Author

Topic: Alpha Technology Litecoin (Scrypt) ASIC Miner Order Batch 1 Now! - page 237. (Read 529093 times)

copper member
Activity: 2898
Merit: 1465
Clueless!
Searing I did the same deal. Used paypal and amex.

I can call amex and cancel the repayment if I want to for 60 days ?

I will play it safe and do it before 45 days

amex said 60 days

some bfl customers got their stuff cancelled from BFL long after 60 days I heard for non-shipping but that was probably because BFL
never replied to Amex ..ie blew them off from what I heard

anyway so what I was told ...by amex rep the day I ordered said 60days

will probably again just do the under 45days if pressed to cancel

still tempted just to keep the dang thing if they can prove they will make one ...even if BFL late and the fall

if such is the case plan on selling some junk in attic to pay for it...hard part here is are they like hashfast all talk or will someone get a unit or two this summer?

but still will probably need more info that I actually WILL get a product of very likely before I keep my order in

they may be just wonderful fellows..but hard to pick the crooks from the champs in all this pre-order stuff the paypal option bodes well on their behalf but gonna have a hard hill to climb to show the goods....BFL and Hashfast and others have scar'd a lot of us

anyway only thing I'm sure of is anything less then 45 days I'm probably ok...,not 100 percent thou they have that refund disclaimer after 30 days not sure how that applies to paypal

anyway show us some stuff that you have the chops alpha tech ..we are or I am a skittish fickle potential customer after last year

Searing
member
Activity: 76
Merit: 10
Searing I did the same deal. Used paypal and amex.

I can call amex and cancel the repayment if I want to for 60 days ?
copper member
Activity: 2898
Merit: 1465
Clueless!
There are just too many facts that are too suspicious. I too was interested in getting the 50 MH/s one. But I can't afford to lose $ 9,000. Also who knows what devices will hit the market by summer/fall of 2014? Better not risk it.

I agree I have ordered the 5mh one..but via paypal and credit card(amex) have 45 days on their list to see how this all shakes out  ...60 days min with amex card probably more for non-shipment as an option to cancel

so yeah...I agree they have like 45 days to show some REAL progress or I and I'm sure many others will punt

just making this point in case anyone from alpha tech is reading this...(doubtful)

Searing
full member
Activity: 168
Merit: 100
There are just too many facts that are too suspicious. I too was interested in getting the 50 MH/s one. But I can't afford to lose $ 9,000. Also who knows what devices will hit the market by summer/fall of 2014? Better not risk it.
sr. member
Activity: 384
Merit: 250
Yeh, I highly doubt this company can produce the hash rates they are saying and for that cost. Increasing the clock speeds will have little to no effect over their "prototype". And their prototype seems to be a copy of https://github.com/kramble/FPGA-Litecoin-Miner. They state that they are using the units ram which means they have not found a way to efficiently store the pseudo-random data on die / chip (in this case in logic gates on FPGA).

I don't think they are using my code. They clearly had their own prototype code back in July (link) and wrote a paper (link) which demonstrated quite poor performance on FPGA. They have improved their implementation since then, but are still persevering with the simple TMTO (LOOKUP_GAP) ratio of 1 which does not make good use of the most critical RAM resource. This may change when they get professional ASIC design engineers involved with the project (they certainly now have the funding from pre-orders to pay for the design work, though the NRE for the mask set will still need to be raised, 28nm is a very expensive process to work with).
hero member
Activity: 583
Merit: 500
I wonder how long the grace period really is. I put in an order, but did not send payment yet. Maybe I'll just keep my bitcoins.


Regardless, never pay with bitcoins! Learn from the HashFast debacle. Those poor saps paid like 45 Bitcoins for the rigs they still haven't received.

