[AMD and Nvidia]
And would a public listed company risk investing heavily in crypto market, which faces regulatory challenges worldwide? It would an extremely adventurous move to say the least.
i wonder that also. its one thing to put out custom mining cards, but for either company to actually design a coin, figure a use for it (store of value, utility, whatever) its algos, and get it traded on exchanges seems pretty risky as now they are directly involved in "printing money" so to speak. thats way out of their normal business. maybe a spinoff company? dunno. and what if the coin bombs? just because amd or nvidia makes it doesnt mean it wont be just another shitcoin. might be a lot of r&d down the tubes. they might be able to (unofficially? but thats risky too if caught) support some devs of an existing or new coin. or be an advisor or something. i can see devs moving to algos that need 10 gb or more (1080ti class memory) in the short term but even that wont hold back asic/fpga for long, looks like its getting much easier to add large chunks of ram to fpga, the uxppr3 has 4 rdimm slots in it. plus its ability to take (the super expensive and stupid fast) qdrii+ memory. im still gonna risk a fpga. i must admit it will be mainly for the fun and to play with it. i dont really see it as a super weapon or major money maker though. but i dont see it as being super terrible, i would think it would eventually make its money back but years vs months? who knows. |
|
|
|
seems right now even "mining" cards are 95% stock video cards, just with no display connectors. everything else is the same except maybe the bios.
i wonder how far a REAL mining card could be stripped down if it didnt need to conform to the pcie spec physically and electrically. only needs a 1x pcie connector, a few pcie power connectors, a GPU, memory and few glue chips and thats it. so it could be much less expensive, fit into smaller (non standard mining only) cases, reach higher rates (tweaked memory and timings and controllers, say 512 bits wide or something [if that can be done i gather memory bus is baked into the gpu, but maybe bank switching or something can be done]). twice the compute power in half the space. would be sweet.
just need someone to step up and order bare gpu dies and go to town.
steal posts much? https://bitcointalk.org/index.php?topic=2786298.msg33011620#msg33011620seems right now even "mining" cards are 95% stock video cards, just with no display connectors. everything else is the same except maybe the bios.
i wonder how far a REAL mining card could be stripped down if it didnt need to conform to the pcie spec physically and electrically. only needs a 1x pcie connector, a few pcie power connectors, a GPU, memory and few glue chips and thats it. so it could be much less expensive, fit into smaller (non standard mining only) cases, reach higher rates (tweaked memory and timings and controllers, say 512 bits wide or something [if that can be done i gather memory bus is baked into the gpu, but maybe bank switching or something can be done]). twice the compute power in half the space. would be sweet.
just need someone to step up and order bare gpu dies and go to town.
reported EDIT: can a DT member paint him with some red trust please. looks like hes trying to meet a bounty post count. |
|
|
|
1500 watts (my self imposed limit, no reason i couldnt do more). its because i want the power efficiency and density that a fpga can give me. plus the versatility.
1.5 kw/h equivalent heat is already no small feat to dissipate... specially in mid range latitude in the summer  . thats one reason for my limit. currently my 7 card gpu rig draws 1100 watts and is in the basement. kept it nice down there last winter. this year im going to get one of those grow tents with powered in and out vents to the basement windows and run it in that for summers. |
|
|
|
Maybe u have too much guts lol. The crazy expensive part is the problem. It is more expensive and less efficient. Versatility just isnt worth it when mining difficulty will go to your coins anyways due to new and current asics. Asic manufacturers can flood the market like crazy. It wont be the first time.
Perhaps u can update on the total expected cost and hashrate of its best coin when u get the info.
guts? heh no not really. but i have no problem tossing 6 grand at an experiment that might bomb. that just makes me a somewhat stupid risk taker. at least if i lose this one im still alive, as opposed to some of my other hobbies. im betting on some coins adopting an anti asic stance. no sense in developing an asic for a coin you know they will change the algo for. difficulty is a known problem and always has been. its part of the game. we have dealt with it with varying degrees of success before and (hopefully) will continue to do so. we will see. i will give all data as to profit/loss etc when i am up and running. |
|
|
|
Summary. FPGA is not workable when one asic controls 64% of GPU mining market. Lets say u use FPGAs and they now control 20% of remaining market via 8 algos. One asic will appear and easily cut it to 15%. They will take the major coin away again... Your FPGA is basically an antminer S7 and when new asic appears, it gets "downgraded" to be an S5. More FPGAs/GPUs will also join in the crowd as u earn less. U may not even be able to be profitable once it gets downgraded to S5. Do note that difficulty increase does not lower profit, it lowers REVENUE. So, u may go negative on mining once it "downgrades to an S5". I strongly doubt it will work because the asics are far cheaper and too much more efficient and majority of the market is centered around a few big algos. This is not yet including the biggest problem which the E3 that can mine eth, which controls 64% of the GPU. It can keep pushing more GPU mining power to the coins u are mining. No point I keep repeating. I find it conceptually just a horrible idea. Hope to see transparency in numbers and figures to brave souls those who attempt this project. Good luck people you keep harping on the e3, sha256 and other currently asic dominated stuff. those are done deals, no ones gonna use a fpga to mine those (or at least long term) thats not why we (or i) would go with this. there are enough non asic algos to make it profitable. part of the fpga strategy is using its reprogramability to move among different algos. basically what gpu miner do now. also i envision we could, if needed, put bounty programs up to get new bitstreams up faster. |
|
|
|
============================== What it is already CHECKMATE FOR GPU/FPGA ============================== U just dont realize it. https://www.reddit.com/r/ethereum/comments/7wufiq/ethereum_is_64_of_all_gpu_mining_by_power_09gw_of/According to this thread, ETH is 64% of all GPU mining power. I always thought it was above 50%. That being said, the antminer E3 is actually checkmate for GPU already. If they just push it down to 800 USD again. Sales of E3 will skyrocket leading to ETH difficulty skyrocket. GPUs will move out of ETH to other coins. Difficulty of other coins will rise too. GPUs will be wiped out and probably sold en mass cheap on ebay. U see, they can manufacture it at too cheaply for gpus to compete. U are actually not competing with your asic competitor with your FPGA. U are at their mercy. I hope u realize this. I am talking about existing asics and not them having to come up with a stronger one yet. FPGA can only win if lets say over 50% of GPU mining power is shared among say 100 crypto coins. Ahhh and then, u would wanna use FPGA because it is too difficult to create 100 different asics to mine 100 different coins. However, that isnt the case in crypto world. Eth holds 64% lol and a few major coins hold most of the other %. That is why crypto manufacturers mostly build Asics. They arent stupid. With 5 asics appearing so far this year to take down 5 gpu coins, it seems they can create a new asic every 30 days now. Thus, it is possible that 90+% of future coins are asic coins. Now, do wise up and stop pointing that small handgun at the tanks and bazookas. U aint gonna win. Oh and u are welcome buddy P.S> U may even wanna consider shorting nvdia and amd stock in july-august before earnings lol. The downdraft should show up by then. Just my 2 cents. you make some good points. i do think you are not taking some things into consideration though, and thats the versatility and longevity of something like this. i will be buying a xupp3r (although i will use ddr400 memory at 1st, as the qdrii+ is crazy expensive). why? its not to compete with asics, you cant. i own none and dont intend to buy any, im GPU only with a power budget of 1500 watts (my self imposed limit, no reason i couldnt do more). its because i want the power efficiency and density that a fpga can give me. plus the versatility. sure any current or new algo can have an asic built, but it will take time. a fpga is more nimble, load a new bitsteam and go. for that period of time of no asic for the algo, the fpga will rule compared to gpus. also there is no reason a xupp3r cant last for years and still be pretty competitive even against other fpgas. sure new ones may come out that are more energy efficient but not by orders of magnitude. as long as it has the capacity to fit the bitsteam it will be useful. thats why i figure ill go with the xupp3r as it seems the most advanced one of those being discussed here, and it has the most connectivity (link up to 4 of them for any algo thats too large to fit in one unit). plus, i like playing with new tech. give me a reason to learn programming (although i will never be good enough to bang out algos) it will be fun just to play around. what can i say, i like shiny new toys. will i lose money on it, possibly. but thats the chance ill take, and if i do lose, no regerts. for me the tech is the most fun. if it loses money, well im sure i could find another use for it, at least something better than a doorstop anyway. |
|
|
|
|
got a quote from bittmex about the xupp3r. about $5900USD with its development kit but with no memory.
and also a price a for xupp3r with development kit and two of the qdrii+ 288 mb modules. about $12500 USD. ouch. the qdrii+ memory may not even help much with algos that are mostly compute heavy, so im wondering if its worth it. have to try and get some numbers for eth.
|
|
|
|
|
prolonging life would be merely OK.
now REVERSING some aspects of age would be the brass ring.
after all living for another 200 years in a 90 years old decrepit body is one thing, but 200 more years in a body with an effective age of 50 is even better.
but ill settle for either when you get right down to it.
|
|
|
|
None of that mate, telegram is account based, not number. And Discord has a public ID where your email is always hidden.
ok will give it a shot later. bunch of other stuff on the table for today though. |
|
|
|
Could't we open a telegram/discord for this? there is so much to ask, answer and share...
Let me know if there are at least 3/4 ppl interested i would be in
im in. never used discord/telegram before but im sure i can figure it out. Just created, give it a try https://discord.gg/DNBH9bwdiscord need voice? eeeeew.no thanks, lemme try telegram EDIT and telegram wants my phone number. bleh. maybe later. |
|
|
|
Could't we open a telegram/discord for this? there is so much to ask, answer and share...
Let me know if there are at least 3/4 ppl interested i would be in
im in. never used discord/telegram before but im sure i can figure it out. |
|
|
|
The intro price (at Avnet) on the VCU1525 is $3995 USD, but it will be going up to around $5K in July. The Bittware XUPP3R-VU9P crypto version is $5895 USD, and has two advantages over the VCU1525: (1) it has four QSFP28 100G ports so you can daisy chain 4 FPGA's together to mine Xevan at 162MH/s, and (2) it has flexible memory options, so you can install either DDR4 or QDRII+ SRAM; the QDR memory gives way faster hash rates on Equihash vs. DDR4. I've been in communication with many members of this forum who are already organizing a group buy.
this difference between the vcu1525 and the XUPP3R-VU9P is mainly the better memory option. how much of a difference is this expected to be? im am so down with this project, i love the fact i can just drop it into my onda board with no mods.. the "just anther videocard" format has me hooked. |
|
|
|
Hey Everyone... as some of you may remember, I have been having issues with GPU 0 on all 4 of my 6x470 rigs. Pretty sure the W10 Fall Creator's Update was the culprit (for some reason i could not stop this from installing on all my rigs), and have tried various versions of Claymore ETHash & Cryptonite-7 miners but always GPU 0 ends up going "open CL call fail restart"... over and over again.
Since I've been out of town most of April, I just turned GPU 0 off mining on all 4 rigs but now I am back and looking to get those 4 running again.
My first try is going to be a driver update, so I wanted to ask the experts on this thread which AMD drivers you are currently using or find to be the best/most stable.
I'm using 17.11.2 on all rigs which is quite old, so hoping a more recent driver will work better and fix that GPU 0 issue.
using 17.11.2 here on my onda rig (mix of 470s and 1070/1080s, 7 cards total). seems super stable, rig stays up for months at a time. using cdm 11.4 on win 10 pro x64 ver 1709. 8 gb ram, 60 gig page file. all updates have been disabled (and so far its worked great, no updates at all since the update services were disabled. about 6 months maybe. i installed 1709 manually. i have a kvm plugged into the iGPU (G4400), so all gpus have nothing plugged into them. sure makes driver updates easier. dunno if you use risers (onda is riserless) but that would be my 1st guess but im sure you already know that. and it happening on 4 rigs at the same times kinda rules that out. any cards bios modded? used pixel patcher to get them recognized? what mobo? any bios updates for it? any changes to the bios been done (pci gen mumber changes, 4g decode etc. even iGPU vs pci video init order changed? my rig is mixed so all amd may have something to do with it? mine has a 470 in slot 0 (closest to cpu) then another 470, 460, 1070, 1070ti then 2x1080tis |
|
|
|
It is necessary to take into account that this thread was opened in 2011, when the price reached the great record of $20, but in general it kept oscillating below $2. I think that when OP opened this thread, he never imagined that the bitcoin would reach the price of $20,000, since his biggest wish was to see the price of $10,000 someday.
Lucky all those who had the great vision of OP when it was possible to acquire many bitcoins for a few dollars, because today they deserve the great wealth they are enjoying.
i joined the forum in 2011 and saw this tread back then. thought the dude was batshit crazy. bit later i remember selling a bunch at $35 USD (which i though was the highest it would probably ever hit) and thinking i was on top of the world. went on a very nice vacation with the wife, all because of "magic internetz money." now i am about to retire early, mainly because of btc. what a ride its been! and the real ride is probably just starting really. |
|
|
|
It will be too hard to get these but undoubtedly this is a great GPU.
turns out you can now basically more or less turn a stock 1080ti into a p102-100. fully reversible, no bios mods needed. just reboot and its back to a regular 1080ti. 50+mhs on eth. https://bitcointalk.org/index.php?topic=3370685.0im using it on my two 1080tis with CDM 11.4 and it works, but its closed source, so its safety (as far as hidden malware etc) is questionable. i run it on a locked down miner though. dyor as to if you want to run it. i have no affiliation with them btw. |
|
|
|
|
if people are using sane settings (voltages and cooling) plus quality PSUs the card will be far obsolete before any damage to caps and such happen.
fans? meh. make sure they are ball bearing. and add external fans if needed. keep cards below 70C and thats not just the gpu chip there are vrms that need to be cooled also (call it 80C for them). the better cards also monitor memory temps. not sure what the highest recommended is but i would guess 70C for them.
people are overthinking this. relax. electronics like stable temps and stable voltages more than anything else.
|
|
|
|
Amateurs.
I'm HODLing to buy entire Solar Systems at a time.
I call dibs on HD 85512.
cool by me, i might want to rent vacation planets on occasion. |
|
|
|
Next bull run, people talk about what helicopter on their yacht’s flight deck.
And the one after that people will talk about booking their first space rocket ship tourist flight. stop thinking small. owning planets are where its at. i want Saturn. lots of moons to build on. plus, you know, the rings are sooo cool. That's not thinking big. Planetary inflation is too high. There are zillions of planets in the universe and just 17mn bitcoins.... 
You're not thinking big enough. Pay to have a new planet custom built to your specifications. Why buy an existing old one when you can have a new one built exactly the way you want it? You would get even cooler rings than Saturn's with more moons.
good points. i need to adjust my thinking. ok i will go for all the ringed planets in known space. then ill branch out to others once theyre all mine. but. as mentioned, might be faster to just buy/build plain planets and build my own ring systems for them. hmmmm. |
|
|
|
Next bull run, people talk about what helicopter on their yacht’s flight deck.
And the one after that people will talk about booking their first space rocket ship tourist flight. stop thinking small. owning planets are where its at. i want Saturn. lots of moons to build on. plus, you know, the rings are sooo cool. |
|
|
|
|
pretty sweet i must say.
generally i dislike running unvetted software but this was too hard to pass up.
hash is in the low 50's on my 2 1080tis on cdm 11.6 11.4. still tweaking. nice.
sent a ltc donation, tx in pm.
|
|
|
|
|
Show Posts
