Bitcoin Forum
June 21, 2018, 11:27:17 AM *
News: Latest stable version of Bitcoin Core: 0.16.1  [Torrent]. (New!)
 
   Home   Help Search Donate Login Register  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 [40] 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 ... 107 »
  Print  
Author Topic: PhoenixMiner 3.0c: fastest Ethereum/Ethash miner with lowest devfee (Windows)  (Read 64014 times)
Bigdrago
Jr. Member
*
Offline Offline

Activity: 182
Merit: 0


View Profile
March 01, 2018, 09:58:32 AM
 #781

I disabled fee in Claymore. Lost 1Mh/s on each gpu.

So it' s worth to pay the fee.
how come 1% is more than 1Mh/s ? what is wrong with you?

I pay 1% fee.
If I disable fee I loose about 4% hashrate (from 30.5 to 29.3. Dualmining).

I havent tried phoenixminer though.
1529580437
Hero Member
*
Offline Offline

Posts: 1529580437

View Profile Personal Message (Offline)

Ignore
1529580437
Reply with quote  #2

1529580437
Report to moderator
1529580437
Hero Member
*
Offline Offline

Posts: 1529580437

View Profile Personal Message (Offline)

Ignore
1529580437
Reply with quote  #2

1529580437
Report to moderator
1529580437
Hero Member
*
Offline Offline

Posts: 1529580437

View Profile Personal Message (Offline)

Ignore
1529580437
Reply with quote  #2

1529580437
Report to moderator
The World's Betting Exchange

Bet with play money. Win real Bitcoin. 5BTC Prize Fund for World Cup 2018.

Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1529580437
Hero Member
*
Offline Offline

Posts: 1529580437

View Profile Personal Message (Offline)

Ignore
1529580437
Reply with quote  #2

1529580437
Report to moderator
Wolf0
Legendary
*
Offline Offline

Activity: 1834
Merit: 1002


Miner Developer


View Profile
March 01, 2018, 10:00:11 AM
 #782

@PhoenixMiner:

you use my GPU kernels in your "fastest" miner, 100% match of binaries. I can prove it if you are not going to confirm it.
So do you confirm it?
   No, we certainly DO NOT confirm it. We believe that you've tried (and apparently you think you succeeded) in extracting our kernels using reverse engineering techniques. That wasn't very nice of you to say the least but we have expected something like this, just not that fast. In some sense it is flattering but also a little annoying.

   The first version of our kernels was obtained from someone on this very forum for quite a bit of coins. They were a little bit slower than the best miner on the market at this time - yours - but we decided to use them to jump start our project anyway. After a lot of tinkering with Polaris GPUs we found out how to increase the speed a bit and improved the kernels a lot, achieving higher speed and/or lower consumption than your miner (in most cases). The original versions were left as a honeypot for particularly persistent reverses which apparently did its job quite well Smiley

   Now, are these old kernels the same as old versions of yours? We don't know for sure but we doubt it, and we can't just take your word for it (especially after throwing accusations like this after basically admitting that YOU are trying to reverse OUR miner), and because they were slower than your miner, and they do not support dual mining at all (and dual mining is still the killer feature). But even if they were, frankly this isn't our problem - we've paid handsomly for them, and we haven't used them in any of the publicly released versions, so we won't lose any sleep over this. If you beleive so, search for a digruntlet employee, or perhaps your unnamed source(s) at AMD?

I decided to check your miner as soon as I noticed that you have same tuning option ("+" and "-" keys in runtime) that works exactly as my -dcri option, also your miner has no "optimized" kernels for same set of chips as my miner. After checking I found 100% match in GPU kernel on GPU.
Now you are talking about some "old unknown" kernels and you don't use them anymore and have super-protection that can fool me and so on... okay.
I will create step-by-step guide how to confirm that you send my kernels to GPU and you will say that your miner uses "old unknown" kernels because it detects dumping but still mines at the same speed. That's funny answer, but you don't have better one, I understand.
So my word against yours. The result is obvious for me, however, since I call you a liar, anyone who wants to confirm by yourself that this miner uses my kernels can PM me and I will send step-by-step guide how to do it, then you can say the results; no newbies please, be at least sr.member on this forum, just to be sure that you don't want to become another "phoenixminer" and spend my time again...

Do what you've got to do of course - obviously you feel strongly about this and believe that this is well worth your time Smiley It's just not exactly clear what you are trying to prove though - we already conceded that our old/honeypot kernels might be the same as your old kernels but this is not very likely. But since you are willing to provide the blobs of your kernels in these PM sessions, our kernel developer guy would certainly like to take a look, so thank you.

Frankly we've got a lot of respect for you, and if you have chosen to contact us via PM with your concerns, we would have respected that and checked if there is any merit in your claims but you've come swinging and accusing us in our thread and if you expect us to just sit there and take it, you are quite wrong.

It is also quite wrong of you to assume these things:
  • That you can't be fooled by our anti-reversing measures (really do you think that we can't detect dll-injection attacks?).
  • And that, if you do succeed in extracting our real/current kernels and leak them in the open, we can't do the same with yours. "We know" a guy that was part of the driver development team of AMD and he is itching to test its kernel-level emulator by trying to extract your current kernels. Heck, if he's successful, we can even put your kernels in our miner and let the users select them explicitly and see for themselves how they compare with ours.  Grin

Anyway, we prefer to concentrate our efforts on improving our miner bit by bit (we are still lacking in many areas) but if you want to play this game, we can play it too. Just keep in mind that we were nobody in comparison with you just a few weeks ago, and you seem resolute to put us on the map so to speak, for which we thank you - it well known that controversial or bad publicity is always better than no publicity at all.  Grin

First off - I'm not accusing you. I said I might look into it - and I found it unlikely. Secondly, sure, I can be fooled. But you can't lie to the device you're executing on. DLL injection? You must be joking. I spend a good amount of time on static analysis before I even bother with runtime. Fact is - the GPU has to have the real thing loaded at some point to it. I'm not going to leak your kernels - real or not. I could have done this to Claymore - but it would be an exercise in aimless cruelty for no reason. I'd be very interested to know about this contact of yours, as they seem to perhaps be a bit flaky: AMD, as far as I am aware, has none of my kernels - not any near current. As a matter of fact, there's extremely few who would... I'm interested, and actually hopeful these ARE your own kernels. I enjoy a bit of competition, even though I've not done a full, clean-room miner which is anywhere near in shape for release in the way yours is. Gives you a bit of advantage, actually - the code in the open source ones have some pretty nasty network/other issues.

By the way - you weren't nobody to me a few weeks ago. I've been poking my head in semi-regularly since someone linked me your thread. :3

Edit: Just noticed your comment about blobs in PM... I've not done this. For... well, any and all recent memory, years. Not for Ethash.

Code:
Donations: BTC: 1WoLFdwcfNEg64fTYsX1P25KUzzSjtEZC -- XMR: 45SLUTzk7UXYHmzJ7bFN6FPfzTusdUVAZjPRgmEDw7G3SeimWM2kCdnDQXwDBYGUWaBtZNgjYtEYA22aMQT4t8KfU3vHLHG
peterboy1
Jr. Member
*
Offline Offline

Activity: 112
Merit: 0


View Profile
March 01, 2018, 10:24:59 AM
 #783

wait... w0lf is claymore?
Wolf0
Legendary
*
Offline Offline

Activity: 1834
Merit: 1002


Miner Developer


View Profile
March 01, 2018, 10:30:10 AM
 #784

wait... w0lf is claymore?

How do you manage to get that far off? I'm a third party - who doesn't have a stake in it, which is why I thought I might look and give my input.

Code:
Donations: BTC: 1WoLFdwcfNEg64fTYsX1P25KUzzSjtEZC -- XMR: 45SLUTzk7UXYHmzJ7bFN6FPfzTusdUVAZjPRgmEDw7G3SeimWM2kCdnDQXwDBYGUWaBtZNgjYtEYA22aMQT4t8KfU3vHLHG
wartywarlock2
Newbie
*
Offline Offline

Activity: 3
Merit: 0


View Profile
March 01, 2018, 10:39:17 AM
 #785

Quote
How do you manage to get that far off? I'm a third party - who doesn't have a stake in it, which is why I thought I might look and give my input.
Because you just replied to a post directed at Claymore acting like it was at you. Easy enough conclusion, if far fetched in reality.

Quote
It's definitely not a hardware problem - this is the anti-debug code activating without good reason apparently. We are trying to make these checks smarter and lighter but obviously we aren't there yet. Still, we are hesitant to relax the checks too much for obvious reasons - apparently there is no lack of highly esteemed developers trying to reverse our miner Wink

No worries, as I said the only difference I can give is that I used the latest ISO for Win10 from MS, but then all the other machines are fully up to date AFAIK, and we have 2 other builds with the exact same hardware (built from ISO from around June 17). The machine itself has crashed/rebooted a couple of times over the last day so maybe it's a problem with the mobo. Will visit the mine later and see about copying bios settings from one of the fully working ones incase some arcane setting balls deep in the menus got missed!
Wolf0
Legendary
*
Offline Offline

Activity: 1834
Merit: 1002


Miner Developer


View Profile
March 01, 2018, 10:52:04 AM
 #786

Quote
How do you manage to get that far off? I'm a third party - who doesn't have a stake in it, which is why I thought I might look and give my input.
Because you just replied to a post directed at Claymore acting like it was at you. Easy enough conclusion, if far fetched in reality.

Quote
It's definitely not a hardware problem - this is the anti-debug code activating without good reason apparently. We are trying to make these checks smarter and lighter but obviously we aren't there yet. Still, we are hesitant to relax the checks too much for obvious reasons - apparently there is no lack of highly esteemed developers trying to reverse our miner Wink

No worries, as I said the only difference I can give is that I used the latest ISO for Win10 from MS, but then all the other machines are fully up to date AFAIK, and we have 2 other builds with the exact same hardware (built from ISO from around June 17). The machine itself has crashed/rebooted a couple of times over the last day so maybe it's a problem with the mobo. Will visit the mine later and see about copying bios settings from one of the fully working ones incase some arcane setting balls deep in the menus got missed!

Oh. Shit. Oops.

Code:
Donations: BTC: 1WoLFdwcfNEg64fTYsX1P25KUzzSjtEZC -- XMR: 45SLUTzk7UXYHmzJ7bFN6FPfzTusdUVAZjPRgmEDw7G3SeimWM2kCdnDQXwDBYGUWaBtZNgjYtEYA22aMQT4t8KfU3vHLHG
Rewqpro
Jr. Member
*
Offline Offline

Activity: 51
Merit: 0


View Profile
March 01, 2018, 11:12:08 AM
 #787

I disabled fee in Claymore. Lost 1Mh/s on each gpu.

So it' s worth to pay the fee.
how come 1% is more than 1Mh/s ? what is wrong with you?

Can't you count? The 1MH/s loss scales too. That is, let's say he has 1k GPUs, at 31MH/s per:

With the fee on, his "effective" hashrate is 31 * 1000 * 0.99 = 30,690MH/s.
With the nofee switch and the inability to modify the code to disable the slowdown: 30 * 1000 = 30,000MH/s.

Pretty clear to me.

oh yeah sorry, I misread the message.

YES, my point was paying fee is better.

What a drama coming!

For those who wants Claymore to lower dev fee - suck it up dudes, he makes constant developments of miner things and he's the only one who always deliver.

If you are so greedy and for you 0.35% makes a huge difference (sick joke, seriously), it would be better to throw yourselves out of the window.

I am 100% sure those who have industrial mines have their own miner software and not have any problem with comission.

So, most of baby cryers outta here have 1 rig with 300$ monthly income -- 3$ per month for actually MAKING this income possible -- it is too much for you guys?




Coming back to reverse engineering thing, I do not appreciate such kind of actions. Hope it has nothing with stealing Claymore's code.

Have a nice day everyone!

I have significant hashpower and believe me, 0.35% makes a substantial difference at current prices

it is your problem you cannot inquire someone about your private miner and have a stable hashrate with no fees.


headshot155
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
March 01, 2018, 11:52:44 AM
 #788

I disabled fee in Claymore. Lost 1Mh/s on each gpu.

So it' s worth to pay the fee.
how come 1% is more than 1Mh/s ? what is wrong with you?

