Pages:
Author

Topic: 1GH/s, 20w, $700 (was $500) — Butterflylabs, is it for real? (Part 2) - page 4. (Read 146930 times)

legendary
Activity: 3878
Merit: 1193
Still if 700 MH/s is what is sustains and it pulls 80W at the wall we are looking at 8.75 MH/W.

Even 80 watts is optimistic:

Consumes between 85 and 90w at full load.
legendary
Activity: 2114
Merit: 1031

Thank you for these enlightening results.

Thus, it seems that the overclocked 7970 can get more than this unicorn BS.

I mean WTF Huh

                                7970   |   BFL
 
Cost :                        550     |  599
Performance :                700   |  700
Warranty :                 3 years |  6 month
Resale value :                 50% | 0%

The ONLY advantage I see is the consumption of power but surely that is not worth it for some US guys that have 0.10 prices. Maybe this is for EU market but VAT and import duty kills that as well so it is LAME right now from the promised holy 1000 MHash/s, 20W, $500 figures ...

BFL deserves a medal for fail of the year.  

Thermal throttling due to bad design FTW !

I think the advantage of plugging it into a USB drive as well as being able to "stack" the singles is advantageous.  My wife is complaining about the noise & effects of me over clocking the GPU (glitchy screen display).
hero member
Activity: 798
Merit: 1000


                                7970   |   BFL
 
Cost :                        550     |  599
Performance :                700   |  700
Warranty :                 3 years |  6 month
Resale value :                 50% | 0%




Wattage:          574W | 80W

http://www.hardocp.com/article/2012/02/08/gigabyte_radeon_hd_7970_oc_video_card_review/8

You missed that little tidbit Wink

Ok BFL uses less wattage (but more than any other FPGA) but 574W?  Really I mean lets as least be slightly realistic shall we?

Blame HardOCP Smiley
legendary
Activity: 2044
Merit: 1000

I am sure they never thought it could deliver 1.05 (remember it wasn't an estimate but 1.05 GH 2 digits of significant value) at 19.8W.  That was simply a tactic to freeze out any sales of competitor products.



Which really cuts to the core of the issue, no?  It does the entire BTC community a diservice.  I imagine at least a couple enterprising individuals ceased developing their own products when they saw the "magical" specs for the bitforce.  Flash forward 5 months and loe and behold the actual product is really not that much more efficient than a undervolted 5970.  

It is too bad, as the more options for FPGA there are, the better it is for all of us as the proverbial "consumer".

I would love to announce a revolutionary vehicle I invented that gets 90 MPG (based on projections), at a price of $17,000.  I would be inundated with pre-orders and have gobs of cash to then develop my product.  Then when I actually delivered, and it only gets 63 MPG (not much better than a Prius), I would be thrown in jail for fraud.  



First of all, the car analogy is crap, since a Prius only gets good milage under a very limited set of circumstances.  Any modern diesel is far superior to the current crop of hybrids as far as fuel efficiency goes.  So comparing an FPGA to a car, you'd need to say XXX FPGA gets 250 MH/s only when you give it EXACTLY 121.742v, at 72F ambient temperature, with the fans spinning at 2311 RPM.  Anything other than that, and it doesn't get those specs.  It's ridiculous is what I'm saying.

I very seriously doubt that any enterprising individuals were/are the least bit dissuaded by another FPGA entering the market.  If anything, it would have spurned them to figure out how exactly it's done and copy the design... it would not put anyone off from developing their own product, that is just ridiculous.

I have absolutely no idea how you are coming up with the statement that it's not much more efficient than an undervolted 5970... lets say, being super, extra, double sugar on top generous, you can power a 5970 with 250w... that gives you ~750 MH/s.  The BFL unit does 800 MH/s at 85w... that is 1/3 the power for more MH/s.  In what world, exactly, is that "not much more efficient?"  In reality, you aren't getting a 5970 to 250w at 750 MH/s, just not happening, and this isn't even including the host system, which to be fair, the BFL unit basically requires as well.. however, you could load up a host system with more BFL units than you can 5970's, so it's more efficient in that area, too.

Quote
Undervolted 5970 can get 5MH/W

This statement is just laughable.  You aren't getting a 5970 to spit out 750 MH/s for 150w... that is just a joke, right?  Or are you, gasp, fudging your numbers to make them look legitimate, when they are, in reality, complete bullshit... just like you're accusing BFL of doing?  

Inaba, I am not making this a personal issue, and I think you would do well not to take it that way.  I don't think the car analogy is crap at all.....BFL came out with projected specs that were wildly inaccurate in regards to both hashing power as well as power consumption.  No one is asking that they be absolutely precise......but being off by 30% and 400% in regards to hashing and power consumption seems more than trivial.

I am not trying to get into an argument over this, I just think they way they went about releasing their product is completely lame.  

Anyway, I hope you get a bunch of them, and they work out well for you.  
donator
Activity: 1218
Merit: 1079
Gerald Davis


                                7970   |   BFL
 
Cost :                        550     |  599
Performance :                700   |  700
Warranty :                 3 years |  6 month
Resale value :                 50% | 0%




Wattage:          574W | 80W

http://www.hardocp.com/article/2012/02/08/gigabyte_radeon_hd_7970_oc_video_card_review/8

You missed that little tidbit Wink

Ok BFL uses less wattage (but more than any other FPGA) but 574W?  Really I mean lets as least be slightly realistic shall we?
hero member
Activity: 798
Merit: 1000


                                7970   |   BFL
 
Cost :                        550     |  599
Performance :                700   |  700
Warranty :                 3 years |  6 month
Resale value :                 50% | 0%




Wattage:          574W | 80W

http://www.hardocp.com/article/2012/02/08/gigabyte_radeon_hd_7970_oc_video_card_review/8

You missed that little tidbit Wink
donator
Activity: 1218
Merit: 1079
Gerald Davis
I have absolutely no idea how you are coming up with the statement that it's not much more efficient than an undervolted 5970... lets say, being super, extra, double sugar on top generous, you can power a 5970 with 250w... that gives you ~750 MH/s.  The BFL unit does 800 MH/s at 85w... that is 1/3 the power for more MH/s.  In what world, exactly, is that "not much more efficient?"

UNDERVOLTED.  250W is at stock.  Actually w/ memclock reduced it is less.  I run 3x5790s.  Full system 2.25 GH/s @ 870W AC (at the wall).  So the idea that it is impossible is just stupid.  

Quote
Undervolted 5970 can get 5MH/W

This statement is just laughable.  You aren't getting a 5970 to spit out 750 MH/s for 150w... that is just a joke, right?  Or are you, gasp, fudging your numbers to make them look legitimate, when they are, in reality, complete bullshit... just like you're accusing BFL of doing?  

Your laughable.    Please educate yourself.  

https://bitcointalksearch.org/topic/undervolting-a-5870-and-a-5770-to-achieve-better-mhj-performance-58912

More than one person has gotten >5MH/W with UNDERVOLTED 5870/5970.  Notice the prefix "under" in undervolted as in voltage less than stock.  Power consumption decreases by the SQUARE of the voltage reduction so a decent voltage reduction (say 15%) reduces power consumption by a significant amount (27% power reduction on a 15% voltage reduction).

Yes you can get >5MH/W on a 5870 or 5970.  You can do the same thing w/ 7970, (and likely a 7990 also).  Can you run it at stock clock?  No and nobody said you could.

Before your start insulting people how about your a) READ and b) MAKE SURE YOUR ARE INFORMED.

Note I never said it the single "barely beats" a 5970.  I don't undervolt.  My electrical rates are cheap I simply experimented w/ 5970 to see if I COULD when the time came and 5970 reached end of life and difficulty increases made mining at stock voltage uneconomical.  
hero member
Activity: 616
Merit: 500
Portland Bitcoin Group Organizer
I don't get it.  Why would a BFL unit have $0 resale value?

Its a very small market, so who would you sell a BFL Single to if the bitcoin experiment fails? your risk with GPUs are mitigated by the resale value that is fairly stable. $0 is a bit of hyperbole, but they make a good point.
legendary
Activity: 1260
Merit: 1000

I am sure they never thought it could deliver 1.05 (remember it wasn't an estimate but 1.05 GH 2 digits of significant value) at 19.8W.  That was simply a tactic to freeze out any sales of competitor products.



Which really cuts to the core of the issue, no?  It does the entire BTC community a diservice.  I imagine at least a couple enterprising individuals ceased developing their own products when they saw the "magical" specs for the bitforce.  Flash forward 5 months and loe and behold the actual product is really not that much more efficient than a undervolted 5970.  

It is too bad, as the more options for FPGA there are, the better it is for all of us as the proverbial "consumer".

I would love to announce a revolutionary vehicle I invented that gets 90 MPG (based on projections), at a price of $17,000.  I would be inundated with pre-orders and have gobs of cash to then develop my product.  Then when I actually delivered, and it only gets 63 MPG (not much better than a Prius), I would be thrown in jail for fraud.  



First of all, the car analogy is crap, since a Prius only gets good milage under a very limited set of circumstances.  Any modern diesel is far superior to the current crop of hybrids as far as fuel efficiency goes.  So comparing an FPGA to a car, you'd need to say XXX FPGA gets 250 MH/s only when you give it EXACTLY 121.742v, at 72F ambient temperature, with the fans spinning at 2311 RPM.  Anything other than that, and it doesn't get those specs.  It's ridiculous is what I'm saying.

I very seriously doubt that any enterprising individuals were/are the least bit dissuaded by another FPGA entering the market.  If anything, it would have spurned them to figure out how exactly it's done and copy the design... it would not put anyone off from developing their own product, that is just ridiculous.

I have absolutely no idea how you are coming up with the statement that it's not much more efficient than an undervolted 5970... lets say, being super, extra, double sugar on top generous, you can power a 5970 with 250w... that gives you ~750 MH/s.  The BFL unit does 800 MH/s at 85w... that is 1/3 the power for more MH/s.  In what world, exactly, is that "not much more efficient?"  In reality, you aren't getting a 5970 to 250w at 750 MH/s, just not happening, and this isn't even including the host system, which to be fair, the BFL unit basically requires as well.. however, you could load up a host system with more BFL units than you can 5970's, so it's more efficient in that area, too.

Quote
Undervolted 5970 can get 5MH/W

This statement is just laughable.  You aren't getting a 5970 to spit out 750 MH/s for 150w... that is just a joke, right?  Or are you, gasp, fudging your numbers to make them look legitimate, when they are, in reality, complete bullshit... just like you're accusing BFL of doing?  
hero member
Activity: 896
Merit: 1000
Seal Cub Clubbing Club
I don't get it.  Why would a BFL unit have $0 resale value?
hero member
Activity: 616
Merit: 500
Portland Bitcoin Group Organizer

I am sure they never thought it could deliver 1.05 (remember it wasn't an estimate but 1.05 GH 2 digits of significant value) at 19.8W.  That was simply a tactic to freeze out any sales of competitor products.



Which really cuts to the core of the issue, no?  It does the entire BTC community a diservice.  I imagine at least a couple enterprising individuals ceased developing their own products when they saw the "magical" specs for the bitforce.  Flash forward 5 months and loe and behold the actual product is really not that much more efficient than a undervolted 5970. 

It is too bad, as the more options for FPGA there are, the better it is for all of us as the proverbial "consumer".

I would love to announce a revolutionary vehicle I invented that gets 90 MPG (based on projections), at a price of $17,000.  I would be inundated with pre-orders and have gobs of cash to then develop my product.  Then when I actually delivered, and it only gets 63 MPG (not much better than a Prius), I would be thrown in jail for fraud. 



Point taken, but no one is holding a gun to your head... Are they?! Are you in danger?!
legendary
Activity: 2044
Merit: 1000

I am sure they never thought it could deliver 1.05 (remember it wasn't an estimate but 1.05 GH 2 digits of significant value) at 19.8W.  That was simply a tactic to freeze out any sales of competitor products.



Which really cuts to the core of the issue, no?  It does the entire BTC community a diservice.  I imagine at least a couple enterprising individuals ceased developing their own products when they saw the "magical" specs for the bitforce.  Flash forward 5 months and loe and behold the actual product is really not that much more efficient than a undervolted 5970. 

It is too bad, as the more options for FPGA there are, the better it is for all of us as the proverbial "consumer".

I would love to announce a revolutionary vehicle I invented that gets 90 MPG (based on projections), at a price of $17,000.  I would be inundated with pre-orders and have gobs of cash to then develop my product.  Then when I actually delivered, and it only gets 63 MPG (not much better than a Prius), I would be thrown in jail for fraud. 

donator
Activity: 1218
Merit: 1079
Gerald Davis
Thank you for these enlightening results.

Thus, it seems that the overclocked 7970 can get more than this unicorn BS.

I mean WTF Huh

                                7970   |   BFL
 
Cost :                        550     |  599
Performance :                700   |  700
Warranty :                 3 years |  6 month
Resale value :                 50% | 0%

Well to be fair 7970 is going to need a host system...

Quote
The ONLY advantage I see is the consumption of power but surely that is not worth it for some US guys that have 0.10 prices. Maybe this is for EU market but VAT and import duty kills that as well so it is LAME right now from the promised holy 1000 MHash/s, 20W, $500 figures ...

I am sure they never thought it could deliver 1.05 (remember it wasn't an estimate but 1.05 GH 2 digits of significant value) at 19.8W.  That was simply a tactic to freeze out any sales of competitor products.

Still if 700 MH/s is what is sustains and it pulls 80W at the wall we are looking at 8.75 MH/W.  I don't see a lot of value in higher energy prices.  Undervolted 5970 can get 5MH/W and best FPGA designs get 20 MH/W+.  So if you want cheap hardware and higher energy costs GPUs look good.  If you want expensive hardware and low energy costs other FPGA looks good.

At $1 per MH and <9MH/W the "single" is kinda in a no mans land in between.

Still to give them the benefit of the doubt it is early.  ztex boards were initially 180MH/s and now are 210 MH/s.
hero member
Activity: 798
Merit: 1000
Thanks Inaba.  I'm glad things are getting a little more productive in this thread now.
legendary
Activity: 1260
Merit: 1000
Quote
@Inaba, have you had a chance to try the ufasoft miner to see if the unit behaves the same? It might provide clues on whether this is a cgminer-only issue, or if it is a fundamental issue with the unit.

Not yet, I will try to get Ufasoft running tonight.  Sonny said the code for Ufasoft is a lot farther along for the BFL Single units than the CGMiner code, so we'll see how that works out.

Quote
So, there is some kind of thermal monitoring that throttles the miner when it gets too hot?

Yes, if the unit overheats, it throttles itself down until the temp is safe.  Sonny said he would reflash my unit to give it a wider tolerance if needed as well, apparently the demo I have refused to give back is restricted on the ranges it will start throttling at.  I'm not sure what/how that code is different from production code/firmware (or even if it is?) .  I'm still traveling, so it's hard to keep up on top of things.  I will be back in town tomorrow.
legendary
Activity: 4592
Merit: 1851
Linux since 1997 RedHat 4
...
@kano, you had mentioned earlier that you were doing some work on icarus support in cgminer; have you seen anything in the code that might suggest a reason for the observed throttling behavior?
...
Nope, got it mostly all cleaned up and working last night.
(well I just found the problems and told him the fixes which he implement with other things he worked out)
But in this case, it was reporting random strange values and hashing at almost half the U: it should, but no they weren't related to the bitforce code that the icarus code is based off (fortunately for us, unfortunately for you)

Also, cgminer has gone through a lot of commits over the last couple of weeks while ckolivas works on getting his 7970 hashing as fast as possible ... including getting rid of an outstanding problem for the last 6 months (more ATI crap finally worked around) so maybe there's something in there? (2.2.7 was just released)
hero member
Activity: 518
Merit: 500
I just ran some numbers for Inaba's Single @ https://eclipsemc.com/mw.php?key=196e69b1afe10927460b22a8de1c0d:

3:23pm CST: 65612 shares
12:08am CST: 70949 shares (8h45m later)

So the average hashrate during that 8h45m period was:

70949-65612 = 5337 shares in 31500 seconds (8h45m)
5337 shares * 2^32 hashes/share is 2.29e13 hashes
2.29e13 hashes / 31500s = 727Mhps

Not too bad, but still 13% lower than the advertised specs of 832Mhps.  Sad

Here's another numbers run, for a 15 hour period to 15:06 CST today. Inaba mentioned there are some issues with his unit that are causing throttling.

@Inaba, have you had a chance to try the ufasoft miner to see if the unit behaves the same? It might provide clues on whether this is a cgminer-only issue, or if it is a fundamental issue with the unit.

@kano, you had mentioned earlier that you were doing some work on icarus support in cgminer; have you seen anything in the code that might suggest a reason for the observed throttling behavior?

start: 00:08CST, shares 70949
end: 15:06CST, shares 79630

So we have:
15:06 - 00:08 = 14h58m = 53880s
79630 - 70949 = 8681 shares
8681 shares * 2^32 hashes per share = 3.73e13 hashes
3.73e13 hashes / 53880s = 692Mhps


Thank you for these enlightening results.

Thus, it seems that the overclocked 7970 can get more than this unicorn BS.

I mean WTF Huh

                                7970   |   BFL
 
Cost :                        550     |  599
Performance :                700   |  700
Warranty :                 3 years |  6 month
Resale value :                 50% | 0%

The ONLY advantage I see is the consumption of power but surely that is not worth it for some US guys that have 0.10 prices. Maybe this is for EU market but VAT and import duty kills that as well so it is LAME right now from the promised holy 1000 MHash/s, 20W, $500 figures ...

BFL deserves a medal for fail of the year.  

Thermal throttling due to bad design FTW !
legendary
Activity: 922
Merit: 1003
I just ran some numbers for Inaba's Single @ https://eclipsemc.com/mw.php?key=196e69b1afe10927460b22a8de1c0d:

3:23pm CST: 65612 shares
12:08am CST: 70949 shares (8h45m later)

So the average hashrate during that 8h45m period was:

70949-65612 = 5337 shares in 31500 seconds (8h45m)
5337 shares * 2^32 hashes/share is 2.29e13 hashes
2.29e13 hashes / 31500s = 727Mhps

Not too bad, but still 13% lower than the advertised specs of 832Mhps.  Sad

Here's another numbers run, for a 15 hour period to 15:06 CST today. Inaba mentioned there are some issues with his unit that are causing throttling.

@Inaba, have you had a chance to try the ufasoft miner to see if the unit behaves the same? It might provide clues on whether this is a cgminer-only issue, or if it is a fundamental issue with the unit.

@kano, you had mentioned earlier that you were doing some work on icarus support in cgminer; have you seen anything in the code that might suggest a reason for the observed throttling behavior?

start: 00:08CST, shares 70949
end: 15:06CST, shares 79630

So we have:
15:06 - 00:08 = 14h58m = 53880s
79630 - 70949 = 8681 shares
8681 shares * 2^32 hashes per share = 3.73e13 hashes
3.73e13 hashes / 53880s = 692Mhps
sr. member
Activity: 446
Merit: 250
There are several issues with the unit I have at the moment.  The windows update, there appears to be a possible bug in cgminer that is sticking the throttled hashrate and not increasing speed once it throttles, and where I have it located might be causing some throttling due to heat.  In retrospect, I probably should have placed it in a slightly different spot, but I was in a rush and it's getting some backwash from one of my PSU exhaust fans. 

So, there is some kind of thermal monitoring that throttles the miner when it gets too hot?
legendary
Activity: 1260
Merit: 1000
There are several issues with the unit I have at the moment.  The windows update, there appears to be a possible bug in cgminer that is sticking the throttled hashrate and not increasing speed once it throttles, and where I have it located might be causing some throttling due to heat.  In retrospect, I probably should have placed it in a slightly different spot, but I was in a rush and it's getting some backwash from one of my PSU exhaust fans. 
Pages:
Jump to: