Bitcoin Forum
November 05, 2024, 06:47:49 PM *
News: Latest Bitcoin Core release: 28.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 [14] 15 16 17 »  All
  Print  
Author Topic: [ANN] labcoin ASIC - v.1 @130nm/300Mhz -16 cores chip - European / Chinese Team  (Read 33658 times)
tucenaber
Sr. Member
****
Offline Offline

Activity: 337
Merit: 252


View Profile
August 01, 2013, 08:01:11 AM
 #261

So, congratulations on a successful IPO. Now let's get some real updates on the project status.

You may want to send TheSeven over here to explain things.
labcoin (OP)
Full Member
***
Offline Offline

Activity: 154
Merit: 100


Labcoin.com


View Profile WWW
August 02, 2013, 02:33:05 AM
 #262

As announced here https://bitcointalk.org/index.php?topic=263445.msg2850861#msg2850861 we do look forward to provide more updates.
A website update is on the way as well, please post any question here or in the security thread as you see fit.

Thanks

Sam

http://labcoin.com - Proprietary ASIC technology
Ytterbium
Full Member
***
Offline Offline

Activity: 238
Merit: 100



View Profile WWW
August 02, 2013, 05:22:20 AM
 #263

My understanding is you've ordered a test first run of ICs at 130nm, and if those chips work you plan on ordering a larger run, correct?

When do you expect to receive the first run, and how many chips will you be getting? 

If the chips work, how long will it take to get them up and running on the network?

With the successful IPO, how many ICs will you order in the second run, and when exactly do you expect to receive them?

Are you still planning on a 65nm design next, when do you expect to tape out, etc?

labcoin (OP)
Full Member
***
Offline Offline

Activity: 154
Merit: 100


Labcoin.com


View Profile WWW
August 02, 2013, 04:02:34 PM
 #264

My understanding is you've ordered a test first run of ICs at 130nm, and if those chips work you plan on ordering a larger run, correct?

When do you expect to receive the first run, and how many chips will you be getting? 

If the chips work, how long will it take to get them up and running on the network?

With the successful IPO, how many ICs will you order in the second run, and when exactly do you expect to receive them?

Are you still planning on a 65nm design next, when do you expect to tape out, etc?


Correct, the first batch of ICs is meant as  a test for our design. Delivery is expected in mid September, in quantities of several thousands pieces.
We expect to have the boards up and running within 7/10 days after we receive the ICs.
65 nm design is currently under development and is expected to go on tape-out in late August or early September.

http://labcoin.com - Proprietary ASIC technology
tucenaber
Sr. Member
****
Offline Offline

Activity: 337
Merit: 252


View Profile
August 02, 2013, 08:27:59 PM
 #265

Correct, the first batch of ICs is meant as  a test for our design. Delivery is expected in mid September, in quantities of several thousands pieces.
We expect to have the boards up and running within 7/10 days after we receive the ICs.
65 nm design is currently under development and is expected to go on tape-out in late August or early September.

So you are not even going to wait to see if the 130nm chips work before going ahead with the next generation? Are you that confident?
labcoin (OP)
Full Member
***
Offline Offline

Activity: 154
Merit: 100


Labcoin.com


View Profile WWW
August 03, 2013, 11:25:15 AM
 #266

Correct, the first batch of ICs is meant as  a test for our design. Delivery is expected in mid September, in quantities of several thousands pieces.
We expect to have the boards up and running within 7/10 days after we receive the ICs.
65 nm design is currently under development and is expected to go on tape-out in late August or early September.

So you are not even going to wait to see if the 130nm chips work before going ahead with the next generation? Are you that confident?

All tests and simulations show that we can be confident on the quality of our work, additionally, we will use this run of 130nm chips to deploy our own mining power and fine tune the hardware, before going for a larger run.
The 65nm design is expected to be ready within the mentioned timeframe, but it's not guaranteed to, we will take the best decisions according to how events will unfold.

http://labcoin.com - Proprietary ASIC technology
polarhei
Sr. Member
****
Offline Offline

Activity: 462
Merit: 250


Firing it up


View Profile
August 03, 2013, 02:11:49 PM
 #267

When we are able to order some hashrate from you? I am worried that there are scams which like to use pre-order to cheat us.
Ytterbium
Full Member
***
Offline Offline

Activity: 238
Merit: 100



View Profile WWW
August 03, 2013, 08:57:16 PM
Last edit: August 04, 2013, 12:04:09 AM by Ytterbium
 #268

I'd advise you not to take pre-orders for whole miners, instead just sell bulk chips and possibly reference PCBs

All the companies that are a little late with actually manufacturing working miners get massive amounts of hate. I'd say to avoid the risk and let third parties build systems and hobbyists build their own miners from components (possibly soldering on their own chips).

That will also create a "filter" that will prevent people who don't know anything from ordering from you and then flipping out all over the forum.

superduh
Hero Member
*****
Offline Offline

Activity: 602
Merit: 500


View Profile
August 03, 2013, 10:43:35 PM
 #269

labcoin,

after 65 design will you begin work on 28nm? (q1/q2 2014?)

ok
tucenaber
Sr. Member
****
Offline Offline

Activity: 337
Merit: 252


View Profile
August 04, 2013, 04:16:59 PM
 #270

Today we have a very important update, the Chinese team is simulating a lot of design simultaneously and worked almost non-stop for the last 48 hours  targeting different process sizes.

The results are more than positive, i will try to outline them in the clearest way possible.

1) The 65nm 500Mhz is still undergoing post-verification phase, while another simulation is ongoing at 600Mhz and we're waiting for the results.

2)  Post simulations yielded positive results on a 130nm, 300Mhz, Power 0.8W, 6.5x6.5mm design.
    
     The team is working on HDL optimizations to get 16 cores for chip.

     Some math, quoted from the tech team
Quote
    "300M*16=4.8G, 0.8*16=12.8W, Area=130,0000*16=2080,0000, make the utilization ratio to 50%, the chip size will be about 4160,0000um2, about 6.5mm x 6.5mm"

     "Power consumption per GHash is 12.8W/4.8G=2.7W/GHash"

      "Estimated selling price for chip, 8-9 USD"

What does this mean ? i think it's not hard to get.  130nm process and 5GH speed at slightly higher power consumption, but competitive prices.

Shoot any question guys

Sam
Labcoin team

How can this version be so much smaller than the earlier 180nm design? I assume this is more or less the 180nm chip design shrunk to 130nm, and the numbers seem right except for the size. How in the world can you fit 16 cores in that small area? I can't make sense of it. The 180nm version would be 5mmx5mm so this one should be a square with 5mm*130/180=3.6mm, no? 16 of these would be 14x14. Can anybody help me figure this out?

VolanicEruptor
Sr. Member
****
Offline Offline

Activity: 392
Merit: 250



View Profile
August 07, 2013, 03:37:26 PM
 #271

wtf is going on here.. nothing for 3 days?  Was that question too tough?

Loredo
Full Member
***
Offline Offline

Activity: 238
Merit: 100


View Profile
August 07, 2013, 04:17:31 PM
 #272

So you are not even going to wait to see if the 130nm chips work before going ahead with the next generation? Are you that confident?
I'm sure that all these questions of design strategy were covered in detail in the Business Planning and Strategy section of the Offering Memorandum that must have been available prior to the IPO...
VolanicEruptor
Sr. Member
****
Offline Offline

Activity: 392
Merit: 250



View Profile
August 07, 2013, 07:43:33 PM
 #273

someone answer tucenabers question, its so legit its creeping me out that its being ignored

Vbs
Hero Member
*****
Offline Offline

Activity: 504
Merit: 500


View Profile
August 08, 2013, 09:50:46 AM
 #274

Today we have a very important update, the Chinese team is simulating a lot of design simultaneously and worked almost non-stop for the last 48 hours  targeting different process sizes.

The results are more than positive, i will try to outline them in the clearest way possible.

1) The 65nm 500Mhz is still undergoing post-verification phase, while another simulation is ongoing at 600Mhz and we're waiting for the results.

2)  Post simulations yielded positive results on a 130nm, 300Mhz, Power 0.8W, 6.5x6.5mm design.
    
     The team is working on HDL optimizations to get 16 cores for chip.

     Some math, quoted from the tech team
Quote
    "300M*16=4.8G, 0.8*16=12.8W, Area=130,0000*16=2080,0000, make the utilization ratio to 50%, the chip size will be about 4160,0000um2, about 6.5mm x 6.5mm"

     "Power consumption per GHash is 12.8W/4.8G=2.7W/GHash"

      "Estimated selling price for chip, 8-9 USD"

What does this mean ? i think it's not hard to get.  130nm process and 5GH speed at slightly higher power consumption, but competitive prices.

Shoot any question guys

Sam
Labcoin team

How can this version be so much smaller than the earlier 180nm design? I assume this is more or less the 180nm chip design shrunk to 130nm, and the numbers seem right except for the size. How in the world can you fit 16 cores in that small area? I can't make sense of it. The 180nm version would be 5mmx5mm so this one should be a square with 5mm*130/180=3.6mm, no? 16 of these would be 14x14. Can anybody help me figure this out?


+1

It's called "designing IC's by multiplication"! Where are the proper simulation results?
tucenaber
Sr. Member
****
Offline Offline

Activity: 337
Merit: 252


View Profile
August 08, 2013, 01:07:49 PM
 #275

Time for some wild speculation. They might abandon the 180 nm core altogether. As the 180 nm tape-out was nearing, they realized it´s not gonna compete, so they stopped it for good. They were simultaneously working on a full-custom 65 nm design and actually based their 130 nm (scaled up) core on that. There were two posts both in the IPO and hardware threads from user labcoin which would back this speculation up:

Quote
1. Chip specifications

There are substantial differences in the way  the BFL chip is produced. They do employ a standard cell ASIC, while we went for a custom design with a focus on performances, achieved via a complex place & route procedure, which took our team almost one full-immersion month of work to complete.
We do confirm that we're expecting to obtain the initially declared performances with the 130nm round, but we will wait for the ICs to be ready, to better assess the yield quality in terms of chip grades.

Quote
First thing to say, while the team in China was developing the first 180nm iteration, it was simultaneously exploring a 65nm IC design as well.

Few days ago, the team was carrying on with the final preparation steps to put our 180nm design out  for final tape-out with UMC fab.

Briefly after hearing about the latest news regarding ASICMINER new deployment of 800-1000 TH within the year, the management decided on a strategy shift.

We stopped at last minute the tape-out of 180nm first iteration test-run, and further development stopped as well.

The Chinese team is now optimizing the initial 65nm design to work at 433Mhz, with over-clocking possible;
we expect to release final estimated hash rate and more precise estimations about power consumption data for this IC in 24 hours.

If all the post-simulations are positive, this IC design will go on tape-out, without test runs but with minimized risks, in a few days;  TSMC has a 65nm CLN65/CMN65 on schedule for 8 and 22 July.
Should we miss the 8 July shuttle we will go for the 22 July one. After about 50 days the deployment will start.

Can´t wait for the announced QA session, it´s already kind of overdue  Sad


Not bad. If you are right then we are in good shape, 'cause that's one awesome chip, it looks like to me. Labcoin, please confirm Wink
TheSwede75
Full Member
***
Offline Offline

Activity: 224
Merit: 100



View Profile
August 08, 2013, 03:24:22 PM
 #276

UPDATE:

The developers are now open for live Q/A on IRC irc.freenode.org - #labcoin

Labcoin is still waiting for the Bitcointalk accounts to be whitelisted. Unfortunately I will be traveling today and have very limited Internet access for a few hours, so if someone from this thread would like to transcribe the Q/A for the forum I would be very grateful.

The Founder and the Developers WILL hold Q/A sessions and answer questions here on the forum as soon as their accounts are white-listed.
TheSwede75
Full Member
***
Offline Offline

Activity: 224
Merit: 100



View Profile
August 08, 2013, 05:14:23 PM
 #277

UPDATE:

Here is TAT's excellent transcript of the first Q/A session:

1. Please give us a detailed specification of the chip and its current status. What is the power efficiency?
Alright,the 130nm IC is designed to work at about 200 Mhz for core, with a total consumption of 0.8W per core, hashing performances of about 200 mhash per core, total speed will depend  on the overall output grade.
The IC covers an area of 6.5mm x 6.5mm and it has been designed following ideas similar to those employed in the bitfury ASIC.

2. How will you show transparency with the 75% retained shares?
the IPO manager will arrange some kind of way for burnside to release that information

3. Are your first run ASICs in production?
the first test run is already out to fab

4. How many chips do you expect to get out of the first batch? When will you have sample chips?
estimate date for receiving sample chip, one week later, if there are no delays, the plan is having boards produced within the first week of september

5. How is it possible to create chips with a such much better performance than the competitors. You've arranged them to fit more on the same size? For that to be possible, not only each Labcoin core would have to be ~42% smaller [65/130*(6.5^2)/(7.1^2)] than each BFL core but also the Labcoin chip would magically operate at a higher frequency (300MHz vs 250MHz) while keeping the same power draw...
rolled cores instead of unrolled, sea-of-hashers approach. as in sea-of-gates. you can google this term and find out more about what it means technically. i remind you guys that the bitfury chip has up to 750 cores. they also use the same approach, we chose this design because one of our guys was close to bitfury developments 2 months ago

6. What is the software used for mining, protocol development ? I am assuming Theseven will be doing the firmware part?
yes theSeven is working on a sophisticated interface for the 65nm version

7. Is Labcoin a hardware sales or mining co?
We will focus on deploying internal hash power initially and won't ship USB miners yet, time is tight

8. How many wafers? How many TH's will be hashing in September? That knowledge will be known as soon as you get the chips, correct?
well not 30-50 TH. not sure what will be the yield of the test-run but it's in the thousands ICs. correct, also,next week we'll have the next batch going out to fab
in september we may have a few TH. i can't give exact figures, by the end of Sept, initial week of october we'll have the full speed online.

9. Is this chip the same design as the 180nm? or a completely new one?
 it's based on similar design choices,but it's indeed different, as it's using a different library,different logic blocks  layout

10. How can you fit 16 cores on 6.5x6.5 ? How many cores on average per chip, what is your assumption
 The cores we have fit are not classical "unrolled cores",the cores as we mean them are the result of our engineering effort. i'll  pass the questions i can't reply to the lead developer

11. Labcoin seems almost small now in the light of new competitors like btcgarden, hashfast, actm, terrahash.. do you think you can keep up?
btcgarden? Well, guys, i would do a reality check before saying such things. Undoubtedly ASICMINER, AVALON and maybe BFL have some kind of advantage. But we don't see how we are inferior to those vaporware companies. What in their announcement makes them more convincing than us? time will tell and the sheer of technical details we have provided is not trivial. it's very unlikely we are making everything up

12. So if everything goes as planned - we'll see Labcoin start hashing in the first week of September? Can you please confirm?
confirm, hashing with moderate speed and testing and within the end of the month, we'll have the full speed online

13. What are the chances that the chips won't work? can you guestimate?
we're not worried about this, as soon as the hashing power comes up, the price will go the right route. there's a small chance that the chip won't work, but very very small. it's much more common to get slightly underperforming chips rather than completely dead ones

14. Who is Samuel Noi?
Sam is our nominee director, Fabrizio is the founder

15. What's the status of the boards? How long will it take after testing to get all the board you need, set them up with chips and get them hashing?
The boards are near their design completion and will be tested as soon as we have the chips. we'll post pictures of the boards very soon.  it will take about 7-10 days to get all the boards up and running probablyand, for the first run, we have 500 boards on order

16. Do you still depend on suppliers or do you have everything in stock? and what tools for p&r, sta, layout
we use  Mentor mostly, and its suite. we still depend on suppliers but we're closing in the gap. we'll be ordering components  according to the volume needed to avoid bottlenecks. 24july first one

17. Where is the lab? Shenzhen or Hongkong? Can we have a chance to visit? When if not now?
lab is located in SZ, at the university facility, we'll be moving shorty to another office. we'll be inviting selected investors to visit our datacenter in september

18. http://www.labcoin.com/presentation.html    are these photos of the real where u r working in?
yes,they are real. it's the university i mentioned

UPDATE:

I hope you felt that the Q/A session was informative. I will also, since he is an Authority that carries a fair amount of 'weight' in the forum make sure to ask TheSeven to take a more active role in the forum and our updates. As you have probably learned by now I am myself not an engineer or involved in development so the process of me asking Labcoin questions and them passing the answers to the forum via myself isn't the best route for technical discussion.

Thank you again!
Ytterbium
Full Member
***
Offline Offline

Activity: 238
Merit: 100



View Profile WWW
August 08, 2013, 07:27:18 PM
 #278

Hmm, I'm a bit nervous about the whole "by the end of September" thing for getting everything online. Hopefully they have a faster internal goal for getting that initial hash power online.  I'm expecting pretty massive diff increases starting in September when the first KnCs hit.

TheSwede75
Full Member
***
Offline Offline

Activity: 224
Merit: 100



View Profile
August 08, 2013, 07:29:42 PM
 #279

Hmm, I'm a bit nervous about the whole "by the end of September" thing for getting everything online. Hopefully they have a faster internal goal for getting that initial hash power online.  I'm expecting pretty massive diff increases starting in September when the first KnCs hit.

Of course the dates/times provided take into consideration lead time etc. I personally also feel that accounting for current max-diff increase per period is more than taking into account a LOT of ASIC delivery.
eve
Full Member
***
Offline Offline

Activity: 210
Merit: 100


View Profile
August 23, 2013, 05:36:03 AM
 #280

Hmm, I'm a bit nervous about the whole "by the end of September" thing for getting everything online. Hopefully they have a faster internal goal for getting that initial hash power online.  I'm expecting pretty massive diff increases starting in September when the first KnCs hit.

Of course the dates/times provided take into consideration lead time etc. I personally also feel that accounting for current max-diff increase per period is more than taking into account a LOT of ASIC delivery.

NO NEWS for the last 2 weeks or so, why? After doing a successful IPO they vanished into the thin air? Why No Updates? What's going On with LabCoin? When are the Chips coming out? Hello, where is Sam??
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 [14] 15 16 17 »  All
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!