Having extra pads for power kinda makes sense, but I wonder why they have dual I/O lines.
In general, number of pads doesn't have to match the number of pins. One could use a power interposer with one ruddy wire to the pin and many thinner wires to the pads. Maybe they were given a choice: QFN before New Year or TO-220 after New Year. Example custom packaging from the same site: http://www.psitechnologies.com/products/custom-packages.phpAs far as dual I/O: maybe they used some ready-made serial I/O blocks that operate in a dual ring, sort of like FDDI? There are pins named "bypass", which kinda suggests ring topology interconnect. |
|
|
|
There is nothing special abut ASIC, most ASIC vendors just use a custom programmed FPGA; this is called FPGA to ASIC conversion.
To get an FPGA/ASIC project of this scale done you will need 2 very good engineers forking full time for a year.
The original post has multiple false premises and therefore makes false conclusions. I'm going to address just the above two. Bitcoin hasher is a spectacular example where full-custom ASIC implementation will be much better than the FPGA implementation. SHA-2 is a rather rare digital circuit that is completely self-testable and observable. All the standard JTAG testing logic required in majority of digital circuits can be omitted. In fact vast majority of the internal D-type flip-flops in the hasher core don't even need the reset signal connected. Order of magnitude less power than FPGA will be easy. Because of self-testability of SHA-2 and repetativeness of brute force hasher the overall design could be done over a couple of lunch breaks by a single engineer familiar with mixed-signal design and with access to the appropriate software tools. In addition to the above the chip is almost completely solipsist: it really doesn't have to obey any well-known interfacing standard, not even with a second copy of itself. It is sufficient to just communicate between the hashing chip and the I/O controller. The "mixed-signal" is a key point here. Although the Bitcoin doesn't by itself use analog signals; the hashing chip is limited primarily by (1) power dissipation and (2) simultaneous switching noise. Because of the above two limitation mixed-signal experience would be a key to designing a chip that will be both efficient and will work on the first tapeout. The optimal package for bitcoin hasher would be something like TO-220 with 7 leads: http://www.psitechnologies.com/products/todo220.phpThe I/O would be serial, the leads would be VccI/O ClkI/O RxD TxD VccHash ClkHash and Reset. Ground would be provided by the heatsink screw pad. One could even omit reset lead by doing serial reset: hold RxD high over (say) 100 I/O clocks. Well, from the choice of packages (all with many more pins) one can surmise that none of the Bitcoin ASIC vendors obtained the advice from the power-analog and mixed-signal designers. I'm not familiar with the commercial toolchains used in ASIC development; but from my past experience with R&D in digital and mixed signal design I'm positive that the main stumbling block would be the learning curve required to understand and learn the tools required. This is a time-to-market or time-to-mine issue. pcm81 didn't make any manufacturing yield claims, but other people did. The Bitcoin hasher is so repetitive that if correctly designed, with a trivial set of clock-disable-bits, the overall yield would be nearly 100% useable chips. Only the chips with faults in the I/O or clock circuitry would have to be rejected. Some other people also made wild claims about testing effort and expense. Well, SHA-2 is essentially self-testing: it either fully works or fails nearly every test. There are no hidden states or data-conditional decision making in the algorithm. The test plan for the chip would be as trivial as it gets. The "millions of dollars" price tags for NRE are just flights of fancy. This really is a project that could be done by a single ASIC engineer over a series of lunch breaks provided that he has both access to and experience with the required toolchain. |
|
|
|
The packages were taken to the Post Office with a moped, they were small and didn't have that much of a dollar value attached, at least not the value Avalons carry.
Where is that "moped" part coming from? ngzhang has posted that he drives CIVIC (I presume Honda) and posted the over-the-dashboard shots of the traffic jam in Beijing.  https://bitcointalk.org/index.php?topic=51371.msg645006#msg645006Was the "moped" part just a putdown or did ngzhang really had to pawn his car to keep the business afloat? |
|
|
|
ok, I assume you've already explored the "open" alternatives such as Java Card which are easier to design for and want to go native for your projects. Then 1) is probably not the issue. I'd say less than 10k$ whatever the vendor. Often less than 5k$. 2) is more tricky. It's fine for a small shop if you can show references (preferred use case, but it might be a chicken and egg issue for you) or are extremely convincing, and in all cases can justify a business plan of a large amount (I'd say 1 million should be fine) of sold units for the next few years, most likely including a large pre-paid order. I'd recommend attending a specialized industry show such as Cartes where you can find all vendors and see who's the most interested in your project. Thank you very much. I was mostly courious if the situation had changed since the last time I researched it. Only costs went somewhat down, other than that things haven't changed much. |
|
|
|
- Not Open Source : smartcard chips and associated toolchains are heavily covered by NDAs, which makes sharing the source a pain. It's of course possible to release an Open Source smartcard wallet application running on an open smartcard platform, but you still need to trust the platform.
OK, lets assume that I have no reservations against signing NDAs and developing wholy closed source. Could you please answer the following questions: 1) What is the cost of one complete development seat for the target platform? 2) Would the toolchain vendor even be willing to consider NDA from an one or two person shop (and no personal assets pledged) as valid? |
|
|
|
No new information, I'm just quoting the offending posts against further deletions by authors, who may try to follow caveden's & jim618's actions. Have you considered adding support for a generic trusted path protocol so any server can send a message and get back an A/B answer in a secure manner?
Do you have any proposal how this should work? Actually device is able to sign bitcoin transactions (pay to address, pay to script hash) and sign bitcoin messages. Yes. Consider the case of online voting. The voting server wants to ensure the following a) The vote is cast by a human, not a virus. b) The vote is cast by each human only once. We can satisfy this use case by using Trezor like this: - Each Trezor has a unique public key (can be RSA) and a certificate signed by the manufacturers, ie, you and stick.
- When you start the voting process, the Trezor sends its certificate to the computer, which then sends it on to the voting server.
- The server checks that this certificate has not been seen before, and then encrypts/signs a protocol buffer that includes a message to display onscreen (any arbitrary text) and the two options available. That signed message gets sent to the Trezor, which shows it on screen.
- The users answer is then signed with the devices private key and uploaded to the voting server which checks it against the public key in the certificate.
In this way, the server can know that no virus interfered with the communication and there is no fake device at work. Of course, it assumes that the devices are somewhat secure against physical tampering. If you can already sign/verify text messages in the standard way, then I guess you could add it on top of that, though the need to communicate certificates and things makes it not really necessary to re-use something designed only for text. You don't have to do this now. As the software is open source, somebody could contribute such a protocol and then you incorporate it into your signed builds. Not content with:
+ redefining the nature of money, see bitcoinj
+ reimagining the meaning of property rights, see video presentation of Bitcoin London 2012
. . . Mike now goes on to reinvent representative democracy.
Please, for the love of God, can someone stop this guy before he redefines the very essence of reality and/ or space-time and we all become - well - the mind boggles.
:-)
yo jim, don't give him ideas!! |
|
|
|
The only people the trolling is good for is those who like setting stuff on fire— not because the stuff is good or bad— but just because they want to watch the world burn.
Actually there is one more class for whom trolling is good: Bitcoin sharks that prey on meek or emotional trolls. Sharks are very polite to trolls, learn about how to herd them and learn how to market to them. Then they just rip their faces off (financially speaking). Either in one single rip (like Pirate@40) or in a gentle, but close to the bone, shave (like Meni Rosenfeld). With this regard this forum is no different than the majority of financial and investment forums. In fact I always thought that this was always the purpose of the free-speech policy of the forum: a steady supply of people who will willingly take the losing side of the trade. Its just business, nothing personal. |
|
|
|
Thanks for the link. I'm somewhat familiar on why Intel would not include HRNG in their CPUs for many years. There are 3 sentences in the article about the reasons: Those analog circuits are also a nuisance when it comes time to improve the manufacturing technology used to make the processor. Every few years, chipmakers modify their fabrication lines to produce integrated circuits at a finer scale, allowing them to pack more transistors into the same area. Making these shifts is pretty straightforward for CMOS digital circuitry, but each new generation of analog circuitry requires careful reevaluation and testing—a major headache.
Essentially even normal routine process improvements would prevent them from generating unbiased bits. This completely perverts their internal performance evaluations: either continue produce low-yielding chips with good randomness or improve yields and ship chips generating weak randomness. The new generation system has sufficient bias-removal circuitry to properly self-test istself and fail with the machine check exception when the manufacturing gets improved too much. |
|
|
|
This thread has gone a bit off topic (Yes - I know I am guilty too).
Could we get it back to Trezor stuff please.
I support Jim in the request for thread-splitting. If possible please start with Mike Hearn's post: https://bitcointalk.org/index.php?topic=122438.msg1466565#msg1466565but exclude the post by commonancestor: https://bitcointalk.org/index.php?topic=122438.msg1468915#msg1468915and change the title to something like: Bitcoiners in support of voter itimidation and vote buying via crypto hardware. Geez, you think this is some sort of conspiracy or something? The Trezor represents an idea, not just a single device. Bitcoin is an experiment just like democracy was an experiment. The latter failed. Let's just give these little wide-eyed optimists the benefit of a doubt to play their own experiment out. You can sit by and make insults, but it doesn't really add anything except to the ignore files. Besides, there is nothing wrong with buying votes directly if you are not allowed to just print your money though monopolistic powers.
|
|
|
|
BFL has always and will likely continue to always respond, not instigate. If anyone is inclined to believe Inaba, here's the reminder how trolling was BFL company policy and not just his personal tactic: https://bitcointalk.org/index.php?topic=88363.0"The thread in which BFL trolls Enterpoint" was split and later preserved by locking by DiabloD3. |
|
|
|
<off-topic> In addition to the above the paper has a whole section entitled "Vote buying" that completely omits the discussion of buying the voting smartcards.
It does, it clearly states that such risk is irrelevant. Buying individual votes like this is much less efficient than buying politicians today. Plus, the keys could expire once each 5 years or something. Also, selling your vote is not much different than blindly delegating it, or blindly voting. Do you really expect somebody that easily sells his vote rights like that to even mind about voting at all? What you seem to be criticizing is democracy per se, not his particular instance. OBS: I'm not sure I support his idea, although I'm inclined to think it might be a good improvement on status quo. A more direct democracy might considerably decrease the effect of "dispersed costs, concentrated benefits", which are pretty much the norm in democracies. But it's still far away from full decentralization of power, which is the "nirvana", let's say. In summary: the whole paper meets the definition of ultra-naivette. May God bless the heart of its author.
Do you honestly believe that the current widely adopted representative democracy models are so superior? Buying politicians is pretty much the norm on them. I guess you're the naive one there if you think otherwise. </off-topic> I made a full quote to protect against further editing. Buying politicians became an expensive requirement when the previous strategy of buying votes (one vote -> one sausage and/or beer) became hard to enforce by combination of voting in public but casting the actual vote in secret. Now we are dscussing a supposed improvement where for the cost of one sausage buyer gets all future votes cast with particular token. I'm just going to quote some select parts from the Wikipedia's article on idiots: An idiot in Athenian democracy was someone who was characterized by self-centeredness and concerned almost exclusively with private—as opposed to public—affairs. [...] Declining to take part in public life, such as democratic government of the polis (city state), was considered dishonorable. "Idiots" were seen as having bad judgment in public and political matters. Over time, the term "idiot" shifted away from its original connotation of selfishness and came to refer to individuals with overall bad judgment–individuals who are "stupid".
I can muse a little about how IQ is not an one-dimensional quantity. An individual can be a competent software (Mike Hearn) or hardware (allten) designer; and yet completely fail the basic civic class material where grade-school children are taught why everyone is asked to vote in public, but the actual vote marking process is obscured and why nobody is allowed to accompany the voter while marking the vote; not even a husband can assist his voting wife (or wives). This central-bank-killing-liquid-democracy device may become the killer gadget for Occupiers world-wide.
Again: may God bless your little hearts. The brown-shirted functionaries of our party will allow you your continuous use of your knees just for the cost of letting them watch while you press the buttons on your trezor in the privacy of your house. If you drop your trezor in our party headquarters you will be paid one sausage for your kindness and cooperation in furthering the efficient democracy. |
|
|
|
Mike. That was a really neat article! Truly among great minds. I feel humbled. This was one of my motivations that got me started on the Bitsafe HW Wallet project. It's so nice to see it on paper. The whole paper is an expose on how the ultra-naive think about democracy. Here's a short quote: They can then walk away with their token, secure in the knowledge that it cannot be tied to their real identity.
Then they walk to the party headquarters where they exchange their voting token for one issue of election sausage. In addition to the above the paper has a whole section entitled "Vote buying" that completely omits the discussion of buying the voting smartcards. In summary: the whole paper meets the definition of ultra-naivette. May God bless the heart of its author. |
|
|
|
From my point of view something is "shipped" once it's scanned into the system of the company handling delivery (or picked up by their couriers). Sometimes this means that it's not actually on a plane going anywhere yet but it's past the point where your supplier has any control over the delivery process any more.
This is completely untrue for pretty much every bulk shipper for every carrier service. Appearance and allocation of package tracking number just means that the customer committed to pay for shipping. The tracking numbers are allocated online and the courier service can then reserve the resources in their logistic system. The actual packages can be stored inside a bulk container that will be opened and scanned only once it gets close to the destination. I don't have actual experience with China, but I do with US, Europe and certain Pacific Rim locales. The "package tracking" is done in bulk by "manifest tracking" and there are specific exception codes used to mark situations where the "manifest" ended up being different than the "packages" inside a bulk container. |
|
|
|
It would be a Sisyphean task to clean up all the BFL treads. It was BFL who started using the "diss" as a form of marketing. Live by the sword diss, die by the sword diss. And "diss" means not only "disrespect" but also "disinformation". Here's just one example of one of the original BFL ASIC threads that now somehow resides in the Off-topic section, and type of marketing practiced by BFL; I'm pointing to my quote of Inaba's conversation with reeses. https://bitcointalk.org/index.php?topic=83985.msg933012#msg933012Butterfly Labs appeared here first time in August of 2011 in the Russian News subforum and immediately aroused suspiction. https://bitcointalk.org/index.php?topic=38922.0Then it was you who actually opened the first English language topic: https://bitcointalk.org/index.php?topic=48863.01GH/s, 20w, $500 — Butterflylabs, is it a scam? and the second one: https://bitcointalk.org/index.php?topic=53530.01GH/s, 20w, $700 (was $500) — Butterflylabs, is it for real? (Part 2) I don't know, maybe you feel somehow responsible for all the BFL-flavoured beer that had been brewed on this forum and feel obliged to drink all of it. But as a reader I feel that BFL should drink their own beer. |
|
|
|
Can someone please explain "how do I move this topic"?
(actually maybe I'll just delete it as it looks as though it will take me more time to explain what the purpose of this topic is than I think I can be bothered with)
I'm sorry, I misunderstood your request. The "move topic" is somewhere on the bottom of the page on the left, next to the "lock topic". |
|
|
|
Oh - how?
start /b /belownormal c:\program files\bitcoin\bitcoin-qt.exe from the command line; or you can make a shortcut: c:\windows\system32\cmd.exe /c start /b /belownormal c:\program files\bitcoin\bitcoin-qt.exe |
|
|
|
In the GPU mining threads many people have shown that the way to get best hash/dollar is not by esotheric cooling but by underclocking and undervolting to meet the available thermal budget. So again, marketing to the OCD and newbie crowds. Oh, on by the way: BlackBox is a recognized brand name with long history in marketing network stuff to newbies. So anyone who feels themselves persuaded by their glossy marketing collateral should maybe adjust their self-awareness appropriately. I'm not going to generally diss BlackBox, because their marketing materials are top notch and may serve as an usefull reference for "maximum non-insane prices". |
|
|
|
* A UPS to condition the power, and keep everything from rebooting when grid power fails and the genset is still starting up. It's OK if the air conditioning is not UPS-backed.
What's the value proposition of that requirement? To me it looks like something to attract obsesive-compulsive nutbags or completely clueless newbies. |
|
|
|
Have you considered adding support for a generic trusted path protocol so any server can send a message and get back an A/B answer in a secure manner?
Do you have any proposal how this should work? Actually device is able to sign bitcoin transactions (pay to address, pay to script hash) and sign bitcoin messages. Yes. Consider the case of online voting. The voting server wants to ensure the following a) The vote is cast by a human, not a virus. b) The vote is cast by each human only once. We can satisfy this use case by using Trezor like this: - Each Trezor has a unique public key (can be RSA) and a certificate signed by the manufacturers, ie, you and stick.
- When you start the voting process, the Trezor sends its certificate to the computer, which then sends it on to the voting server.
- The server checks that this certificate has not been seen before, and then encrypts/signs a protocol buffer that includes a message to display onscreen (any arbitrary text) and the two options available. That signed message gets sent to the Trezor, which shows it on screen.
- The users answer is then signed with the devices private key and uploaded to the voting server which checks it against the public key in the certificate.
In this way, the server can know that no virus interfered with the communication and there is no fake device at work. Of course, it assumes that the devices are somewhat secure against physical tampering. If you can already sign/verify text messages in the standard way, then I guess you could add it on top of that, though the need to communicate certificates and things makes it not really necessary to re-use something designed only for text. You don't have to do this now. As the software is open source, somebody could contribute such a protocol and then you incorporate it into your signed builds. One trezor, one vote! Slush, if you are going to start a political party, remember that I was always your friend. Hopefully my trezors will have low serial numbers.  |
|
|
|
|
Show Posts
Actually that's an old interest. Here's a paper I wrote about 6 years ago:
