Topic: BFL 6 May 2013 ASIC Update - page 6. (Read 23357 times)

You are trying to deflect what was a pretty straightforward question. As far as I am concerned the Avalon chip DIY based project are not  an option until the designs and chips are shipping from Avalon, and then starts the actual building of products like the Klondike and assessment of market penetration. At that point you can start making claims as to BFL's relative success in the marketplace, before that it's wild speculation.

No more vapourware to those on the back end of the BFL order train right? Especially given what we are seeing from the open source designers. There is more than just an idea here it is moving forward and this will happen. And to be brutally honest who you would put more faith in at this point? Avalon releasing the specs and DIY'ers making these boards work or BFL?

Understood, but the Avalon chips are vaporware atm, possibly months off. Have they published the schematic and open source board design yet? It was meant to be May

This will be mass produced (if it works) and replicated locally across the world... BFL will not get a second run at this. It could be game over for them and given their delivery delays I can't see anyone going with them ever again when these cheaper and better built alternatives come out in a few months.

I am waiting on my 540 chips... and my plan when they are up and mining is to reproduce the boards locally here in Indonesia for the Indonesian market only. I won't be the only person doing this trust me things have changed and BFL will have to either change or face the fact they will be pushed completely out of the market.

So if BKKcoins actually builds a functional Klondike, a 16 chip board that could clock to 300 Mhz per chip on a 10 cm x10 cm square at 32W as a DIY project in a few months what the hell has BFL been doing these past 10 months?



And costing roughly https://github.com/bkkcoins/klondike/blob/master/docs/Parts%20List.pdf what only $26.48 per manufactured board with runs of 25+ board minimum and $8 for an Avalon chip and then in heatsinks, the psu and time and sweat you could put have the following.

16 Avalon Chips............................  $128.00
1 Board......................................... $ 23.68
Corsair CX-600 PSU.......................... $70.00
1 Extruded Al Heat sinks....................  $6.00
Blood Sweat Tears.........................  $100.00
USB cable........................................ $2.00
32+ Watts.
4.8 GH/s
No extra shipping charges
Easily scales to 72 GH/s

For 330 USD?

Hmmm.

I see a tough road ahead for any BFL products if the Avalons DIY'ers get this working. This might prove the end of BFL considering this will be open source.

So instead of 8 chips, the single will use 16 chips, it is a nice move that leaves a lot of room for chips. I remember that they said the test for running at 500Mhz is problematic but chips can run stable at 300Mhz, e.g. 4.8GH per chip

Each Klondike board has a 6 pin PCI Express power connector allowing efficient powering of many boards from a decent ATX PSU. A 16 chip board should be capable of 4512 MH/s and consume about 32W power. A low cost Corsair CX-600 PSU should easily power 16 boards using readily available power splitters. This provides for 72 GH/s off one low cost ATX PSU.

Another useful BFL thread ruined by trolls.
Are there no mods?

I am use to a forum with over 10,000 active members, where mods would have deleted half the posts in this thread in seconds as off topic or abusive.

There are some changes we are making to the metal layers to hopefully reduce power. In the meantime, we have several of the initial wafers that are going through packaging and we are dialing in the packaging so that our yield rate is acceptable... that takes time and the packaging house we were using is performing... less than adequately. We've changed packaging houses and the first chips off the line made it through today and the yield rate is @#$#@ spectacular and The Josh is pleased. Those chips should be making their way to the testing/binning facility and then on to Chicago on the 9th to be mounted. It's just another handful of chips to make sure everything is kosher as far as things go, and assuming all goes well, we should have a ton of ships queued up and ready to go once we are confident that the new packaging facility can get into gear. We are still chewing through the first 5 wafers, but slowly so we don't waste any more chips than we have to, while we dial in the packaging and substrate manufacture. I'm not sure how many of the 5 wafers have been used so far. The next 5 have some modifications which should be ready now I believe and the remaining wafers have some additional modifications in the metal layer to reduce power even further, so that will be the real interesting test once those arrive in KC.

After those chips, we have some more on order that have more substantial changes in the upper layers to further reduce power as well as several other irons in the fire, but all that is a couple months away, about the time we are ready for more chips and have worked our way through the first set.

Same, if you check the forum stats here https://bitcointalk.org/index.php?action=stats  you will see that the worst trolls of BFL threads are also the turbo posters right across these forums.

I have most of them ignored, what gets me is the retards who quote them.