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

[michaelmclees]

But they avoided that issue. Interesting info though. So what's this further unexplained delay?

*yawn* Reported. Tired of all the trolling. We can't even have a conversation in here without you butting in and being a dick. Don't you have a job, a hobby or something?

I don't see how that message is trolling.  BFL claims to have taken a course of action that would avoid potential delay... and then pushed back the release another 4-6 weeks.  What gives?

[1713969881]

1713969881

[1713969881]

1713969881

[n0ne]

But they avoided that issue. Interesting info though. So what's this further unexplained delay?

*yawn* Reported. Tired of all the trolling. We can't even have a conversation in here without you butting in and being a dick. Don't you have a job, a hobby or something?

I don't see how that message is trolling.  BFL claims to have taken a course of action that would avoid potential delay... and then pushed back the release another 4-6 weeks.  What gives?

Your right. When companies plan to release stuff, everything goes to plan and it's ALWAYS out on time! No exceptions! This seem fishy now that I think about it! /s

Good thing people don't get uppity every time a company is late on there release dates, oh wait.

[DeathAndTaxes]

Did they really push it back that far? If no how far back is the delay?

I don't think they said.  Product page says new orders will ship in 4 to 6 weeks.

Hopefully BFL learned a lesson.  None of this would even have been an issue if they had kept testing internally, not made promises based on simulations, and not offered pre-orders until board was in final production.

At this point I am torn.  Are they:
a) utterly incompetent.  I mean can anyone be that bad at launching a product?
b) manipulative and knew they wouldn't be ready for launch until Feb or March.  So they used the false promises/claims, and fake Nov delivery timeframe to freeze out sales on competing products until they were ready to launch.

Sadly I wish it was A (there business incompetence has little relevence on their electronic competence) but given the string of PR failures (the rig box still has what likely are impossible specs) I am leaning towards B.

[Inaba]

From my interactions with them, I don't think it was B.  I honestly think they had an overly optimistic design and release schedule.

[Matthew N. Wright]

From my interactions with them, I don't think it was B.  I honestly think they had an overly optimistic design and release schedule.

That's the feeling I got as well. We'll find out on our side in a week or so for the Bitcoin Magazine.

I hope anyone who has had any experience with the singles (no speculation, just actual experience) sends us their comments, complains and professional opinions in email to bittalk.tv@gmail.com

If they're fake, we'll expose them. If they're real, we'll support them. Whatayagonnado.

[DeathAndTaxes]

From my interactions with them, I don't think it was B.  I honestly think they had an overly optimistic design and release schedule.

Well I hope so.  Lesson to anyone else launch a product or company; keep your initial (and likely way overly optimistic) specs and timelines internal.  That way when you inevitably fail to meet them you only disappoint yourselves.  If somehow you do meet them you can "shock & awe" the market. 

I mean imagine if the launch had been like this:
Nothing.  BFL website didn't even exist pubicly.  Then one day a post by BFL saying they had a 800 MH @ 80W product for $800.  the 8-8-8.  Obviously a lot of "suspicion" so they already have 4 prototypes to give to senior members of the forum to test.  Inaba says "hey I live in same town".  They invite him over, he snaps some shots of their work area, sees single in production/testing, meets some of engineers, and gets a good impression.  He takes home a BFL single and post results that night.  Maybe to supplement that BFL hooks a BFL single to Deepbit, makes the API public, and provides local logs on a public website. 

1000 orders overnight, and unbelievable positive buzz.  Nobody would even know the product was 6 months late.  400% over original power spec and 20% reduced hash rate, and 20% higher priced than they could offer it at.  Now compare that to how the situation was handled.

[DeathAndTaxes]

That's the feeling I got as well. We'll find out on our side in a week or so for the Bitcoin Magazine.

Are you going to interview the company for next issue? 

I hope anyone who has had any experience with the singles (no speculation, just actual experience) sends us their comments, complains and professional opinions in email to bittalk.tv@gmail.com

No product has shipped so I believe Inaba is the only one who has had physical access.  Luke also IIRC had remote access to a single.

[Matthew N. Wright]

That's the feeling I got as well. We'll find out on our side in a week or so for the Bitcoin Magazine.

Are you going to interview the company for next issue? 

I hope anyone who has had any experience with the singles (no speculation, just actual experience) sends us their comments, complains and professional opinions in email to bittalk.tv@gmail.com

No product has shipped so I believe Inaba is the only one who has had physical access.  Luke also IIRC had remote access to a single.

Yep. They agreed to meet us at our UK offices where Vladimir Marchenko will do his testing and confirm capabilities etc.

We were planning on doing the review in this issue, but it's not looking like it's going to happen in time 

[P4man]

a) utterly incompetent.  I mean can anyone be that bad at launching a product?

Be fair. When was the last time AMD launched a product that was not delayed at least as much as BFLs product has been to date?  Delays are usually not measured in weeks or months, but in quarters.

[rjk]

a) utterly incompetent.  I mean can anyone be that bad at launching a product?

Be fair. When was the last time AMD launched a product that was not delayed at least as much as BFLs product has been to date?  Delays are usually not measured in weeks or months, but in quarters.

Heh, Christmas launch of the 7970, anyone? That's what they get for going fabless.

[DeathAndTaxes]

a) utterly incompetent.  I mean can anyone be that bad at launching a product?

Be fair. When was the last time AMD launched a product that was not delayed at least as much as BFLs product has been to date?  Delays are usually not measured in weeks or months, but in quarters.

AMD doesn't
a) public guaranteed specs based on simulations
b) provide official word on launch dates
c) indicate product will ship in 4-6 weeks for 4 months.

Often data will leak about a product and product will either not live up to that leak or meet the launch timeline.

It is far different if the company puts out an official spec or date and then fails to live up to their own promise.

The point is your right delays happen, product re-orgs happen, anticipated performance fails to materialize.

Since you brought up AMD here are two counter examples.  The 6000 series was suppose to be 32nm which would have made them roughly 50% more powerful in GFLOPS / W and GFLOPS / $.  However AMD never promised this, enver provided official specs, never set a deliver date.  The fab couldn't get 32nm working so AMD scrapped those plans, redesigned the chips and released then as 45nm parts.  The performance was lower (barely better than 5000 series) and the launch was late but AMD never broke any promises.

AMD obviously wanted the 7970 to launch before Christmas.  I mean it is Christmas.  The fab couldn't get it done.  AMD was disapointed.  Gamers were disapointed but once again AMD didn't break any official promise. 

[DeathAndTaxes]

Yep. They agreed to meet us at our UK offices where Vladimir Marchenko will do his testing and confirm capabilities etc.
We were planning on doing the review in this issue, but it's not looking like it's going to happen in time 

Nice.  Glad I subscribed.

[P4man]

AMD doesn't
a) public guaranteed specs based on simulations
b) provide official word on launch dates
c) indicate product will ship in 4-6 weeks for 4 months.

Often data will leak about a product and product will either not live up to that leak or meet the launch timeline.

It is far different if the company puts out an official spec or date and then fails to live up to their own promise.

AMD does most of the above to its customers, which is not us, but OEMs and OBMs. Not too pick on AMD, nVidia does the same, and I happen to know almost literally what you wrote above happened for both Tegra2 and 3. I know that from someone at a large mobile phone manufacturer. Of course large companies are a bit more experienced and therefore somewhat more careful in their wording, if for no other reason than huge contractual damages, but essentially they drop the ball the same way BFL did quite often.

[RandyFolds]

But they avoided that issue. Interesting info though. So what's this further unexplained delay?

*yawn* Reported. Tired of all the trolling. We can't even have a conversation in here without you butting in and being a dick. Don't you have a job, a hobby or something?

Nice. Did you put "I'm a whiny little bitch" in the comments box?

Go fuck yourself, fella. Feel free to report this post as well.

[bulanula]

Yeah. I am pretty sure they launched this like last year and with some impossible specs to ward off any potential competitors from developing anymore.

But that fully custom ASIC miner that gets 10 ghash/s at 10W and costs $100 is still on the horizon so BFL beware !

[DiabloD3]

But they avoided that issue. Interesting info though. So what's this further unexplained delay?

*yawn* Reported. Tired of all the trolling. We can't even have a conversation in here without you butting in and being a dick. Don't you have a job, a hobby or something?

Nice. Did you put "I'm a whiny little bitch" in the comments box?

Go fuck yourself, fella. Feel free to report this post as well.

Ahh, a fine afternoon in off-topic, indeed.

[ZodiacDragon84]

to /b/, or not to /b/, that is the question?

[makomk]

[Look the Statix III is a 65nm chip.  Alterra isn't even making it anymore.  They want to sell Stratix IV and soon they will need to sell Stratix V.  They want that inventory gone.  Gone from their website, gone from wholesalers, gone from retail outlets.  when Stratix V is in full production it becomes the high end part and the Statix IV becomes the value segment.  The Stratix III is just a third wheel.

I'm not sure it works that way, though - there's always customers out there who've built designs around older chips and for whatever reason either can't or don't want to move over to the latest and greatest. From what I've seen the FPGA vendors are really slow at taking older chips out of production for this exact reason.

Case in point: The CEO (and probably sole proprietor) of ZTEX did.

Before the 1.15x module, with its 8 Amp core voltage supply, came the 1.15d module, and the originally recommended supply for it
http://www.ztex.de/usb-fpga-1/pwr-1.0.e.html only sported a 3 Amp core voltage supply!

He found out the hard way that all these unrolled loops of SHA-2 cause something like 50% of all flip-flops on the FPGA to switch simultaneously, and
thus blowing even the most conservative power estimations out of the water.

That board wasn't really designed for Bitcoin mining in the first place - he's been selling them on eBay for ages as generic FPGA development boards.

[RandyFolds]

But they avoided that issue. Interesting info though. So what's this further unexplained delay?

*yawn* Reported. Tired of all the trolling. We can't even have a conversation in here without you butting in and being a dick. Don't you have a job, a hobby or something?

Nice. Did you put "I'm a whiny little bitch" in the comments box?

Go fuck yourself, fella. Feel free to report this post as well.

Ahh, a fine afternoon in off-topic, indeed.

...and an education in LUTs, bitstreams, and chip stats. It really can't be beat.

To the FPGA guys here: Why is it 'rolled' and not 'furled'? It seems way more appropriate.

[DeathAndTaxes]

[Look the Statix III is a 65nm chip.  Alterra isn't even making it anymore.  They want to sell Stratix IV and soon they will need to sell Stratix V.  They want that inventory gone.  Gone from their website, gone from wholesalers, gone from retail outlets.  when Stratix V is in full production it becomes the high end part and the Statix IV becomes the value segment.  The Stratix III is just a third wheel.

I'm not sure it works that way, though - there's always customers out there who've built designs around older chips and for whatever reason either can't or don't want to move over to the latest and greatest. From what I've seen the FPGA vendors are really slow at taking older chips out of production for this exact reason.

They are slow which is why the Stratix III is still around despite being released almost 5 years ago.  Still you notice Stratix II isn't available.  Companies tend not to like to keep 3+ generations going at the same time.  As the Stratix V nears volume production the company will try to transition customers to the newer products.  For large customers they will even provide incentives, demo units, and design assistance to move their bitstreams to newer products.