Hodl your coins man!
legendary
Activity: 3416
Merit: 1912
The Concierge of Crypto
I wonder how long the grace period really is. I put in an order, but did not send payment yet. Maybe I'll just keep my bitcoins.
sr. member
Activity: 448
Merit: 250
Yeh, coins like YaCoin use sCrypt-Jane which is a slighly modified version of sCrypt using different mix functions. sCrypt-Jane is cool because it dynamically changes the memory requirements of the algo therefore it will become more CPU friendly and less GPU friendly as the memory requirements increase.

AFAIK Litecoin and most Litecoin forks use sCrypt however sCrypt does include tuneable values so the Coin Maker can chose how much memory, latency and parallelism the algo will require.
Most AltCoins are Copy and Replace Jobby's so I wouldn't expect much difference in the algo.

Great info - thanks!
newbie
Activity: 49
Merit: 0
Yeh, I highly doubt this company can produce the hash rates they are saying and for that cost. Increasing the clock speeds will have little to no effect over their "prototype". And their prototype seems to be a copy of https://github.com/kramble/FPGA-Litecoin-Miner. They state that they are using the units ram which means they have not found a way to efficiently store the pseudo-random data on die / chip (in this case in logic gates on FPGA).

Like I said " I highly doubt ". I mean sCrypt ASIC...... Don't you think a larger established ASIC mining equipment company like BFL would have done it already if it was so easy that Alpha just has to have the units "optimised".......

sCrypt is MEMORY HARD people = Process / Memory ratio of about 1-1, I.e. you need x amount of memory per x random logic. This equates to actual space requirements on die and compared to the costing requirements on other algo's is one of the highest.

I have racked my brain many a night on ways to produce a unit that is more cost effective than plain GPGPU units and it never works out, the only plus with some systems is power usage but at an incredibly higher cost than GPGPU units.

Read Colin Percivals paper on his sCrypt Algo http://www.tarsnap.com/scrypt/scrypt.pdf then make your mind up as to if you want to give this company money.

It's not impossible to build a unit like they are advertising, just WAY more costly than building units for SHA256 and due to the memory requirements ASIC (in its literal form) is highly unlikely.

I don't wish to slander this company, merely to make people aware of the facts.


+1 Excellent points and great link - thanks.

But i did hear somewhere that the sCrypt flavour used in cryptocurrencies is different sCrypt 'proper'.
do you know anything about that?

Yeh, coins like YaCoin use sCrypt-Jane which is a slighly modified version of sCrypt using different mix functions. sCrypt-Jane is cool because it dynamically changes the memory requirements of the algo therefore it will become more CPU friendly and less GPU friendly as the memory requirements increase.

AFAIK Litecoin and most Litecoin forks use sCrypt however sCrypt does include tuneable values so the Coin Maker can chose how much memory, latency and parallelism the algo will require.
Most AltCoins are Copy and Replace Jobby's so I wouldn't expect much difference in the algo.
hero member
Activity: 714
Merit: 500
They have not wheres near to capital to produce an ASIC unit.  The natural progression is to go to FPGA first.  Recoup the cost of FPGA and then is the FPGA profits to invest is ASIC.  Moving directly to ASIC with "unproven code" means higher failure rates.
sr. member
Activity: 448
Merit: 250
Yeh, I highly doubt this company can produce the hash rates they are saying and for that cost. Increasing the clock speeds will have little to no effect over their "prototype". And their prototype seems to be a copy of https://github.com/kramble/FPGA-Litecoin-Miner. They state that they are using the units ram which means they have not found a way to efficiently store the pseudo-random data on die / chip (in this case in logic gates on FPGA).

Like I said " I highly doubt ". I mean sCrypt ASIC...... Don't you think a larger established ASIC mining equipment company like BFL would have done it already if it was so easy that Alpha just has to have the units "optimised".......

sCrypt is MEMORY HARD people = Process / Memory ratio of about 1-1, I.e. you need x amount of memory per x random logic. This equates to actual space requirements on die and compared to the costing requirements on other algo's is one of the highest.

I have racked my brain many a night on ways to produce a unit that is more cost effective than plain GPGPU units and it never works out, the only plus with some systems is power usage but at an incredibly higher cost than GPGPU units.

Read Colin Percivals paper on his sCrypt Algo http://www.tarsnap.com/scrypt/scrypt.pdf then make your mind up as to if you want to give this company money.

It's not impossible to build a unit like they are advertising, just WAY more costly than building units for SHA256 and due to the memory requirements ASIC (in its literal form) is highly unlikely.

I don't wish to slander this company, merely to make people aware of the facts.


+1 Excellent points and great link - thanks.

But i did hear somewhere that the sCrypt flavour used in cryptocurrencies is different sCrypt 'proper'.
do you know anything about that?
sr. member
Activity: 448
Merit: 250
That's a pretty quick change of heart -- do you know something the rest of us don't?
Sometimes it's just that gut feeling .... at least I got my refund back. I will wait for version 2 perhaps. See if they truly come through first. In the meantime I will gpu mine. Plus there is another company that was advertising a 25 mh/s for 800 gbp I believe...

Scrypt ASIC International (SAI) are offering 25MH/s for GBP 800 and delivery in February.

Yeah, sure. And i am Satoshi Nakamoto...

Do your due diligence before handing over any coins to anyone.

https://bitcointalksearch.org/topic/captchadd-is-very-likely-the-owner-of-scryptasicorg-scam-312714

legendary
Activity: 1862
Merit: 1002
How many units are expected to be sold in batch 1? Does anyone know?
newbie
Activity: 49
Merit: 0
Yeh, I highly doubt this company can produce the hash rates they are saying and for that cost. Increasing the clock speeds will have little to no effect over their "prototype". And their prototype seems to be a copy of https://github.com/kramble/FPGA-Litecoin-Miner. They state that they are using the units ram which means they have not found a way to efficiently store the pseudo-random data on die / chip (in this case in logic gates on FPGA).

Like I said " I highly doubt ". I mean sCrypt ASIC...... Don't you think a larger established ASIC mining equipment company like BFL would have done it already if it was so easy that Alpha just has to have the units "optimised".......

sCrypt is MEMORY HARD people = Process / Memory ratio of about 1-1, I.e. you need x amount of memory per x random logic. This equates to actual space requirements on die and compared to the costing requirements on other algo's is one of the highest.

I have racked my brain many a night on ways to produce a unit that is more cost effective than plain GPGPU units and it never works out, the only plus with some systems is power usage but at an incredibly higher cost than GPGPU units.

Read Colin Percivals paper on his sCrypt Algo http://www.tarsnap.com/scrypt/scrypt.pdf then make your mind up as to if you want to give this company money.

It's not impossible to build a unit like they are advertising, just WAY more costly than building units for SHA256 and due to the memory requirements ASIC (in its literal form) is highly unlikely.

I don't wish to slander this company, merely to make people aware of the facts.
sr. member
Activity: 433
Merit: 250
Looks like Hasfast is struggling http://www.coindesk.com/asic-manufacturer-hashfast-faces-legal-action/

I don't know if Alpha is a scam but do they have the expertise to get hundreds, maybe thousands of asics into production on time and on budget? I'm not so sure. I guess time will tell.
member
Activity: 99
Merit: 10
I'm moving back up to Liverpool on Saturday so will be paying these guys a visit very soon..
newbie
Activity: 56
Merit: 0
That's a pretty quick change of heart -- do you know something the rest of us don't?
Sometimes it's just that gut feeling .... at least I got my refund back. I will wait for version 2 perhaps. See if they truly come through first. In the meantime I will gpu mine. Plus there is another company that was advertising a 25 mh/s for 800 gbp I believe...
hero member
Activity: 714
Merit: 500
 Alpha Technologies... Butterfly Labs in sheep's clothing.
hero member
Activity: 574
Merit: 500
That's a pretty quick change of heart -- do you know something the rest of us don't?
Jump to: