Topic: First GENUE LTC FPGA (Read 3376 times)
FPGA's kickstarted the arms race in SHA256, they'll do the same for sCrypt and fade into oblivion. Their high price vs even 10X better ratio won't be able let them stick their ground in sCrypt race beginning next year sometime. People are probably already planning ASIC's with on die memory, something FPGA's won't be getting, more so for 'salvaged' FPGA's
I am doubtful that the miner - even if real - is a big deal. The secret to high performance to threaten CPU and GPU mining I'd massive parallelization. Can this even be achieved on any FPGA that doesn't cost a royal fuckton?
I suspect his price/performance figures rely on sourcing very cheap components (eg reclaimed FPGAs from scrap). Good for him (assuming he's just building his own farm), but its not going to affect the overall network much.
I am doubtful that the miner - even if real - is a big deal. The secret to high performance to threaten CPU and GPU mining I'd massive parallelization. Can this even be achieved on any FPGA that doesn't cost a royal fuckton?
Yeah. I stalked up a bit more and I find this to be correct. Great scott, though. He needs help with not looking like a scammer. he looks like a perfect setup of someone who doesnt know shit about hardware trying to make people think he has a product that can mine LTC at astronomical rates so that they preorder. Just being paranoid. That's all.
Apologies.
I think he just doesn't give a shit about what others think about him.
When it will be ready the product will speak for itself.
Well, yeah...is not hard to be paranoid on bitcointalk, is a scammer paradise. I never seen such an scammers infested place before.
.............................
You show little literacy about hardware development and failed to answer a technical question that would not release any vital information about the supposedly privately commissioned product.
Honestly I think you sound a lot like a scammer. Videos don't provide any real proof as they can easily be faked and your grammar and word usage sounds like you are 14 years old.
What do you say?
You show little literacy about hardware development and failed to answer a technical question that would not release any vital information about the supposedly privately commissioned product.
Honestly I think you sound a lot like a scammer. Videos don't provide any real proof as they can easily be faked and your grammar and word usage sounds like you are 14 years old.
What do you say?
Not a scammer. This guy made his custom FPGA boards waay back, at the time we were mining with 2-3 GPU cards he had about 70GH or so in FPGA.
He doesn't talk much about his work, actually I am surprised.
As for the grammar and word usage, English is not his native language, so...
Yeah. I stalked up a bit more and I find this to be correct. Great scott, though. He needs help with not looking like a scammer. he looks like a perfect setup of someone who doesnt know shit about hardware trying to make people think he has a product that can mine LTC at astronomical rates so that they preorder. Just being paranoid. That's all.
Apologies.
.............................
You show little literacy about hardware development and failed to answer a technical question that would not release any vital information about the supposedly privately commissioned product.
Honestly I think you sound a lot like a scammer. Videos don't provide any real proof as they can easily be faked and your grammar and word usage sounds like you are 14 years old.
What do you say?
You show little literacy about hardware development and failed to answer a technical question that would not release any vital information about the supposedly privately commissioned product.
Honestly I think you sound a lot like a scammer. Videos don't provide any real proof as they can easily be faked and your grammar and word usage sounds like you are 14 years old.
What do you say?
Not a scammer. This guy made his custom FPGA boards waay back, at the time we were mining with 2-3 GPU cards he had about 70GH or so in FPGA.
He doesn't talk much about his work, actually I am surprised.
As for the grammar and word usage, English is not his native language, so...
FPGA's had been around before cryptocur and since they can be reprogrammed unlike ASIC's you could just sell them away, right?
Quite right, and there is a significant market in reclaimed FPGAs from scrap boards. Though as a seller you'll just be making pennies on the dollar. But its your decision, once bitcoin costs more in electricity to mine than its worth, do you sell your boards for scrap, or use them to mine litecoin? The electricity consumption is an order of magnitude less than GPUs, so that's not an issue. You won't make much return at just 15kHash/sec, but perhaps more than the scrap value, and it still helps the network to some extent.
I remember reading about another guy trying to get some FPGA to work on scrypt, it kinda looked like the spartan, doing 37 kH/s. And this, and also this scam http://scryptasic.org/ Something big is surely coming along *grinning*
I think that was my open source project (just using the bare FPGA, and NOT using external RAM). One of the contributors got it running on a pretty big altera chip at around that speed. The spartan LX150's are topping out at around 8kHash/sec and probably won't get very much faster than that. It just a hobby project really, but might find some use for the old bitcoin fpga miners rather than just scrapping them (only 15kHash/sec on a Lancelot though, so its not really that useful).
FPGA's had been around before cryptocur and since they can be reprogrammed unlike ASIC's you could just sell them away, right?
wow
I remember reading about another guy trying to get some FPGA to work on scrypt, it kinda looked like the spartan, doing 37 kH/s. And this, and also this scam http://scryptasic.org/ Something big is surely coming along *grinning*
I think that was my open source project (just using the bare FPGA, and NOT using external RAM). One of the contributors got it running on a pretty big altera chip at around that speed. The spartan LX150's are topping out at around 8kHash/sec and probably won't get very much faster than that. It just a hobby project really, but might find some use for the old bitcoin fpga miners rather than just scrapping them (only 15kHash/sec on a Lancelot though, so its not really that useful).
Being paranoid often pays off. First they laugh at you for being paranoid. But then they don't laugh. because they're dead, and you're not, because you were paranoid. 
I can only agree on this.
Stop agreeing with someone who is trying to oppose you. XD Thats not how it works.
Being paranoid often pays off. First they laugh at you for being paranoid. But then they don't laugh. because they're dead, and you're not, because you were paranoid.
I can only agree on this.
Well I'll be damned.
Weird stuff.
A google search query of the device ID's ("idVendor=0x1234" "idProduct=0x1010") yields many links to z.download.csdn.net where such information is referenced almost exactly, some kind of chinese internet forum. Apparently this has something to do with "STM32", which is a model of ARM Cortex MCUs. It seems to me the device id's referenced might be the ones of those devices.
Or maybe it's a generic idVendor and idProduct, but those were the only matches I could find.
Otherwise I could not find any correlation between what he posted and anything currently findable by google.
I have to say, it's a mystery. Maybe he does have an fpga miner, but it's single-threaded and does not resolve any of the scrypt fpga-mining issues that makes it so unprofitable so is more or less exactly worthless.
Weird stuff.
A google search query of the device ID's ("idVendor=0x1234" "idProduct=0x1010") yields many links to z.download.csdn.net where such information is referenced almost exactly, some kind of chinese internet forum. Apparently this has something to do with "STM32", which is a model of ARM Cortex MCUs. It seems to me the device id's referenced might be the ones of those devices.
Or maybe it's a generic idVendor and idProduct, but those were the only matches I could find.
Otherwise I could not find any correlation between what he posted and anything currently findable by google.
I have to say, it's a mystery. Maybe he does have an fpga miner, but it's single-threaded and does not resolve any of the scrypt fpga-mining issues that makes it so unprofitable so is more or less exactly worthless.
0x1234 was default in the usb stack I used. 0x1010 si for visual, I had many of this boards connected to a linux and I needed a way to quick recognize them in lsusb output.
But all this effort and you didn't notice that it uses a derivation of BFL SC communication protocol..
Well, I have not used BFL devices.
Either way, I'm not really jealous. If it's real, then fantastic, scrypt will have an even more secured future. Dedicated devices are fundamental for the long-term safety of a coin, I think.
I just think it all sounds like a big setup for a big scam.
Being paranoid often pays off. First they laugh at you for being paranoid. But then they don't laugh. because they're dead, and you're not, because you were paranoid.
I remember reading about another guy trying to get some FPGA to work on scrypt, it kinda looked like the spartan, doing 37 kH/s. And this, and also this scam http://scryptasic.org/ Something big is surely coming along *grinning*
you can get such hardware at 1$ - 1.5$ (closer to 1)
Is it the retail price for new components? No, I am doing it for money, so most information are classified till I sell for good reason
Why post it at all at this point?
I am looking to inform community about this project, while keeping sensitive information, which may affect project development if disclosed, private.
No, I am doing it for money, so most information are classified till I sell for good reason
Why post it at all at this point?
They`re just jealous ...keep on working on the project and when it will hit the market even the jealous ones will buy 
it look very ugly, when complete can you give better look?
It is designed the way it is designed for the same reason an AK47 it is designed the way it is designed.
All three rams are working at max possible speed. And will probably have another iteration, and this is not the only RAM board I have ready.
Well I'll be damned.
Weird stuff.
A google search query of the device ID's ("idVendor=0x1234" "idProduct=0x1010") yields many links to z.download.csdn.net where such information is referenced almost exactly, some kind of chinese internet forum. Apparently this has something to do with "STM32", which is a model of ARM Cortex MCUs. It seems to me the device id's referenced might be the ones of those devices.
Or maybe it's a generic idVendor and idProduct, but those were the only matches I could find.
Otherwise I could not find any correlation between what he posted and anything currently findable by google.
I have to say, it's a mystery. Maybe he does have an fpga miner, but it's single-threaded and does not resolve any of the scrypt fpga-mining issues that makes it so unprofitable so is more or less exactly worthless.
Weird stuff.
A google search query of the device ID's ("idVendor=0x1234" "idProduct=0x1010") yields many links to z.download.csdn.net where such information is referenced almost exactly, some kind of chinese internet forum. Apparently this has something to do with "STM32", which is a model of ARM Cortex MCUs. It seems to me the device id's referenced might be the ones of those devices.
Or maybe it's a generic idVendor and idProduct, but those were the only matches I could find.
Otherwise I could not find any correlation between what he posted and anything currently findable by google.
I have to say, it's a mystery. Maybe he does have an fpga miner, but it's single-threaded and does not resolve any of the scrypt fpga-mining issues that makes it so unprofitable so is more or less exactly worthless.
0x1234 was default in the usb stack I used. 0x1010 si for visual, I had many of this boards connected to a linux and I needed a way to quick recognize them in lsusb output.
But all this effort and you didn't notice that it uses a derivation of BFL SC communication protocol..
Jump to: