Pages:
Author

Topic: Mini Rig announcement by Butterfly Labs - 25gh/s - page 3. (Read 34251 times)

sr. member
Activity: 336
Merit: 250
I almost get the feeling they arbitrarily delay postage. like they wanna mine themselves as much as they can before shipping it. Ofcourse they deny this.
The current wait is because of a lack of cases, apparently the boards are ready to ship but there's nothing to put them in.

Yea that's the Singles, but what about Mini-rig, they demonstrated all the boards, some guy confirmed he saw the Mini-Rig cases at their office. They promised to start shipping end of may.
rjk
sr. member
Activity: 448
Merit: 250
1ngldh
I almost get the feeling they arbitrarily delay postage. like they wanna mine themselves as much as they can before shipping it. Ofcourse they deny this.
The current wait is because of a lack of cases, apparently the boards are ready to ship but there's nothing to put them in.
sr. member
Activity: 336
Merit: 250
I almost get the feeling they arbitrarily delay postage. like they wanna mine themselves as much as they can before shipping it. Ofcourse they deny this.
hero member
Activity: 504
Merit: 500
FPGA Mining LLC
Here's a quick photo update for you...

This is the Mini Rig processor card.  It's a 1.5 GH/s unit which operates at a board temperature of 36C with a 22C ambient.  

http://www.butterflylabs.com/wp-content/uploads/2012/05/MR-BF2_burn.png
http://www.butterflylabs.com/production-update/
That SATA port looks interesting.

I was about to post the same thing, but it's labeled "X-LINK" and since a SATA port doesn't make sense (the bandwidth is not that high and the mini rig will have a Raspberry Pi as a controller, not a a PC motherboard with 18 SATA ports), I assume that they just re-purposed the SATA connector and SATA cables for something else. The question is: For what?

I know I'm late, but the answer is: For that!
mrb
legendary
Activity: 1512
Merit: 1028
Can't edit post above but it seems there is this writing on the chips this time around :

NDBAN20242
4PDHATWOA

Can't find anything about these but they look very strange indeed for a FPGA chip.

These are Altera's Lot Number and Trace Code. https://bitcointalksearch.org/topic/mini-rig-card-2-x-altera-arria-ii-ep2agx260-84651
sr. member
Activity: 348
Merit: 250
What I'd like to know is how this is going to interact with the host unit. Since it's "BitForce V 2.0" it seems like it would just work with cgminer, et al. just like the Singles. Does anybody know for sure? I haven't seen BFL make a statement on this as such. Maybe they could (wink wink)? Also, would it be a problem if it was reported to this host as one huge device doing 25,000 mhash? These are things I would wonder about if I were to drop 15.5k on dedicated mining hardware.

These are valid questions, but with them showing the Raspberry Pi inside I was thinking it was its own host unit.

My understanding was that the Raspberry Pi was optional - that it had the mounting area for it but that it didn't ship with the unit.

That's correct.  Sonny told me yesterday that you can either run the mini-rig with a Raspberry Pi or via USB to a host PC.
member
Activity: 80
Merit: 10
What I'd like to know is how this is going to interact with the host unit. Since it's "BitForce V 2.0" it seems like it would just work with cgminer, et al. just like the Singles. Does anybody know for sure? I haven't seen BFL make a statement on this as such. Maybe they could (wink wink)? Also, would it be a problem if it was reported to this host as one huge device doing 25,000 mhash? These are things I would wonder about if I were to drop 15.5k on dedicated mining hardware.

These are valid questions, but with them showing the Raspberry Pi inside I was thinking it was its own host unit.

My understanding was that the Raspberry Pi was optional - that it had the mounting area for it but that it didn't ship with the unit.
donator
Activity: 490
Merit: 500
What I'd like to know is how this is going to interact with the host unit. Since it's "BitForce V 2.0" it seems like it would just work with cgminer, et al. just like the Singles. Does anybody know for sure? I haven't seen BFL make a statement on this as such. Maybe they could (wink wink)? Also, would it be a problem if it was reported to this host as one huge device doing 25,000 mhash? These are things I would wonder about if I were to drop 15.5k on dedicated mining hardware.

These are valid questions, but with them showing the Raspberry Pi inside I was thinking it was its own host unit.
legendary
Activity: 1666
Merit: 1000
cgminer has handled well over 25 Gh/s in one instance - it is ready...
member
Activity: 80
Merit: 10
What I'd like to know is how this is going to interact with the host unit. Since it's "BitForce V 2.0" it seems like it would just work with cgminer, et al. just like the Singles. Does anybody know for sure? I haven't seen BFL make a statement on this as such. Maybe they could (wink wink)? Also, would it be a problem if it was reported to this host as one huge device doing 25,000 mhash? These are things I would wonder about if I were to drop 15.5k on dedicated mining hardware.
sr. member
Activity: 448
Merit: 250
So the Single is an FPGA so can be reused for a similar task, but mini-rig is a proper ASIC so application specific, no reuse possible ?

It's my best guess - don't take it for a fact just yet.
sr. member
Activity: 336
Merit: 250
So the Single is an FPGA so can be reused for a similar task, but mini-rig is a proper ASIC so application specific, no reuse possible ?
sr. member
Activity: 448
Merit: 250
I think this time it is a structured ASIC.

They handed over some 300 grand to Altera to have a HardCopy chip (structured ASIC) made.
Naturally, without the parasitic capacitance of all the interconnect transistors, the HardCopy is faster and more power efficient.
hero member
Activity: 518
Merit: 500
Can't edit post above but it seems there is this writing on the chips this time around :

NDBAN20242
4PDHATWOA

Can't find anything about these but they look very strange indeed for a FPGA chip.
hero member
Activity: 518
Merit: 500
OK. Anyone have an idea of the chips they are using now ?

750 MHash/s for a single chip really is strange and very high.

I am afraid this is the death of GPUs Cry unless 7990 comes out soon.

Any update for EU buyers and VAT avoidance Huh I really want to buy but the VAT kills it.
legendary
Activity: 922
Merit: 1003
I have this theory that at least some of the cooling problems of [few] "problem Singles" originate from a lack of co-planarity between the two FPGAs, leading to a void between heat sink and FPGA at one or even both of the FPGAs. OK, not literally a void, but an area where the thermal paste is very thick, and that's not good.

That is a good theory, Inspector, and I agree with you. Cooling performance of the Rev3 units (single heatsink servicing both FPGAs) is dependent to some degree on the co-planarity of the chips, which is difficult to achieve in practice. If the 2 FPGAs each had their own heatsink the cooling performance might be better. It seems to be sufficient, however, to achieve the rated 832Mhps at 22C ambient with the stock fan(s). Perhaps not ideal or optimized, but sufficient. At this point I tend not to complain.
legendary
Activity: 938
Merit: 1000
What's a GPU?
If I could get them sooner without the enclosure and the Raspberry Pi, I'd rather have my 18 boards as-is. OK, heat sinks would be nice.

Speaking of:
I have this theory that at least some of the cooling problems of [few] "problem Singles" originate from a lack of co-planarity between the two FPGAs, leading to a void between heat sink and FPGA at one or even both of the FPGAs. OK, not literally a void, but an area where the thermal paste is very thick, and that's not good.

It also seems common for the heatsinks to slip while shipping.

BFL, what are you doing to accommodate this?
sr. member
Activity: 448
Merit: 250
If I could get them sooner without the enclosure and the Raspberry Pi, I'd rather have my 18 boards as-is. OK, heat sinks would be nice.

Speaking of:
I have this theory that at least some of the cooling problems of [few] "problem Singles" originate from a lack of co-planarity between the two FPGAs, leading to a void between heat sink and FPGA at one or even both of the FPGAs. OK, not literally a void, but an area where the thermal paste is very thick, and that's not good.
rjk
sr. member
Activity: 448
Merit: 250
1ngldh
Thank you for the update!

Anyone noticed the voltmeter measurement points at the top right ?
Singles have those too. But I see all kinds of cool goodies on the board, to be honest.

sr. member
Activity: 336
Merit: 250
Thank you for the update!

Anyone noticed the voltmeter measurement points at the top right ?
Pages:
Jump to: