Author Topic: Alpha Technology Litecoin (Scrypt) ASIC Miner Order Batch 1 Now!  (Read 360088 times)

Offline AlphaTechnology

  • Jr. Member
  • **
  • Posts: 42
    • Alpha Technology
Re: LTC FPGA Miner
« Reply #15 on: July 24, 2013, 01:07:55 AM »
Update: We have been able to get the Litecoin algorithm working in FPGA. However, after further analysis we realised that FPGA is not the correct platform for this algorithm in terms of high-speed hashrate, it is possible however not financially or efficiently viable. This is primarily, because of the design of the Scrypt algorithm, there are few loops which require a lot of memory and all operations have to be sequential which cannot be made parallel, so you will not be able to get high speeds. The algorithm was designed in this way to avoid hacking.

The problem with FPGA is that it cannot work at very high frequency. An example is we have seen 800 Mhz in Zynq series. The fastest memory available to an FPGA is on itself which is the block ram. And that we don't have in abundance. This is something which is present in GPU boards though.

You need to physically implement the algorithm on an FPGA in order to actually see how in depth the algorithm works, simulations are not enough. So our efforts have been beneficial in order to come up with a better method of litecoin hashing. We will be releasing a full analysis report as a proof of concept of the algorithm on an FPGA anyway in the near future.

The implementation has been beneficial to us as a team and most importantly our well established partners, in order to get fully familiar with the algorithm itself. So it has given us an in depth knowledge in our new design to overcome the limitations of an FPGA in this case. (will be disclosing who our partners are after our official press releases).

Therefore, We believe the demand is high enough to go ahead with this. We are working on a fabric based on a GPGPU design, as the only way to speed up a sequential algorithm is by overclocking or choosing a fabric which can run at very high speeds and has got lots of memory. As GPGPU implementations have different syntaxes than that of FPGA's, we will be taking advantage of our partner's expertise and connections in manufacturing in this area.

Not only is this a better high speed choice as an end product for miners, it is also business wise for us as we can take advantage of economies of scale, and lower cost per unit therefore lower price for the customers (FPGA's are much more expensive). Our actual design will be reaching speeds in the 1000s Kilohashes/second.

We are in early development and much more information will be coming soon. For now visit our website or follow us on twitter for early development updates.

 

Website: www.alpha-t.net

Twitter: www.twitter.com/UKAlpha

Company Name: Alpha Technology (INT) LTD

company Number: 8483921

Established Partner (Will be disclosing soon)
« Last Edit: July 26, 2013, 08:36:43 PM by AlphaTechnology »

Offline kramble

  • Full Member
  • ***
  • Posts: 247
    • Github
Re: LTC FPGA Miner
« Reply #16 on: July 24, 2013, 01:16:04 AM »
Huge Update: We have been able to get the Litecoin algorithm working in FPGA. However, after further analysis we realised that FPGA is not the correct platform for this algorithm in terms of high-seed hashrate

Tell us something new, scrypt was designed that way. As for the rest of your post, its just marketing speak. Given my (very limited) experience with fpga litecon mining, I'm backing jasinlee's project as the most likely to come up trumps. Lets just wait for some real verified results shall we? (I can give 1kHash/sec as a starter, since noone else seems to have posted anything verifiable  :P )

Offline mhysa

  • Newbie
  • *
  • Posts: 18
Re: LTC FPGA Miner
« Reply #17 on: July 24, 2013, 02:09:13 AM »
Our actual design will be reaching speeds in the 1000s Kilohashes/second.
At what cost? What price point are you aiming for? And can you narrow down what you mean by 1000s? Is it closer to 2,000 Kh/s or greater than 10,000 Kh/s?

Established Partner (Will be disclosing soon)

I hope your secret established partner isn't BFL.

Offline AlphaTechnology

  • Jr. Member
  • **
  • Posts: 42
    • Alpha Technology
Re: LTC FPGA Miner
« Reply #18 on: July 24, 2013, 02:38:02 AM »
Thanks for your reply, as you already know we are still in early development, and it's good if you understand the finance and manufacturing side of things before you ever think to pre-order or purchase shares in an 'IPO'. First of all Bitcoin algorithm (double SHA-256) could easily be implemented in an FPGA, it is not memory intensive at all and Bitcoin was NOT developed to be FPGA resistant. You could maximise hardware efficiency with a Bitcoin FPGA. Now lets look at it from a manufacturing point of view, at the time when Bitcoin FPGA's were released, cost-efficient powerful FPGA's were being used (Spartan-6 LX150). However, even then cost wise FPGA's are expensive, as they are really designed from prototyping. Example ModMiner Quad 900 mh/s for $1,069. The cheapest was the Bitforce BFL FPGA, 832 mh/s for $600. BFL having a lot of funding and being a somewhat big player in the community must of had connections with a limited supplier to be able to sell their product for that price, as the number of BFL FPGA's being sold was very limited in order to sell that cheap, and indeed it was. Look at the volume of ASIC's they are selling now, which is easily financially viable due to the LOW per unit cost of an ASIC (high NRE).

We had this in mind, but thought there could be a way to shortcut the FPGA implementation after actually physically doing it. But it isn't possible for the reasons i stated in my last post, and trust me, we are working with our partners who have over 150 employees, hardware and software experts in FPGA design with many years of experience (i wish i could tell you who they are, but you will have to wait a few weeks before they release official news release). That being said, now look at Litecoin, it contains not only Scrypt, but also many other algorithms, and has forcefully made some operations that cannot be made parallel. If Bitcoin FPGA's were barely profitable and high priced and only limited in quantity (they just paved the way for ASIC), what makes you think an unknown developer with no hardware experience making an IPO (which has clear fraud issues) can even compete with power/cost of a Bitcoin FPGA at the time Bitcoin FPGA's were up for sale. Which was generally even higher than 1$ per megahash. It's impossible as the Litecoin algorithm also contains SHA-256 algorithm aswell as many others including Scrypt, which ALL needs to be hashed, so Bitcoin is always going to be multitude times FPGA cost effective. We will release a proof of concept to show this and a full report, if anyone is stating they have exact FPGA hashing figures in the 2MH/s even at $5,000 (thats minimum) area that means they are in the late stages of development, and can easily show you a video/image proof of that.

On the other hand we are taking a far realistic approach specialized for the characteristics of the actual algorithm. Not only that our partners who are a great addition (most likely biggest) to the crypto-currency hardware market, which gives us access to facilities/manufacturing/economies of scale and are direct partners with companies like Texas Instruments.

I'm not bashing any competitors, i'm just being realistic from a hardware manufacturing point of view, which is the best point of view you'll get in this matter. And what to expect when certain projects fail, due to unrealistic expectations or fraud.
« Last Edit: July 24, 2013, 02:42:19 AM by AlphaTechnology »

Offline jasinlee

  • Scammer
  • Elite Member
  • ************
  • Posts: 1549
Re: LTC FPGA Miner
« Reply #19 on: July 24, 2013, 03:44:27 AM »
Thanks for your reply, as you already know we are still in early development, and it's good if you understand the finance and manufacturing side of things before you ever think to pre-order or purchase shares in an 'IPO'. First of all Bitcoin algorithm (double SHA-256) could easily be implemented in an FPGA, it is not memory intensive at all and Bitcoin was NOT developed to be FPGA resistant. You could maximise hardware efficiency with a Bitcoin FPGA. Now lets look at it from a manufacturing point of view, at the time when Bitcoin FPGA's were released, cost-efficient powerful FPGA's were being used (Spartan-6 LX150). However, even then cost wise FPGA's are expensive, as they are really designed from prototyping. Example ModMiner Quad 900 mh/s for $1,069. The cheapest was the Bitforce BFL FPGA, 832 mh/s for $600. BFL having a lot of funding and being a somewhat big player in the community must of had connections with a limited supplier to be able to sell their product for that price, as the number of BFL FPGA's being sold was very limited in order to sell that cheap, and indeed it was. Look at the volume of ASIC's they are selling now, which is easily financially viable due to the LOW per unit cost of an ASIC (high NRE).

We had this in mind, but thought there could be a way to shortcut the FPGA implementation after actually physically doing it. But it isn't possible for the reasons i stated in my last post, and trust me, we are working with our partners who have over 150 employees, hardware and software experts in FPGA design with many years of experience (i wish i could tell you who they are, but you will have to wait a few weeks before they release official news release). That being said, now look at Litecoin, it contains not only Scrypt, but also many other algorithms, and has forcefully made some operations that cannot be made parallel. If Bitcoin FPGA's were barely profitable and high priced and only limited in quantity (they just paved the way for ASIC), what makes you think an unknown developer with no hardware experience making an IPO (which has clear fraud issues) can even compete with power/cost of a Bitcoin FPGA at the time Bitcoin FPGA's were up for sale. Which was generally even higher than 1$ per megahash. It's impossible as the Litecoin algorithm also contains SHA-256 algorithm aswell as many others including Scrypt, which ALL needs to be hashed, so Bitcoin is always going to be multitude times FPGA cost effective. We will release a proof of concept to show this and a full report, if anyone is stating they have exact FPGA hashing figures in the 2MH/s even at $5,000 (thats minimum) area that means they are in the late stages of development, and can easily show you a video/image proof of that.

On the other hand we are taking a far realistic approach specialized for the characteristics of the actual algorithm. Not only that our partners who are a great addition (most likely biggest) to the crypto-currency hardware market, which gives us access to facilities/manufacturing/economies of scale and are direct partners with companies like Texas Instruments.

I'm not bashing any competitors, i'm just being realistic from a hardware manufacturing point of view, which is the best point of view you'll get in this matter. And what to expect when certain projects fail, due to unrealistic expectations or fraud.

Wow, you really tried hard to pull a passive aggressive offhanded insult there.

Offline polishboy1098

  • Newbie
  • *
  • Posts: 5
Re: LTC FPGA Miner
« Reply #20 on: July 24, 2013, 02:45:01 PM »
The only one that I think committed any fraud was you. You claimed that you have over 150 employees working on this project yet you only now realize that scrypt is memory intensive, and not able to be done in parallel. That is one of the main selling points of scrypt that it is memory intensive and FPGA/asic resistant because this. Either your whole team had been doing NOTHING until now or you have no team.

Offline Wolf0

  • Sr. Member
  • ****
  • Posts: 1041
Re: LTC FPGA Miner
« Reply #21 on: July 25, 2013, 06:38:43 AM »
The only one that I think committed any fraud was you. You claimed that you have over 150 employees working on this project yet you only now realize that scrypt is memory intensive, and not able to be done in parallel. That is one of the main selling points of scrypt that it is memory intensive and FPGA/asic resistant because this. Either your whole team had been doing NOTHING until now or you have no team.

Or, giving him the benefit of the doubt, he has terrible communication with his engineers.

Offline AlphaTechnology

  • Jr. Member
  • **
  • Posts: 42
    • Alpha Technology
Re: LTC FPGA Miner
« Reply #22 on: July 26, 2013, 04:40:15 AM »
Thanks for your reply. I merely stated our partner has over 150 employees. That doesn't mean all of them are working simultaneously on this specific project. Just stating our range of experience in different areas of development. And even though it is memory intensive, it is still a good idea to test all hardware options and variables before committing to the final architecture. We have been told by our partners to not release any information on our partnership till more work is done. But when we do in the coming weeks hopefully your concerns will be put to rest.

We also believe it is a good idea for us to only release a development update when we have reached bigger milestones. So no more information or replies will be given till that stage.

Offline jk_14

  • Sr. Member
  • ****
  • Posts: 572
Re: LTC FPGA Miner
« Reply #23 on: July 26, 2013, 09:38:13 AM »
And even though it is memory intensive, it is still a good idea to test all hardware options and variables before

Also is a good idea, even earlier - to search a bit litecoin forum...
Just to avoid such sense announcements: "After our deep research - we've discovered scrypt is memory intensive"
« Last Edit: July 26, 2013, 09:45:29 AM by jk_14 »

Offline AlphaTechnology

  • Jr. Member
  • **
  • Posts: 42
    • Alpha Technology
Re: LTC FPGA Miner
« Reply #24 on: July 26, 2013, 12:23:39 PM »
Our Scrypt implementation on an FPGA (proof of concept) has been fully analysed and documented in a report for public access. You can download the .pdf file on this link https://mega.co.nz/#!AtFC3YLY!ZK_DB2kyXsasPHiZlV9uNH1V_cVYAlO908ut_7qR7qY
« Last Edit: November 01, 2013, 07:58:17 PM by AlphaTechnology »

Offline kramble

  • Full Member
  • ***
  • Posts: 247
    • Github
Re: LTC FPGA Miner
« Reply #25 on: July 26, 2013, 01:16:09 PM »
You may want to amend that report slightly, I don't recall any place named "Manchestar, England" (sic), and I only live a hundred or so miles away. Otherwise it looks fairly accurate. I'll give you another data point for your maximum hash rates. I'm getting 1.12kHash/sec at 30MHz on a DE0-Nano ALTERA EP4CE22 (and that's only an initial naive implementation).

Offline razorfishsl

  • Jr. Member
  • **
  • Posts: 54
Re: LTC FPGA Miner
« Reply #26 on: August 05, 2013, 11:58:48 PM »
Yep
 I read the report and the mix seems out.....

I'm seeing ~225us @100mhz...
rom mix 1024 times x 2 Salsa x 11

RF

Offline xus

  • Jr. Member
  • **
  • Posts: 26
Re: LTC FPGA Miner
« Reply #27 on: August 07, 2013, 01:32:20 PM »
26/07/2013 Update Our Scrypt implementation on an FPGA (proof of concept) has been fully analysed and documented in a report for public access. You can download the .pdf file on this link http://alpha-t.net/blogs/news/8373831-litecoin-miner-development-update-26-07-2013

Such things should be really done in LaTeX... would look way more professional.

Offline razorfishsl

  • Jr. Member
  • **
  • Posts: 54
Re: LTC FPGA Miner
« Reply #28 on: September 01, 2013, 01:33:14 AM »
26/07/2013 Update Our Scrypt implementation on an FPGA (proof of concept) has been fully analysed and documented in a report for public access. You can download the .pdf file on this link http://alpha-t.net/blogs/news/8373831-litecoin-miner-development-update-26-07-2013

Such things should be really done in LaTeX... would look way more professional.

I don't see how some guy in a rubber diving suit is going to make them look any more professional.....

Offline Wolf0

  • Sr. Member
  • ****
  • Posts: 1041
Re: LTC FPGA Miner
« Reply #29 on: September 02, 2013, 04:25:49 AM »
26/07/2013 Update Our Scrypt implementation on an FPGA (proof of concept) has been fully analysed and documented in a report for public access. You can download the .pdf file on this link http://alpha-t.net/blogs/news/8373831-litecoin-miner-development-update-26-07-2013

Such things should be really done in LaTeX... would look way more professional.

I don't see how some guy in a rubber diving suit is going to make them look any more professional.....

I hope you're joking. Either that or you're incredibly stupid.