HashFast launches sales of the Baby Jet

[gnar1ta$]

We did it!! Shipped the first Baby Jets and Sierras today.

Then why did my email not have a tracking number but some big long refund or wait policy instead?

[miaviator]

It's been a while since I looked on this forum. I'm really disappointed to read about what's happened with Hashfast as it looks as though a lot of trusting customers may end up losing a fair portion of their hard earned money. That's inexcusable as it would seem that HF got more than enough funding to develop a 28nm chip and build their customers' systems. How much they took out of the company in expenses and salaries is anyone's guess.

It's also going to make life very difficult for any new suppliers that  might come along, and that leaves a bigger problem.

I've always felt that mining rigs are grossly overpriced, even at the current lowest offering of $3 per gigahash. It's simply not competitive with what the 'invisible' mining corporations can build capacity for, ie less than $1.50/GH. Think I'm kidding? Well, let me give you some insight....

Recently a previous colleague of mine got in contact. He works for a group of Venture Capitalists in the UAE and had been approached by some of my fellow countrymen with a business plan for the development of a 'consumer' electronic system - he wouldn't tell me what the application was as it was all under Non Disclosure agreements, but basically he asked me to check out their technical plan -edited, of course - and their costings. When I went over the facts and figures it became obvious that the plan centered on a custom silicon device; from the power specs, huge heatsinks and unusually low clock speeds it spelt out SHA256, or 'transaction engine' as my contact called it.

To cut to the chase, it was a very comprehensive and well thought out technical plan with realistic timescales, although I thought some of the costings were a bit on the pessimistic side. It's a real pity the plan I got was so heavily edited as it referred to 'commercial **** companies dominating the market with proprietary systems' and had several graphs and tables totally blacked out. It did, however, refer to a 'Network ##### Rate' and how much they expected it to grow in 2014 - to over 180,000####### by Dec 2014, no less. Whoever these 'companies' were, the guys that wrote the plan seemed to really disapprove of them.

But here's the killer - in their plan a basic system using one 'transaction engine' would retail at 'around' $300 with a manufacturing price well below $200 - I can't give you the real figure - including one-off costs! This system had a 'rated capacity' of 200 billion ###### per second, I'm pretty sure I don't have to spell the rest out.

My UAE contact said his principals were very impressed with the plan, but only if the systems were used in-house, as it were, due to their  potential ROI - I never got to see that part. They most certainly didn't want them sold to the public under any circumstances. He hinted that a very attractive offer was made to the group, but was politely declined. He wouldn't tell me who the group were or how to contact them.

It's a very interesting thought - a 200GH miner for $300?? Anyone heard any hint of this?

Anyone who has heard or spoke of it has violated an NDA and risks a lot of fines and possible criminal prosecution.

Sounds legit to me.

[brontosaurus]

It's been a while since I looked on this forum. I'm really disappointed to read about what's happened with Hashfast as it looks as though a lot of trusting customers may end up losing a fair portion of their hard earned money. That's inexcusable as it would seem that HF got more than enough funding to develop a 28nm chip and build their customers' systems. How much they took out of the company in expenses and salaries is anyone's guess.

It's also going to make life very difficult for any new suppliers that  might come along, and that leaves a bigger problem.

I've always felt that mining rigs are grossly overpriced, even at the current lowest offering of $3 per gigahash. It's simply not competitive with what the 'invisible' mining corporations can build capacity for, ie less than $1.50/GH. Think I'm kidding? Well, let me give you some insight....

Recently a previous colleague of mine got in contact. He works for a group of Venture Capitalists in the UAE and had been approached by some of my fellow countrymen with a business plan for the development of a 'consumer' electronic system - he wouldn't tell me what the application was as it was all under Non Disclosure agreements, but basically he asked me to check out their technical plan -edited, of course - and their costings. When I went over the facts and figures it became obvious that the plan centered on a custom silicon device; from the power specs, huge heatsinks and unusually low clock speeds it spelt out SHA256, or 'transaction engine' as my contact called it.

To cut to the chase, it was a very comprehensive and well thought out technical plan with realistic timescales, although I thought some of the costings were a bit on the pessimistic side. It's a real pity the plan I got was so heavily edited as it referred to 'commercial **** companies dominating the market with proprietary systems' and had several graphs and tables totally blacked out. It did, however, refer to a 'Network ##### Rate' and how much they expected it to grow in 2014 - to over 180,000####### by Dec 2014, no less. Whoever these 'companies' were, the guys that wrote the plan seemed to really disapprove of them.

But here's the killer - in their plan a basic system using one 'transaction engine' would retail at 'around' $300 with a manufacturing price well below $200 - I can't give you the real figure - including one-off costs! This system had a 'rated capacity' of 200 billion ###### per second, I'm pretty sure I don't have to spell the rest out.

My UAE contact said his principals were very impressed with the plan, but only if the systems were used in-house, as it were, due to their  potential ROI - I never got to see that part. They most certainly didn't want them sold to the public under any circumstances. He hinted that a very attractive offer was made to the group, but was politely declined. He wouldn't tell me who the group were or how to contact them.

It's a very interesting thought - a 200GH miner for $300?? Anyone heard any hint of this?

Anyone who has heard or spoke of it has violated an NDA and risks a lot of fines and possible criminal prosecution.

Sounds legit to me.

Just to say there's no NDA been violated - the NDA in question related to a specific design implementation and the detailed commercial figures (which I never saw). My contact was very careful to screen his information - if I had never heard of Bitcoin or sha256 I would have assumed this was some kind of encryption device.

But thanks for your comment, you're absolutely right that NDA violation can result in heavy civil penalties.

[brontosaurus]

‏@HashFast
https://twitter.com/HashFast/status/418107074979971072

A big thank you to Amy for camping out in Montreal for the last week!



We did it!! Shipped the first Baby Jets and Sierras today.

I'm sorry if I sound a little cynical about all this, but Hashfast's end-of-2013 announcements seem a bit odd, to put it mildly.

Firstly, they claim chip performance of 'up to' 664GH/sec. Real engineers don't do 'up to', they quote maximum and minimum and specify under
what conditions each is valid.

Secondly, they say this performance test was " conducted by running a single GN die directly from bechtop power supplies, as opposed to powering it through the module". I'm assuming by die they mean one of the 4 functional blocks. Why not use the actual system power supplies? Something wrong with them?

Thirdly, because "This approach allows us to obtain data about what the ASIC itself can do, without having to make subjective estimates regarding the efficiency of the power supply on the module. However, doing things this way also has it’s own set of disadvantages.For example, the reason we are “only” able to announce a top speed of 664 Ghash per chip is purely because that’s the point at which we ran out of power to put through the chip. " then that means their chip, with all four cores running will use 664GH x 0.67w/GH = 442 watts, all from a  silicon ares of 664/2 (their figures of 2GH/mm2 of silicon), ie 332 mm2. This, frankly, is impossible. You would need a heatsink with a NEGATIVE thermal coefficient to keep the die junction temperature below 75 degrees C. As far as I'm aware, none exist.

Can't they afford a decent PSU?

Anyone else care to add their observations?

[cedivad]

That they changed the specs of the chip overnight. From 735 to 664 GH/s

https://bitcointalk.org/index.php?topic=262052.msg4247859#msg4247859

[ninjarobot]

That they changed the specs of the chip overnight. From 735 to 664 GH/s

https://bitcointalk.org/index.php?topic=262052.msg4247859#msg4247859

Yes, looks like HF made some stealth-edits to those specs already:

Original: http://web.archive.org/web/20140101023059/http://hashfast.com/were-shipping-2013/

The Golden Nonce is:

The fastest Bitcoin mining chip in the world today — up to 735 Ghash/s per chip!
The most energy efficient mining chip in the world today — 0.59 watts per Ghash when run for maximum efficiency
The most silicon-efficient chip in the world — producing up to an astonishing 2.27 Ghash out of every square millimeter of silicon!
We couldn’t be prouder of these results – and can’t wait to see what the community can do with it.

Edited: http://hashfast.com/were-shipping-2013/

The Golden Nonce is:

The fastest Bitcoin mining chip in the world today — up to an unprecedented 664 Ghash/s per chip!
The most energy-efficient — down to 0.67 watts per Ghash when run for maximum efficiency!
And the most silicon-efficient — Each square millimeter of silicon on the GN chip produces an astonishing 2+ Ghash!
We couldn’t be prouder of these results – and can’t wait to see what the community can do with it.

[Puppet]

664GH x 0.67w/GH = 442 watts, all from a  silicon ares of 664/2 (their figures of 2GH/mm2 of silicon), ie 332 mm2. This, frankly, is impossible. You would need a heatsink with a NEGATIVE thermal coefficient to keep the die junction temperature below 75 degrees C.

Not sure what makes you say that. 442W/332mm²= ~1.3W/mm². Thats not enormous. A typical highend AMD or Intel desktop CPU has comparable thermal density, and in fact, for an x86 chip most of that power is concentrated in a few hotspots that are just a few mm². Not that 75+C should be a problem anyway.

I do find it funny they claim measurements of their PSU would somehow be "subjective".

[brontosaurus]

That they changed the specs of the chip overnight. From 735 to 664 GH/s

https://bitcointalk.org/index.php?topic=262052.msg4247859#msg4247859

Yes, looks like HF made some stealth-edits to those specs already:

Original: http://web.archive.org/web/20140101023059/http://hashfast.com/were-shipping-2013/

The Golden Nonce is:

The fastest Bitcoin mining chip in the world today — up to 735 Ghash/s per chip!
The most energy efficient mining chip in the world today — 0.59 watts per Ghash when run for maximum efficiency
The most silicon-efficient chip in the world — producing up to an astonishing 2.27 Ghash out of every square millimeter of silicon!
We couldn’t be prouder of these results – and can’t wait to see what the community can do with it.

Edited: http://hashfast.com/were-shipping-2013/

The Golden Nonce is:

The fastest Bitcoin mining chip in the world today — up to an unprecedented 664 Ghash/s per chip!
The most energy-efficient — down to 0.67 watts per Ghash when run for maximum efficiency!
And the most silicon-efficient — Each square millimeter of silicon on the GN chip produces an astonishing 2+ Ghash!
We couldn’t be prouder of these results – and can’t wait to see what the community can do with it.

Well, lets be honest - if you can beat the laws of thermodynamics then specs become meaningless. So why bother tying yourself down?

If I was a Hashfast customer, I'd be more than a little annoyed that the public seem to get 'new' information at the same time I do. If I've paid over my cash, I'd like to get the info first.

[brontosaurus]

664GH x 0.67w/GH = 442 watts, all from a  silicon ares of 664/2 (their figures of 2GH/mm2 of silicon), ie 332 mm2. This, frankly, is impossible. You would need a heatsink with a NEGATIVE thermal coefficient to keep the die junction temperature below 75 degrees C.

Not sure what makes you say that. 442W/332mm²= ~1.3W/mm². Thats not enormous. A typical highend AMD or Intel desktop CPU has comparable thermal density, and in fact, for an x86 chip most of that power is concentrated in a few hotspots that are just a few mm². Not that 75+C should be a problem anyway.

I do find it funny they claim measurements of their PSU would somehow be "subjective".

Well, it's all to do with the thermal impedance of the chip (theta jc). That determines how many watts can be transferred from the die junction to it's 'case', or in this case the back of the die, per degree centigrade rise of the junction temperature. An Intel Core i7-970 cpu has a die area of 239 mm2 and a design power dissipation of 130W, approximately 0.54w/mm2 averaged across the whole die area. Don't know about you, but I hold Intel's engineering in very high esteem, they know how to make ultra high volume consumer silicon products reliable. So when a 'nobody' suggests than they can get 2.5x better thermal performance per square mm, I'm more than a little concerned. 1.25x, maybe. Just.

Of course chips can exceed junction temperatures of 75 degrees C, power devices go up to 150 routinely but they're built with processes designed to operate at this level.

[cedivad]

Thanks for sharing, brontosaurus. It's really interesting.

[Puppet]

An Intel Core i7-970 cpu has a die area of 239 mm2 and a design power dissipation of 130W, approximately 0.54w/mm2 averaged across the whole die area.  

Sure, but have a look at thermal image of a cpu:



Almost all of that 130W is consumed in an area that is is maybe 10mm². A bitcoin asic would be much more uniform. Silicon is a decent thermal conductor, but not as good as the aluminium heatspreader.

[brontosaurus]

Thanks for sharing, brontosaurus. It's really interesting.

Very kind of you to say so, thanks.

I have no personal interest in Hashfast or any other supplier of rigs, but I really don't like the way that some companies assume that their audience will swallow any old technical rubbish they serve up. I appreciate that a lot of people don't know a lot about the detailed technology and that's why forums are good for everyone. Plus there is a lot of accumulated knowledge out there - nobody knows it all and we can all learn from others experience, the rig companies included.

Companies wanting customer's money on trust have an implicit duty to provide them with proper specifications - not one off measurements or guesses. By all means give them estimated performance but base it on proper maths and technical parameters and specify HOW you have arrived at your data.


 

[CYPER]

It's a very interesting thought - a 200GH miner for $300?? Anyone heard any hint of this?

1.5$ per GH/s is to be expected from all ASIC miners to be made available in about 6 months time.
I would not be very surprised if the KnCMiner Neptune manages 6000GH/s, which equals to $1.66 per GH/s

I know it is highly optimistic, but then again the Jupiter was advertised as 250GH/s when announced and it is currently hashing at 675GH/s, which is a 170% increase.
A 3000GH/s miner with the same percentage increase would be 8100GH/s, so 6000 sounds possible.

[driver]

What I would like to know is what type video output does it have and is it male or female plug ? You just never know what will come on a machine when it is not listed in the specs .
 I assume it is vga with female  but who knows .
 Also assume it has it's own ethernet card  . Wonder what speed that is .  They need to put more info on their website about their products .

its a Raspi (supposedly). Last i checked they use  hdmi

thanks . good thing when I was buying cables I ordered a hdmi just because it was on sale . Now if I can just get it before it is outdated and only good as a paper weight .

[driver]

Now since the only payment option is in BTC Will I get the same ammount of BTC back should you fail to deliver by December 31st?

Orders are taken in BTC, in the unlikely event we get to refunds they will be given in BTC.

HashFast has now announced they'll be missing their "guarantee date", and are attempting to renig on the above. Now customers must choose to lock in an 86% loss or lose all potential of refund:

Batch 1 customers who would like to initiate a voluntary refund request, can do so by sending the following information to: refunds@hashfast.com. Please note this is a request only and not a guarantee of refund eligibility. As orders were priced in USD, refunds will be issued in the equivalent amount of USD. If a refund is to be paid in BTC, the USD to BTC market exchange rate on the date of refund will be used to calculate the amount of BTC to be refunded.

This cancellation and refund is Buyer's sole and exclusive remedy for HashFast failing to deliver by the December 31, 2013 guaranteed delivery date, and Buyer must cancel the order by January 15, 2014 to avail itself of this remedy.

hashfast must return the same amount of BTC in value not in coins. you give them 2.5 coins worth 2500.00 and the value on BTC drops then they must give you back more than 2.5 coins  to compensate . At least in theory . I wonder if there is any exception in the laws to provide them to only give you back the exact amount of coins you gave them or in the case of cash the same amount of bills you gave them even if the dollar lost value . 

[driver]

‏@HashFast
We did it!! Shipped the first Baby Jets and Sierras today.

It's too bad your terms of sale specify refunds are available for all Batch 1 customers who don't receive their miners by... oh look... today!

Good luck on your illegal scammy attempts to refund people a small fraction of the BTC they paid you, too. I sure hope you enjoy getting sued.

One thought I had is that since BTC is not a " real currency " that they may only have to give back as many as you gave them and not the amount to make you whole . Like if you gave them 5 apples they would only have to give you back 5 even if apples price dropped way down . but they would have to give you back apples of the same quality i.e. not rotten so maybe there is some room for legal standing .
 My guess is they are just trying to make you think you will be given back less than you paid so you will not ask for a return and that buys them more time to deliver .

[Gyrsur]

‏@HashFast
We did it!! Shipped the first Baby Jets and Sierras today.

It's too bad your terms of sale specify refunds are available for all Batch 1 customers who don't receive their miners by... oh look... today!

Good luck on your illegal scammy attempts to refund people a small fraction of the BTC they paid you, too. I sure hope you enjoy getting sued.

One thought I had is that since BTC is not a " real currency " that they may only have to give back as many as you gave them and not the amount to make you whole . Like if you gave them 5 apples they would only have to give you back 5 even if apples price dropped way down . but they would have to give you back apples of the same quality i.e. not rotten so maybe there is some room for legal standing .
 My guess is they are just trying to make you think you will be given back less than you paid so you will not ask for a return and that buys them more time to deliver .

exactly. they have to refund the payment of xxx BTC as it is mentioned in the TOS.

[aneutronic]

Thanks for sharing, brontosaurus. It's really interesting.

Very kind of you to say so, thanks.

I have no personal interest in Hashfast or any other supplier of rigs, but I really don't like the way that some companies assume that their audience will swallow any old technical rubbish they serve up. I appreciate that a lot of people don't know a lot about the detailed technology and that's why forums are good for everyone. Plus there is a lot of accumulated knowledge out there - nobody knows it all and we can all learn from others experience, the rig companies included.

Companies wanting customer's money on trust have an implicit duty to provide them with proper specifications - not one off measurements or guesses. By all means give them estimated performance but base it on proper maths and technical parameters and specify HOW you have arrived at your data.

+1

Thanks for sharing your knowledge.   

[aerobatic]

I'm sorry if I sound a little cynical about all this, but Hashfast's end-of-2013 announcements seem a bit odd, to put it mildly.

Firstly, they claim chip performance of 'up to' 664GH/sec. Real engineers don't do 'up to', they quote maximum and minimum and specify under
what conditions each is valid.

Secondly, they say this performance test was " conducted by running a single GN die directly from bechtop power supplies, as opposed to powering it through the module". I'm assuming by die they mean one of the 4 functional blocks. Why not use the actual system power supplies? Something wrong with them?

Thirdly, because "This approach allows us to obtain data about what the ASIC itself can do, without having to make subjective estimates regarding the efficiency of the power supply on the module. However, doing things this way also has it’s own set of disadvantages.For example, the reason we are “only” able to announce a top speed of 664 Ghash per chip is purely because that’s the point at which we ran out of power to put through the chip. " then that means their chip, with all four cores running will use 664GH x 0.67w/GH = 442 watts, all from a  silicon ares of 664/2 (their figures of 2GH/mm2 of silicon), ie 332 mm2. This, frankly, is impossible. You would need a heatsink with a NEGATIVE thermal coefficient to keep the die junction temperature below 75 degrees C. As far as I'm aware, none exist.

Can't they afford a decent PSU?

Anyone else care to add their observations?

both your assumptions could be argued arent safe assumptions...

first, since hf were testing just one die in isolation (presumably with the others turned off) then they were specifically benchmarking one die and it should be treated more as academic interest than a marketing statistic.  Its an interesting and exciting statistic but it ignores the reality of the power supply, the thermal characteristics of operating 4 dies concurrently in the same package, and presumably also avoiding the thermal limits of the package & cooling system that would be a different scenario with all four dies turned on in one package.

Its a very exciting marketing statistic, but by testing one die in isolation and then presumably multiplying the result by four... does that still count as a legitimate benchmark for what the system is capable of?   I say YES, provided all four dies, when run together can also achieve the same number... but conversely if that cant be achieved with all four dies, then testing one die in isolation and multiplying the result by four could be argued that its an artificial performance metric of academic interest only.  Much the same as if you have a 4 core intel cpu and turn off 3 of the cores, that the remaining single core also will run much faster than when all 4 cores are turned on together.

Of course, they could redesign the substrates and package and just put one die in each package... and that would allow them a board re-design.. and then have 4 chips in a baby jet instead of one big one...  (which follows the Bitmine argument that using multiple smaller chips may achieve a better outcome than using fewer bigger chips).

is it valid to measure the performance of just one die, and then multiply the result by 4 to give you what the total of 4 dies wouldve couldve done (.. in a perfect world where they had infinite power and cooling available to them)... when those 4 dies, in the same package, when run together, may not be able to achieve the same result?   And, i should stress, if it CAN.. thatd be awesome and extremely impressive...!

then there's the issue of the two stats.  the two data points.  The 664GH performance claim, And the 0.67w/GH power consumption are two separate stats.   Hashfast didnt link them together.   You did (incorrectly assuming they were done using the same conditions).   HF didnt claim that when they were running at 664GH they were ALSO only consuming 0.67w/gh.. though thatd be simply fantastic if true!  Those two are independent stats and its safer to assume that the 0.67w/gh ultra low power achievement was probably achieved when running at a lower voltage setting than when when the benchmark was showing a die running at the 664gh (/4) equivalent performance achievement.

also, as they also identified... the tests were achieved when running off a bench power supply, without the inefficiencies of the dc/dc converters nor the limits of atx power supplies... so its an isolated measure of performance.  Its testing the die on its own, but not testing the dies, in situ in the system as it will be supplied.   we of course would love to know what the die will do on its own... but the more important statistic for us as customers is what the chip will do, when its on a production board using production power supplies and production cooling... and though its exciting to hear what it can do (with the wind behind it) in the lab, connected to a bench power supply, isolated from the other dies.  thats a special case scenario that isnt necessarily representative of what will be achieved in the real world use case.

heck, if they want to quote even higher performance numbers (quite legitimately) they should be pouring liquid nitrogen down a tube directly onto the die... for the ultimate in cooling - the way that pc overclockers do it.  But you have to bear in mind that Intel never makes claims as to the performance that the overclocking teams hit, as theyre doing it using extreme methods that arent available to regular customers.

[brontosaurus]

I'm sorry if I sound a little cynical about all this, but Hashfast's end-of-2013 announcements seem a bit odd, to put it mildly.

Firstly, they claim chip performance of 'up to' 664GH/sec. Real engineers don't do 'up to', they quote maximum and minimum and specify under
what conditions each is valid.

Secondly, they say this performance test was " conducted by running a single GN die directly from bechtop power supplies, as opposed to powering it through the module". I'm assuming by die they mean one of the 4 functional blocks. Why not use the actual system power supplies? Something wrong with them?

Thirdly, because "This approach allows us to obtain data about what the ASIC itself can do, without having to make subjective estimates regarding the efficiency of the power supply on the module. However, doing things this way also has it’s own set of disadvantages.For example, the reason we are “only” able to announce a top speed of 664 Ghash per chip is purely because that’s the point at which we ran out of power to put through the chip. " then that means their chip, with all four cores running will use 664GH x 0.67w/GH = 442 watts, all from a  silicon ares of 664/2 (their figures of 2GH/mm2 of silicon), ie 332 mm2. This, frankly, is impossible. You would need a heatsink with a NEGATIVE thermal coefficient to keep the die junction temperature below 75 degrees C. As far as I'm aware, none exist.

Can't they afford a decent PSU?

Anyone else care to add their observations?

both your assumptions could be argued arent safe assumptions...

first, since hf were testing just one die in isolation (presumably with the others turned off) then they were specifically benchmarking one die and it should be treated more as academic interest than a marketing statistic.  Its an interesting and exciting statistic but it ignores the reality of the power supply, the thermal characteristics of operating 4 dies concurrently in the same package, and presumably also avoiding the thermal limits of the package & cooling system that would be a different scenario with all four dies turned on in one package.

Its a very exciting marketing statistic, but by testing one die in isolation and then presumably multiplying the result by four... does that still count as a legitimate benchmark for what the system is capable of?   I say YES, provided all four dies, when run together can also achieve the same number... but conversely if that cant be achieved with all four dies, then testing one die in isolation and multiplying the result by four could be argued that its an artificial performance metric of academic interest only.  Much the same as if you have a 4 core intel cpu and turn off 3 of the cores, that the remaining single core also will run much faster than when all 4 cores are turned on together.

Of course, they could redesign the substrates and package and just put one die in each package... and that would allow them a board re-design.. and then have 4 chips in a baby jet instead of one big one...  (which follows the Bitmine argument that using multiple smaller chips may achieve a better outcome than using fewer bigger chips).

is it valid to measure the performance of just one die, and then multiply the result by 4 to give you what the total of 4 dies wouldve couldve done (.. in a perfect world where they had infinite power and cooling available to them)... when those 4 dies, in the same package, when run together, may not be able to achieve the same result?   And, i should stress, if it CAN.. thatd be awesome and extremely impressive...!

then there's the issue of the two stats.  the two data points.  The 664GH performance claim, And the 0.67w/GH power consumption are two separate stats.   Hashfast didnt link them together.   You did (incorrectly assuming they were done using the same conditions).   HF didnt claim that when they were running at 664GH they were ALSO only consuming 0.67w/gh.. though thatd be simply fantastic if true!  Those two are independent stats and its safer to assume that the 0.67w/gh ultra low power achievement was probably achieved when running at a lower voltage setting than when when the benchmark was showing a die running at the 664gh (/4) equivalent performance achievement.

also, as they also identified... the tests were achieved when running off a bench power supply, without the inefficiencies of the dc/dc converters nor the limits of atx power supplies... so its an isolated measure of performance.  Its testing the die on its own, but not testing the dies, in situ in the system as it will be supplied.   we of course would love to know what the die will do on its own... but the more important statistic for us as customers is what the chip will do, when its on a production board using production power supplies and production cooling... and though its exciting to hear what it can do (with the wind behind it) in the lab, connected to a bench power supply, isolated from the other dies.  thats a special case scenario that isnt necessarily representative of what will be achieved in the real world use case.

heck, if they want to quote even higher performance numbers (quite legitimately) they should be pouring liquid nitrogen down a tube directly onto the die... for the ultimate in cooling - the way that pc overclockers do it.  But you have to bear in mind that Intel never makes claims as to the performance that the overclocking teams hit, as theyre doing it using extreme methods that arent available to regular customers.

Hi Aerobatic, good of you to take the time to reply. Like I said before, I have no interest in Hashfast other than that I'd like to see them fulfill their obligations to their paying (or rather paid up) customers, and I'm sure said customers would agree. So to me it's odd to trumpet what they 'may' be able to achieve rather than supplying tracking numbers for what they have dispatched, especially when those number don't add up - from the engineering point of view.

I'd like to address your points in more detail, but it's 9 pm in the UK just now and I'm just about to watch Sherlock Holmes (the new series) on BBC. Don't know if you get it where you live, but it's well worth a watch on the iPlayer. So I'll reply tomorrow, but one thing I can confirm to you is that the watts/GH figure remains constant no matter what clock speed is used. It's generated from the equation:

watts/Hash = Ng * Pg * F / (Nc * F) where:

Ng = number of gates switching per clock cycle - a design constant which depends upon the pipeline stage architecture
Pg = average switching power per gate per Mhz - a silicon process constant; about 0.6 nanowatts per Mhz for most LP 28nm processes
F = clock frequency (variable)
Nc = number of cores (pipelines) in the device; each core produces one result (hash) out of the pipeline every clock cycle, the pipeline latency is
       ignored as it's irrelevant in practice. So hashes = NC * F.

Or to simplify, P = Ng*Pg/Nc.  F cancels out, ie frequency is irrelevant. Static device power is also ignored here as it's relatively low in comparison.

If your pipeline design is very efficient then P goes down. Inefficient design = up.

Hope this helps.