"Shipping the week after Thanksgiving" ... but not really, we actually mean "The first week of December cause we can't read calendars." (OOps, maybe not, we don't know when we are shipping yet!) We will ship the "5th week of November"!? Someone can't throw any stones in their glass house... I think every ASIC manufacturer should avoid any stones, since the only one that hasn't had a delay is Avalon and they aren't close enough to shipping that they would need to announce a delay yet. Get your shit out the door and then you can mock the competitor and steal his customers. Until then, focus on your own product. Sage advice if I ever heard it! |
|
|
|
"Shipping the week after Thanksgiving" ... but not really, we actually mean "The first week of December cause we can't read calendars." (OOps, maybe not, we don't know when we are shipping yet!) We will ship the "5th week of November"!? Someone can't throw any stones in their glass house... |
|
|
|
Orders canceled. Not dealing with this shit (BFL) anymore.
Well, at least you stood your ground and gave them every opportunity to come through.... Looks like the first (safe) wave of refunds have begun or are ongoing. The second wave of refunds is where the risk is at (IMO). |
|
|
|
According to a BFL representative, the reason why they were late was exactly because of those changes to their specs. So no, not as crazy as it sounds.
Didn't BFL say at the time that they'd originally intended to release a post-delivery firmware update which would have taken the specs to 60 GH/s? That was the original implication. Later followed by apparent hardware level changes to support the increased speed. Hence, the extra delays. BFL might have originally shipped at their original spec much sooner. Though that train of logic is broken (or borked) by the problem they appear to still face: namely, sourcing components. They appear to be suffering from a component shortage or late delivery of components [other than the ASICs[. Everyone's pretty much assumed since then that all ASIC providers would have the capacity to offer firmware updates which would deliver higher specs. The question was really whether they'd do it pre or post-delivery and how much they'd try to keep in reserve. There is no simple answer to this. We don't really know if it is even possible to just push a firmware in and make it work. We would have to ask for confirmation from each vendor to figure that out. If you want to screw your competitors you want them to start delivery while you still have time to make hardware modifications which increase your maximum specs beyond theirs, not just do something they can match with a firmware tweak. Hell, if you want to be really nasty, you don't even announce how much better your specs really are until you start shipping.
Actually, many people would probably prefer to mine first and hope for upgrades in the form a simple firmware upgrade. But there is little proof that A) This feature of firmware upgrades is as simple as a CLI application. (could be like a J-Tag style of firmware upgrades) B) That the competition would be negatively impacted by simply announcing your specs late. It pays much more (IMO) to delay the competition by forcing them to change parts and components. Everytime they do, they have to make hardware or software changes which costs time and money. In this latest round of upgrades, BFL has basically stated they will not change their specs or design. They instead will prioritize their shipments first. Whether a customer can remotely upgrade their hardware is yet to be seen. This isn't exactly an Intel/AMD or Broadcomm chipset and design implementation. So who knows how easy it is to upgrade. It is something we should ask the Vendors directly to get a straight answer. |
|
|
|
BFL did that to Avalon.
FTFY Avalon was originally supposed to sell their Pre-orders for 1299, and later sell their hardware for 1999. They could get away with this because they were 50% faster than BFL, and over 2x as fast as the bASIC. When BFL upped their specs (due to the bASIC upping theirs), Avalon had to lower their price to remain competitive. And the reason they went from 40Gh/s to 60 is obvious. It wasn't more than a week or two ago that one BFL representative mentioned why they kept waiting on new revisions. Because of the speed bump increase. |
|
|
|
Nah mate it's ridiculous.
Ok, but when were the changes implemented on Tom's design? Was it in late November? Don't assume, simply ask Tom when his device obtained the 72Gh design changes. |
|
|
|
That sounds really dumb.  You haven't read many of hahahafr's posts, have you? Actually, it is not so far fetched. Avalon did that to BFL. And there was a long delay as BFL responded to the changes in their hardware. According to a BFL representative, the reason why they were late was exactly because of those changes to their specs. So no, not as crazy as it sounds. ------------------------ The question is, when did Tom get the changes in and when did his engineering firm ask him to consider the change to 8 chips? He doesn't give a real timeline as to when that all happened. |
|
|
|
|
@ Those last two posts above me.
The sad thing is that the rep laughs at his former customer choice for going to "Plan B". Then begins to create the impression that "Plan B" Vendor is a total scum of the earth. Portraying his competition as a continual liar of sorts. (sorta)
Not realizing that his customers was "forced" to use "Plan B" because in "Plan A" [the Vendor] was horribly unable to keep their deadline and has indeed given inaccurate dates again and again, never quite fulfilling their orders on a scheduled time.
Question: What does the "Plan A" rep have to gloat about in this circumstance??
The irony is in there somewhere. Can you see it?
|
|
|
|
He used to annoy me, back when I thought he was a for realz company rep, but now he cracks me up.  The fail volume is off the charts with nearly everything he posts. I should be thanking him, if it weren't for his antics I'd probably be near the back of BFL's waiting list instead of Tom's. Whew, bullet dodged. TY JOSH! BFL_Josh quote: If there is work to be done on T-Day, we will be doing it. No turkey shall stand in the way of ASIC progress, this I promise you. The main update is that next week our Customer Service staff will be off Thursday and Friday to celebrate Thanksgiving, so get your questions and orders in early in the week. https://forums.butterflylabs.com/blogs/bfl_jody/47-im-still-here.htmlGuess it didn't happen.  There was nothing to do on T-Day, so they took their holiday vacation and played some minecraft. |
|
|
|
|
For a multi-million dollar business, this shouldn't even be an actual issue....well, not for such a long and protracted period of time.
That it is...
In retrospect, they probably should have spent the cash on the fabs that do charge more but will actually stay on schedule (most of the time).
|
|
|
|
|
By the way, No BFL representative has to actually go to China. They can hire a Professional Agent overseas to do the nagging and inspections of the work being done.
|
|
|
|
What exactly do you want me to do? Fly to Asia and knock on someones door?
Let's be realistic here. I can't very well force someone to give me a date that's accurate, since by forcing them I might be putting them in a position to give me a false date just to get me off their ass... and that helps no one. Flying to Asia and knocking on someone's door might not be a bad idea at this point. You've basically implied that even if they give you a date you can't trust it to be accurate - which says something about your suppliers in and of itself. Indeed. High fab prices keep those little companies orders at bay. (well for the low nm scales) I realise that some of the figures regarding the value of BFL pre-orders are probably widely over-stating the reality, but I'd also be willing to bet that the value of this particular contract is significant. It's not just your reputation and future business which is on the line - your supplier's reputation is on the line too and you can leverage that shit. No comments. "We'll send you the chips when we get around to it and we can't tell you when that will be" is not an acceptable response and you know that. Only when you pay for it. You pay less, you get worse quality, thats always been the case in most industries. I hope to hell you have penalty clauses in your contract because for all your bravado you must realise that this is hurting BFL and that "they won't give us a date" isn't going to fly with your users. This is a bullet run, FFS. Some ass needs to be kicked and I can't think of anyone more qualified than you to do it.
If it is a cheapo fab...why would they add clauses where they would lose money on a consistent basis? Hell they probably have a clause to cover them against all liabilities. [Wild Speculation and joking opinion only] |
|
|
|
What exactly do you want me to do? Fly to Asia and knock on someones door?
Let's be realistic here. I can't very well force someone to give me a date that's accurate, since by forcing them I might be putting them in a position to give me a false date just to get me off their ass... and that helps no one.
What reputable business would not be able to give you an estimate of when they'll have the chips, that's just absurd. Are you guys using a 3rd world sweat shop or something to pump these things out? Also... why is BFL too cheap to pay for an expedited order? EDIT: You don't even have to respond to those questions. I know why, because you guys have yet to design a working chip. You're probably on your 3rd revision by now and it doesn't even work. It is cheap for a reason? (65nm) Imagine if everyone could pop out 65nm chips at a cheap and affordable rate (like BFL has sourced [allegedly]). Why wouldn't their chosen fab be backed up? |
|
|
|
Well heck, by that criteria BFL is fine, too... since we've always maintained that we would would target October but it would be as late as December (hence the January 1st refund policy, which has been there since day one.)
Thanks for the confirmation that BFL is on target timeframe wise as well, I appreciate it! ...and it's posts like this that pushed me away from BFL. Who is Inaba and why does he say "we've" when he's talking about BFL? I don't know who you are talking about? Whenever I see that screen name I see the name Inability. See take a look  Oh that's right he wears two hats but consistently forgets to swap them when acting as a BFL mouthpiece. BTCFPGA's potentially slipping shipping schedule in no way parallels BFL's estimateS and missed shipping scheduleS. You should watch some training videos on how to represent a company. The McDonalds around the corner has more professional personalities representing them at the drive thru window.
Well, thats pretty harsh. |
|
|
|
I could see how power consumption could be a big deal for someone paying .2 or .3, but I don't think that most people pay that much, especially in the US.
{snip...]Or maybe all miners will relocate their machines to US data centres in states with cheap electricity (that'd be my preferred option). But without data on how many miners pay a certain amount for electricity, you really can't know if there's going to be a "chilling effect" on the network hashrate any time soon. How much does this cost? Is it affordable to do this? If so, how does someone go about doing what you described? (collocating I think it is called.) It would simply involve shipping your hardware to the datacenter location - and then paying a housing fee (think electricty + bandwidth costs + a small margin for profit). Personally, I'd be happy to house your mining equipment - you can manage it via remote session and I'm sure we could work out a percentage based on what your current power costs are. For us to find a starting point, what are your current power draw and cost? If I am not mistaken it is around 0.095 cents per kwh. (It fluctuates as the electrical company here changes the rates every few months.) Edit: I have heard that Datacenters pay the cheaper wholesale rate of .045 -> .01 per kwh. I also asked because I have just purchased my first [used] blade server and would consider collocation. I don't mine so my current costs are zero in that vein. |
|
|
|
Historically fpga to asics upgrades assume same.
price points are price points.
Explain? I traded my BFL FPGA Single in for $599 on a 60gh ASIC Single,then just paid the difference of $1299=$700  BFL did say they would do trade in's on first gen ASIC's to second gen ASIC's,whether it's the same deal as the FPGA's wasn't laid out,but I'm assuming it is. Just make sure you get them to say "specifically" what they will offer you in writing and you are good to go. (Always cover your bases as the saying goes.) An FPGA can be re-purposed so it has value. A first Gen ASIC is kinda of a paperweight after its life cycle. (Unless the Smithsonian wants a batch of them for their museum? (theme, the first successful cryptocurrency?) |
|
|
|
I could see how power consumption could be a big deal for someone paying .2 or .3, but I don't think that most people pay that much, especially in the US.
{snip...]Or maybe all miners will relocate their machines to US data centres in states with cheap electricity (that'd be my preferred option). But without data on how many miners pay a certain amount for electricity, you really can't know if there's going to be a "chilling effect" on the network hashrate any time soon. How much does this cost? Is it affordable to do this? If so, how does someone go about doing what you described? (collocating I think it is called.) |
|
|
|
Overclocking a video card that's $200 and you have a huge amount of previous overclocking information from other people at your disposal is one thing. Overclocking a few thousand dollars ASIC rig that noone has any idea how it'll handle that short or long term is a whole different ballgame.
I agree, they should just do a batch revision in the interim after shipping what they now have and then offer an upgrade for people who bought the 60. Maybe upgrade them to 80 or 90 and then resell the 60's to other customers as a separate SKU. It would suck to get a rig and one month later end up with a dead chip(s). That would be pretty hard to fix and there would be almost no workaround for that. (RMA's and the confidence hit wouldn't be worth it) |
|
|
|
Historically fpga to asics upgrades assume same.
price points are price points.
Explain? |
|
|
|
|
Show Posts
