLINDING!