Ok, I just wanted to make sure that you're really working on this, not just a joke  it's not a joke. but the feasibility check will take 2-3 month, and the prototype will take another 2-3 month. i hope it will come out in 2012. |
|
|
|
I received the Icarus package today from ngzhang  Shipping to Europe took 9 days. It was well packed and complete as promissed. The package contained also a USB stick with all software needed for mining, experimenting and dev. Within 10 minutes my Icarus dual fpga board was online and mining at Coinotron. During mining the device uses 26 watt and the short breaks in between blocks 8 watt. The hashrate at Coinotron moves between 200 and 400 Mhs/s, but now the pool is down so I think it's a pool problem.  and btcguild:  My first thoughts after mining for one hour are that ngzhang made a great fpga board with a lot of potential. One point of criticism, the tiny cooling fans make noise. I think I have to replace the fans for one bigger low noise fan. Thanks Zhang for your hard work and I will be the first to buy the 1.5 Ghash fpga from you, with 10% discount Some pics:      good, first. network conditions is deadly important for the real world mining speed. by my test, a 300-500ms ping speed to deepbit.net will generate a 330MH/s real-world hashing power.  there must be somthing wrong with the power consume. Icarus never uses more than 21 watt, less than 20 is a typical. and the idle state consuming 2-4w typ. i uses a 1% grade meter for the test, my be one of the meter is not dependable? and, please, please keep this thread to Icarus. the 1.5G rig is really really at a plan now, the SPEC./ speed/ price .... and so on is not tested in real world and not decided yet, i will not announce it officially until i have at lest 5 of them for mining 1 month +, i don't want to be called as scamers.  the development just started and looks like at least will take half year or even longer time to finish, it's a very hard and complex design, really. so please don't ask any time table or something details like that.
Good news. Looks like a high quality product. What size are the fans ? 40mm ?
it's a 25mm fan... |
|
|
|
2nd batch's manufacture has been finish today, i will test them in the after week. after first batch's response posted on the forum, 2nd batch pre-order will start. coming soon.  |
|
|
|
How come NG's is not updated here yet? =)
NOT shipping  |
|
|
|
It was never possible.
and by there "excellent" marketing communications, nearly all other unfirm FPGA developers scared and canceled their project. good job. |
|
|
|
let me show you how a 15KM distance is a big problem... i drive my CIVIC to the delivery station. actually 20KM away. this is the maxim speed... i leave home at 2PM and back at 6PM... 5 packages delivered successfully. Beijing's traffic is a disaster. really, really.  Hey ngzhang, you should consider applying the same techniques that your product uses. Instead of posting the boards yourself, you could UNROLL and hire 2x 64 delivery boys! You would send them to the post office, each with another packet, and CLOCK their departures at a 1 minute pace. At 19kph, each of them would need slightly more than 2 hours for the FULL round-trip. Yet, after filling the PIPELINE, one packet per minute would be dropped off at the post office. Also, every minute one IDLE delivery boy would show up again at your place. That's just in time to smoke a cigarette and pick up the next leaving packet!! Sustaining that schedule, you would effectively dedicate just 1 minute to delivery issues rather than several hours! That looks like more than 100 times better! PS: Also don't disregard the potential optimizations. For example you could instruct the boys to throw the packets through the post office door from the streetside, instead of entering. That would save them a minute for sure. You could then hire just 2x 63 boys instead of 2x 64, yet maintain the same 1 minute pace. That would be more efficient. Watch out for other opportunities as well! this is my first time to do international shipment, next time i will just send the package to the postoffice by a local express (1$ pre kilo). |
|
|
|
Any plans for a 540 or 720 Mhash/s version with 3 FPGAs or 4 FPGAs etc. ?
there is a 1.5Gh/s version in plan. is it a scam?  |
|
|
|
要买几个才有折扣？
现实的问题是买得到还是买不到的问题，而不是折扣多少的问题。 |
|
|
|
I didn't find any documentation in https://github.com/ngzhang/Icarus, just very brief README files. Is that the location that you were referring to? Can you point me more precisely to where I can find the answers to my previous questions? Thank you. sorry, technical detailes please re-check the https://github.com/ngzhang/Icarussite, all project source files are already released there, so i won't explain them everywhere. |
|
|
|
you mean 1.5MH/$? yeah, it's extremely possible in the next 6-8week. Meaning $120 for a fully assembled/tested ~180MH/s 6s150 FPGA miner?  -rph or a $1200 for a 1.8Gh/s miner |
|
|
|
You mention that each of your boards generates about 360MH/s.
Is that one mining instance running at 180MHz in each device, or two instances running at 90 MHz in each device, or some other configuration?
How full are each device? How many slices are used in each device? How many flip-flops?
Icarus project is a GPL'd open-sourced project, please check github for these questions. |
|
|
|
Well stable, the BFL unit is > 4x the hashrate than ztek and more than double ngzhang. Power draw is also more than double ngzhang (Dunno what ztek power draw is). Even allowing for a 10% efficiency decrease for "real world" scenario vs the test data, the numbers still hold. it means that if i change Icarus's price to less than 350$ than BFL's life is over? more like 250$ if you want almost the same amount of $/MH you mean 1.5MH/$? yeah, it's extremely possible in the next 6-8week. |
|
|
|
Well stable, the BFL unit is > 4x the hashrate than ztek and more than double ngzhang. Power draw is also more than double ngzhang (Dunno what ztek power draw is). Even allowing for a 10% efficiency decrease for "real world" scenario vs the test data, the numbers still hold. it means that if i change Icarus's price to less than 350$ than BFL's life is over? ADD@2011/12/10 there must be soming misunderstanding here, i mean some other product(my me or other) reach <1$/M, not Icarus. sorry. |
|
|
|
When will the second batch be ready? Thanks
I am very interested in these as well. Anxious to hear the reviews!
I'm also anxious to here the reviews. so i decided to start the second batch after some reviews posted on this thread. 2nd batch has 30 boards, i will get them next week. |
|
|
|
I have to do my groceries at a distance of 15 km so I don't understand why this can be a problem ? Oh... no... you don't have a car?
let me show you how a 15KM distance is a big problem... i drive my CIVIC to the delivery station. actually 20KM away. this is the maxim speed...  i leave home at 2PM and back at 6PM... 5 packages delivered successfully. Beijing's traffic is a disaster. really, really. |
|
|
|
ah, first user appeared. i just give 1 board to xiangfu @ beijing by personal.
That's nice, but I would like to play also with Christmas, why don't you ship the orders that are paid for ? oh, i must go to a post-office 15KM away to mail internationally, my plan is after all buyer's payment is finished, than ship them all together. this morning, last one payment is finished, and i pack all icarus to box tonight, tomorrow they will be all send out. I have to do my groceries at a distance of 15 km so I don't understand why this can be a problem ? Oh... no... you don't have a car? fell at ease. I do, I'm just eager to make coins with your fpga board so i send them @ 9/12, 1 day earlier than i announced (10/12). |
|
|
|
ah, first user appeared. i just give 1 board to xiangfu @ beijing by personal.
That's nice, but I would like to play also with Christmas, why don't you ship the orders that are paid for ? oh, i must go to a post-office 15KM away to mail internationally, my plan is after all buyer's payment is finished, than ship them all together. this morning, last one payment is finished, and i pack all icarus to box tonight, tomorrow they will be all send out. fell at ease. |
|
|
|
您的比特币在MTGOX出售后获得到的美元，怎么才能无损失无手续费的回到中国国内？ 目前网上兑换渠道都要收取手续费，非常令人不舒服。 现在好了，我决定零手续费收购您的MTGOX美元，按照中国银行当日美元兑人民币的牌价进行收购，人民币通过支付宝立即付款进入您的帐户。 在兑换过程中，我可以保证，如果您MTGOX帐户中有100美元，如果当日中国银行外汇牌价是6.33:1，那么我可以保证您的支付宝帐户中出现633元人民币！真正的零手续费！ 信誉：我是爱墙网 www.lovewall.net 的站长李光耀，爱墙网是2005年建立的，是我的生命，我想让她千秋万代的传下去，因此我的信誉是和爱墙网捆绑在一起的。为证明我的身份，我在爱墙网首页底部增加了一个连接“站长信息”，页面内有我的姓名“李光耀”和支付宝账号“powercaogen@sina.com”，我的QQ：821706149（草根的力量）。 如果您不放心的话，还可以分次交易。即100美元分成10次转入我的MTGOX帐户，每次您转账完毕后，我查看我的MTGOX帐户无误后，即向您的支付宝帐户打入63.3元人民币。然后再进行第二次交易。。。 是的，我完全值得您信赖！ 联系我的QQ：821706149 （草根的力量） 太黑了。都比外汇牌价高的。  |
|
|
|
Based on the OP, (sorry I haven't re-read the whole thread recently), each FPGA has to dissipate less than 10W.
Are fans really necessary?
It seems that a pair of small passive heat sinks should do just fine.
Reducing both cost and total power consumption (admittedly minimally) and more importantly the noise level.
This is really a question, but there are some necessary reasons. XC6SLX150 's FGG484 package has a very high thermal resistance, the cooling system must keep the chip surface at a very low temp. so, a heatsink and a small fan is really needed. i also tried a larger heatsink without fan, but result is not very good. this function limit the power below approx. 5W. |
|
|
|
|
Show Posts
