Topic: BFL ASICs - page 137. (Read 557782 times)

БФЛ выслало Luke-jr первый асик 31 марта. (все кто делал ставки что не выпустит до 1 апреля проиграли)

выслало Luke-jr? в EXIFe одной из картинок автор - Josh
это только прототип  

что не понятно, так это каким образом 179 ватт рассеивается через такие маленькие радиаторы без вентилятора?

БФЛ выслало Luke-jr первый асик 31 марта. (все кто делал ставки что не выпустит до 1 апреля проиграли)

start shipping next week

site:forums.butterflylabs.com "start shipping next week"

We should receive the completed boards in KC a couple days after they make it through assembly, so it's unlikely we'll start shipping next week at this point; probably the following week.

A couple things I would like to address:

The mental giants (You know who you are) who are crying about the "lights dimming" at 17 seconds into the video are apparently confused about how autoexposure on a phone camera works. It does this throughout the video, cooincidentally when I point the camera at the bright computer screen. This is one of the reasons I did not want to just "break out the cell phone and take a video" because of stuff like this. A more polished video is what's needed, but I went ahead and did it, and now I get to listen this kind of crap from people that can't even tell a reflection from substrate. Sorry trolls, but yet another fail.

The board used in this test is drawing more current than it normally would due to a variety of reasons. Yet another reason why I did not want to shoot a premature video - this is not a production board. This is a test board that is not designed for these kinds of power loads. We already have a revised board finished and produced, I just don't have it in the labs yet. We are going through another revision as we speak and those boards should be ready by the time the chips arrive in Chicago for mounting. The final shipping revision of the boards will likely consume far less power at idle.

We will not be shipping a board that draws 195w anytime in the near future. We probably won't even be shipping a board that consumes 160w. A more realistic target right now is somewhere around 100w - 120w.

The aluminum heatsink on the board is what all boards except the first few hundred were going to eventually ship with. We switched over from copper to aluminum due to cost and the fact that they perform virtually the same. We did this quite awhile ago and our entire stock of heatsinks are the aluminum variety, except for the first batch we ordered in all copper.

The remaining 4 wafers are being packaged next week and should be by the end of the week and shipped to the assembly facility at that point. We were hoping they'd be finished by this weekend, but that doesn't look like it's going to happen. We should receive the completed boards in KC a couple days after they make it through assembly, so it's unlikely we'll start shipping next week at this point; probably the following week.

Кто нибудь знает почту Джоша? Я вам точно отвечу, врет он или нет  Б...дь не смог найти даже на их тупом продвинутом форуме.

Возможен вариант уменьшения тактовой частоты и увеличения кол-ва чипов?

Если серьезно, то никакой дизайн СО ничем не поможет, если энергопотребление получилось сильно выше расчетного. Потому что в таком случае придется либо разбираться с энергопотреблением (с токами утечки, что решается редизайном ядра / тюнингом техпроцесса), либо менять корпус. И то и другое - непростые задачи и потребуют кучу времени. Если кто-то пишет в стиле "у нас пластиковый корпус и энергопотребление 120Вт, но мы ща быстро это решим" - он врет и у него просто ничего нет.