|
Will pay 2 BTC (shipped to SoCal) for a copy in good condition of Bitcoin Magazine issue #1.
Reply to this thread, or by email (m.bevand at gmail). No PM.
|
|
|
|
5 5970 is impossible because of 8 gpu limit, 7 gpu limit on 32bit os
Some people have PSUs powering cards across different machines. |
|
|
|
|
Typically, I ran either 4x5970 or 3x6990 per PSU. With memory downclocked on the cards.
Both configs draw about 970-1000W out of the PSU (which is ~90% efficient, so 1080-1110W at the wall).
I have never tried it, but you could go as far as 5x5970, as it would be just about 100% of its rated capacity, and the AX1200 is a well-engineered piece of hardware able to tolerate such a continuous load. But only do this if you know what you are doing, ie. if you have a clamp meter to measure current on the 12V rails to be sure you are not going over specs.
|
|
|
|
The information is apparently in Bitcoin Magazine, and subscribers have reported getting the magazine already. A quick readthrough and post is worth 40BTC.
So, who has received the magazine? |
|
|
|
I have a ton of computer gear for sale. - 20 x PSU Corsair AX1200 1200W (sold out)- 23 x Radeon 5970s capable of 650 Mhash/s at stock clocks (sold out)- 10 x Radeon 6990s capable of 700 Mhash/s at stock clocks (sold out)- 2 x Radeon 7950s capable of 420 Mhash/s at stock clocks (sold out)- I also have defective 5970s for sale in this thread (sold out)- 20 x Mobo MSI 890FXA-GD70 with 6 PCIe slots (5 x16 and 1 x1) (sold out)- I include the CPUs & Heatsink/Fan & RAM for free with each mobo (AMD Sempron 140 2.7GHz + 2GB DDR3-1066)Edit: sold 1 5970 on Nov 3. Edit: sold 5 5970s on Nov 3. Edit: sold 1 6990 on Nov 4. Edit: sold 2 5970s on Nov 4. Edit: sold 2 mobos on Nov 4. Edit: sold 9 6990s on Nov 5. Edit: sold 7 PSUs on Nov 5. Edit: sold 4 mobos on Nov 5. Edit: sold 1 mobo on Nov 5. Edit: sold 1 mobo on Nov 8. Edit: sold 1 5970 on Nov 8. Edit: sold 1 5970 on Nov 9. Edit: sold 2 mobos on Nov 10. Edit: sold 1 PSU on Nov 11. Edit: sold 4 5970s on Nov 13. Edit: sold 7 5970s on Nov 14. Edit: sold 2 5970s on Nov 15. Edit: sold 1 PSU on Nov 20. Edit: sold 1 mobo on Nov 20. Edit: sold 1 PSU on Nov 25. Edit: sold 2 PSUs on Nov 27. Edit: sold 1 PSU on Dec 5. Edit: sold 2 7950s on Dec 19. Edit: sold 1 mobo on Dec 20. Edit: sold 2 5970s on Jan 3. Edit: sold 1 mobo on Jan 4. Edit: sold 2 mobos on Jan 5. Edit: sold 1 mobo on Jan 7. Edit: sold 1 mobo on Jan 17. Edit: sold 1 mobo on Jan 18. Edit: sold 2 mobos on Jan 24. Edit: sold 2 PSUs on Jan 24. Edit: sold 1 PSU on Feb 7. Edit: sold 1 PSU on Mar 4. Edit: sold 1 PSU on Mar 31. Edit: sold 2 PSU on Apr 1. Edit: all hardware is now sold out. Shipping is extra. I ship from Los Angeles. I very much prefer taking bitcoins as payment. Paypal or Dwolla only if you are really trusted. I am an old member in this community. I have sold to multiple forum members. I am a verified Paypal user. And I have a good rating on bitcoin-otc: http://bitcoin-otc.com/viewratingdetail.php?nick=mrb_No PM, please email me: m.bevand at gmail.comI have all purchase receipts, ask me a copy when you buy, and I will give them to you. The warranty is in theory not transferrable, but manufacturers don't always care if the name on the purchase receipt does not match the name of the person exercising the warranty: - PSUs have been purchased on Newegg or Amazon between Dec 2010 and Jul 2011, and have a 7 year warranty - mobos have been purchased on Newegg between Dec 2010 and Jul 2011, and have a 3 year warranty |
|
|
|
|
Good news. But wake me up when Coinbase allows transfers of at least $1000/day.
|
|
|
|
yes I expect 150 TH to be shipped by the end of January.
I doubt it. BFL already admited being 2 months behind schedule in delivery the SC line (Oct -> Dec). |
|
|
|
It's not unknown that Nvidia cards don't even pay for their own electricity. I dont know why this would change just because you're mining with 18 thousand Nvidia GPUs.
What's the difference between mining with 1 and 18 thousand Nvidia GPUs? In the latter case, you are losing money 18 thousand times faster. Someone should tell davidspitzer that he does not really want that to be his mining rig =:) |
|
|
|
|
In theory it is possible to design an SSL accelerator vastly outperforming CPUs on RSA. But nobody made one because the market is just too small (<100 companies in the world need to do gazillions of SSL handshakes per sec). Also, the handshake itself is not that expensive as the connection can be re-used / kept open (eg. HTTP keep-alive) and the speed bottleneck usually becomes whatever else needs to be done over SSL (rendering a dynamic HTML page, performing a search in your mailbox, etc).
|
|
|
|
However you have also made a convenient assumption that it will utilise two usb ports to power it, allowing it to have twice as much power, for a max of 5 watts. That is abit of a stretch to assume that and why the math to me does not add up for it to do 3.5 Gh/s at 2.5W and is what I stated.
See Lethos, I was right. I correctly predicted 3 months ago that it would use 2 USB ports, and BFL confirmed it: The Jalapeno draws it's power from two USB port connectors. Both need to be connected to achieve the full 4.5 GH/s performance. (At the users discretion, a single USB connector can be used if it has a 2nd head split off as is common for use with laptop DVD drives).
|
|
|
|
(If not: does somebody have any idea what is it good for then?) They are good for "[enabling] organizations to cost-effectively improve Web server performance" http://www.tech.proact.co.uk/ncipher/ncipher_nfast_800_ssl_accelerator.htmSo if you have a web server that uses SSL, these cards help offload the cryptographic operations involved in encrypting traffic to the clients. In fact, they don't even help with this because modern CPUs are so fast that they outperform most SSL accelerators. nFast: 800 1024-bit RSA decryption/second according to http://www.asiapeak.com/download/nfast.pdfA quad-core Intel Core-2 3.0GHz does 7680 decryption/second (1920 per core), so it is 10x faster: $ openssl speed rsa
[...]
sign verify sign/s verify/s
rsa 1024 bits 0.000521s 0.000028s 1920.7 35894.9
I like to call them "SSL decelerators". They only made sense 10+ years ago when CPUs were 32-bit (modular exponentiation in RSA is much slower when done with 32-bit ints instead of 64-bit). |
|
|
|
ASIC's are not a natural technological generation leap like going from CPU to GPU was. ASIC's are simply specialized processing units made specifically for Bitcoin. Which i do not beleive is following the original intentions of Satoshi, for many reasons.
Satoshi was an intelligent man. He was well aware of the possibility of a hardware mining development arms race. Also one can make the following observation: There was only 1 GPU chip manufacturer who dominated mining farms: AMD. Today there are 2 FPGA chip manufacturers who are at the basis of all FPGA farms: Altera (in the BFL Single), and Xilinx (Spartan6-LX150 in all others). And soon there will be 3 vendors of ASIC mining chips: BFL, Tom's, and nghzang's companies. So we are actually becoming more decentralized over time. IMHO the number of owners of the IP (intellectual property) of the technologies at the core of Bitcoin mining is a better indicator of how resilient (which is what you mean by decentralized I think) Bitcoin mining can be, assuming hypothetical scenarios where hostile parties would try to take control of, or negatively affect, this technology. It is not the number of resellers or distributors that matters (like you point out there are many AMD graphics cards resellers). |
|
|
|
That would be 200m$ worth of BFL hardware... Not going to happen, leaving quite a margin for less efficient device to cover their investments ... at some point.
Are you one of these guys who think that "640kB ought to be enough for anyone"? If Bitcoin continue to succeed, it will come to a point where $200M will have been spent on mining hardware. Just consider that today alone, $20M has been spent on CPU/GPU/FGPA mining hardware. Going to $200M is only a 10x increase. |
|
|
|
its all a get rich scheme to run off with 1/10 of a penny from the community  No. Like the OP said, the address is from block #2. It very, very likely belongs to Satoshi Nakamoto himself, not to the OP. |
|
|
|
Probably 90nm. Seems to be the most cost effective.
Prepare for the improbable.  I believe my prediction of 65nm will be shown to be right |
|
|
|
|
SHA-256 is a one-way hashing algorithm, not encryption algorithm.
If you encrypt N bytes, you get N bytes out. And it is possible to decrypt these bytes back to the original value.
If you hash N bytes, you always get a fixed-size hash as output. And it is not possible to compute the original value from the hash (well, unless you bruteforce), hence the name "one-way hash".
|
|
|
|
hmm.. the real question is can i slip another 350-ish buck charge onto my paypal visa without the wife-to-be noticing... :-D
Looking for pseudo-anonymity? I heard about some newfangled cryptocurrency... |
|
|
|
Wow, Tom! I appreciate your generous offer of free money and I am absolutely delighted to take that bet from you with regards to the accuracy of our respective power estimates (I'll even send you a condolences cake with the proceeds).
However there's a few caveats, which I'm sure you'll agree do not alter the intent of the bet at all:
I will wager that what I've said is true about BFL's power efficiency if you wager that what you've said about your power efficiency is true as well. You've said that your power will be "competitive." To commence this bet, you will need to define "competitive." How many factor increase in power consumption is considered "competitive" against ours? Twice the power usage? Three times? Four? Is that still "competitive?" What happens if you're 6x as power hungry as our device? Do you lose the bet then? I want to know what you mean by competitive, and until you define that, well... I can't bet on it.
Honestly though, after you calling me out in your thread, calling me a liar about our power... I want to know why you aren't being honest and open about what you expect your power to be. It's simply not credible to say you don't know what your estimates are. Even if you really have no idea at all, how can you then promise it'll be competitive?
Also, the last caveat is that I will not stay out of your thread. You've been... shall we say... less than truthful about a number of things in your thread and I will be here to call BS when it's required. But you define competitive and release the stipulation that I stay out of your thread and you've got a bet, cowboy! I will absolutely wager 1000 BTC against your 1000 BTC any day of the week.
Let me put it this way. You refuse an easy 1000 BTC win (per your words) because you disagree on minor details like voluntarily leaving this thread? Wow. |
|
|
|
If I were to test a large batch of ASICs with that magnitude of hashrate, I would solo mine with a few different addresses so in lists like that it would just be attributed to "Unknown" and the addresses would be different.
Blockorigin usually reports about 12% of blocks found are unknown. This has been pretty consistent for months. for now yes. if one of ASIC companies will mining solo + tor / proxy has no way to find out which company increases the difficulty. right ? ASIC mining through TOR? Yeah, that would work well, suuuuure... It would work well. An ASIC miner would mine solo, so the numerous getwork RPC calls per second would not get routed over TOR. He could even connect to only one peer-to-peer node over TOR to further reduce traffic. |
|
|
|
|
Show Posts