Can't you count? The 1MH/s loss scales too. That is, let's say he has 1k GPUs, at 31MH/s per:

With the fee on, his "effective" hashrate is 31 * 1000 * 0.99 = 30,690MH/s.
With the nofee switch and the inability to modify the code to disable the slowdown: 30 * 1000 = 30,000MH/s.

Pretty clear to me.

oh yeah sorry, I misread the message.

YES, my point was paying fee is better.

What a drama coming!

For those who wants Claymore to lower dev fee - suck it up dudes, he makes constant developments of miner things and he's the only one who always deliver.

If you are so greedy and for you 0.35% makes a huge difference (sick joke, seriously), it would be better to throw yourselves out of the window.

I am 100% sure those who have industrial mines have their own miner software and not have any problem with comission.

So, most of baby cryers outta here have 1 rig with 300$ monthly income -- 3$ per month for actually MAKING this income possible -- it is too much for you guys?




Coming back to reverse engineering thing, I do not appreciate such kind of actions. Hope it has nothing with stealing Claymore's code.

Have a nice day everyone!

I have significant hashpower and believe me, 0.35% makes a substantial difference at current prices

it is your problem you cannot inquire someone about your private miner and have a stable hashrate with no fees.



you misunderstand me - 0.35% is the difference between Phoenix Miner and Claymore - that difference IS significant
Rewqpro
Jr. Member
*
Offline Offline

Activity: 51
Merit: 0


View Profile
March 01, 2018, 11:57:02 AM
 #789

I disabled fee in Claymore. Lost 1Mh/s on each gpu.

So it' s worth to pay the fee.
how come 1% is more than 1Mh/s ? what is wrong with you?

Can't you count? The 1MH/s loss scales too. That is, let's say he has 1k GPUs, at 31MH/s per:

With the fee on, his "effective" hashrate is 31 * 1000 * 0.99 = 30,690MH/s.
With the nofee switch and the inability to modify the code to disable the slowdown: 30 * 1000 = 30,000MH/s.

Pretty clear to me.

oh yeah sorry, I misread the message.

YES, my point was paying fee is better.

What a drama coming!

For those who wants Claymore to lower dev fee - suck it up dudes, he makes constant developments of miner things and he's the only one who always deliver.

If you are so greedy and for you 0.35% makes a huge difference (sick joke, seriously), it would be better to throw yourselves out of the window.

I am 100% sure those who have industrial mines have their own miner software and not have any problem with comission.

So, most of baby cryers outta here have 1 rig with 300$ monthly income -- 3$ per month for actually MAKING this income possible -- it is too much for you guys?




Coming back to reverse engineering thing, I do not appreciate such kind of actions. Hope it has nothing with stealing Claymore's code.

Have a nice day everyone!

I have significant hashpower and believe me, 0.35% makes a substantial difference at current prices

it is your problem you cannot inquire someone about your private miner and have a stable hashrate with no fees.



you misunderstand me - 0.35% is the difference between Phoenix Miner and Claymore - that difference IS significant

Here I have completely understood your point, you have unlimited hashpower - blah blah.

But in the case you have zillion hashes per second, why aren't you able to buy a private miner for your needs and remove fee, approximately, forever?

So, your words are just blabbering to me.
headshot155
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
March 01, 2018, 12:07:01 PM
 #790

I disabled fee in Claymore. Lost 1Mh/s on each gpu.

So it' s worth to pay the fee.
how come 1% is more than 1Mh/s ? what is wrong with you?

Can't you count? The 1MH/s loss scales too. That is, let's say he has 1k GPUs, at 31MH/s per:

With the fee on, his "effective" hashrate is 31 * 1000 * 0.99 = 30,690MH/s.
With the nofee switch and the inability to modify the code to disable the slowdown: 30 * 1000 = 30,000MH/s.

Pretty clear to me.

oh yeah sorry, I misread the message.

YES, my point was paying fee is better.

What a drama coming!

For those who wants Claymore to lower dev fee - suck it up dudes, he makes constant developments of miner things and he's the only one who always deliver.

If you are so greedy and for you 0.35% makes a huge difference (sick joke, seriously), it would be better to throw yourselves out of the window.

I am 100% sure those who have industrial mines have their own miner software and not have any problem with comission.

So, most of baby cryers outta here have 1 rig with 300$ monthly income -- 3$ per month for actually MAKING this income possible -- it is too much for you guys?




Coming back to reverse engineering thing, I do not appreciate such kind of actions. Hope it has nothing with stealing Claymore's code.

Have a nice day everyone!

I have significant hashpower and believe me, 0.35% makes a substantial difference at current prices

it is your problem you cannot inquire someone about your private miner and have a stable hashrate with no fees.



you misunderstand me - 0.35% is the difference between Phoenix Miner and Claymore - that difference IS significant

Here I have completely understood your point, you have unlimited hashpower - blah blah.

But in the case you have zillion hashes per second, why aren't you able to buy a private miner for your needs and remove fee, approximately, forever?

So, your words are just blabbering to me.


wow, why the agression. I've never claimed to have unlimited hash power - just that 0.35% is significant
wtfonly16
Full Member
***
Offline Offline

Activity: 253
Merit: 100


View Profile
March 01, 2018, 01:18:49 PM
 #791

when ogodaoffset release for free?? we need linux undervolt!@!!!

ye i aint bares
Claymore
Donator
Legendary
*
Offline Offline

Activity: 1302
Merit: 1207

Miners developer


View Profile
March 01, 2018, 01:38:27 PM
 #792

I may look into it, just to check. While they have only a Win version, where I'm less practiced, I don't need nor want to go through the logic itself - location and extraction of the binary GPU kernel will tell me. Even if they based it on yours and made modifications - certain things in it stand out, so it shouldn't be ambiguous at all.
While I *know* it's quite possible to do better than Claymore's kernels - recently, he's made improvements resulting in them being quite a bit better, which cuts some of my own advantage. The optimizations in regards to core computations I've done result in me requiring far lower core clocks for the same hashrates, so I'll take the power savings. Despite this, I don't see Claymore blatantly lying about something so easily proved.

It would be nice if you can confirm my words. No modifications at all, the kernels are the same, taken from v10. I know you dumped my binaries for v10 so you can compare them easily.

It is also quite wrong of you to assume these things:
  • That you can't be fooled by our anti-reversing measures (really do you think that we can't detect dll-injection attacks?).
  • And that, if you do succeed in extracting our real/current kernels and leak them in the open, we can't do the same with yours. "We know" a guy that was part of the driver development team of AMD and he is itching to test its kernel-level emulator by trying to extract your current kernels. Heck, if he's successful, we can even put your kernels in our miner and let the users select them explicitly and see for themselves how they compare with ours.  Grin

1. Yes you cannot detect my runtime attacks, at least not all, and 100% you cannot detect driver level attacks. In fact, your protection is really weak, I spent some more time and now have three different ways to get kernels and they all return same binary. Of course you will state that your super-protection even detects my system driver, it's ridiculous.
2. My dumped kernels are on this forum already in public, so I don't care about your "kernel guy", "a guy from AMD" etc. The only thing I care is to prove that you are the liar.


Please read Readme and FAQ in the first post of this thread before asking any questions, probably the answer is already there.
List of my miners: https://bitcointalk.org/index.php?topic=3019607
wtfonly16
Full Member
***
Offline Offline

Activity: 253
Merit: 100


View Profile
March 01, 2018, 02:31:06 PM
 #793

its 2018... we have idiots who think asus b250 is a trash board. mining has really gone to shit. to many new "gamers" who have no idea wtf their doing.

so u calling pheonix a lier is like calling a retarded person retarted... just file lawsuit and be done with it.

ye i aint bares
Wolf0
Legendary
*
Offline Offline

Activity: 1834
Merit: 1002


Miner Developer


View Profile
March 01, 2018, 02:45:10 PM
 #794

Also, if you really had serious ties to AMD - it's not likely to be something you'd flaunt. It sounds awesome - yeah. There's two tragedies in life: one is not getting what you want; the other one is getting it.

Code:
Donations: BTC: 1WoLFdwcfNEg64fTYsX1P25KUzzSjtEZC -- XMR: 45SLUTzk7UXYHmzJ7bFN6FPfzTusdUVAZjPRgmEDw7G3SeimWM2kCdnDQXwDBYGUWaBtZNgjYtEYA22aMQT4t8KfU3vHLHG
tenmoi
Jr. Member
*
Offline Offline

Activity: 54
Merit: 0


View Profile
March 01, 2018, 04:21:07 PM
 #795

Also, if you really had serious ties to AMD - it's not likely to be something you'd flaunt. It sounds awesome - yeah. There's two tragedies in life: one is not getting what you want; the other one is getting it.

May I chime in? There’s this hidden tragedy named demion with his nodevfee sword. So stop quarreling and lower the fee before the life blood is cut. Grin
wtfonly16
Full Member
***
Offline Offline

Activity: 253
Merit: 100


View Profile
March 01, 2018, 04:48:35 PM
 #796

or just release private binaries already keeping same 1% fee lol

if the whole world has it u as the dev get more cause everyone mining faster.

ye i aint bares
Wolf0
Legendary
*
Offline Offline

Activity: 1834
Merit: 1002


Miner Developer


View Profile
March 01, 2018, 04:53:00 PM
 #797

Also, if you really had serious ties to AMD - it's not likely to be something you'd flaunt. It sounds awesome - yeah. There's two tragedies in life: one is not getting what you want; the other one is getting it.

May I chime in? There’s this hidden tragedy named demion with his nodevfee sword. So stop quarreling and lower the fee before the life blood is cut. Grin

I didn't get this kernel so good because of that. That was just study and hard work. There's not a shitton behind GCN's compute logic besides what is public - matter of fact, I can't think of any.

It sucks for other reasons - related to the ACTUAL benefits. Tongue

Code:
Donations: BTC: 1WoLFdwcfNEg64fTYsX1P25KUzzSjtEZC -- XMR: 45SLUTzk7UXYHmzJ7bFN6FPfzTusdUVAZjPRgmEDw7G3SeimWM2kCdnDQXwDBYGUWaBtZNgjYtEYA22aMQT4t8KfU3vHLHG
pickaxe123
Newbie
*
Offline Offline

Activity: 5
Merit: 0


View Profile
March 01, 2018, 05:04:07 PM
 #798

I've been mining since 2.4, now on 2.7b, on a single Vega 56, W10, latest drivers, and it's been mostly fine. I've only experienced one recurring problem where mining stops (driver crash AFAIK) then restart and hangs at detecting cards for 60-80 minutes before resuming mining. Before 2.7b, my wattman overclock settings were reset to default, now with 2.7b I get to keep my settings which is much better already.
    Thank you for the logs. Unfortunately we can't determine much from there just that the GPU thread is freezing after the crash. You've got quite a bit of memory overclock though maybe try dialing it back a little. We also had problems with Vega because of the card high power consumption - even 600W quality PSU led to sporadic crashes when powering a single Vega56 even with big underclock. It seems like Vega has high peak currents that lead to crashes if the PSU is not ridiculously overpowered. We have some work to do with hardware controls support for Vega though, as it appears to use some new (and undocumented) ADL functions and structures.

Ok I'll try lowering mem clocks but in the meantime do you have an idea why recovering takes 60-80 minutes in that case?

Here's another log from this morning:

Code:
17591:09:24:46.146: main Phoneix Miner 2.7b Windows/msvc - Release
17591:09:24:46.146: main Cmd line: -pool eu1.ethermine.org:4444 -pool2 us1.ethermine.org:4444 -wal xxx -cdm 0 -amd -fanmin 10 -fanmax 20 -tmax 65 -powlim 50 -cclock 997 -cvddc 935 -mclock 1125 -mvddc 950 -minRigSpeed 39
17591:10:36:41.383: main Available GPUs for mining:

Thanks for your help
flatounet
Jr. Member
*
Offline Offline

Activity: 53
Merit: 0


View Profile
March 01, 2018, 05:57:16 PM
 #799

tested OC by phoenix too ,

i dont have the same minning speed .... ,i think dont have constant voltage
i have it in afterburner ,but not in phoenix ....

back to afterburner

OC with AB = 1880w
OC with phoenix = 1840w 

minning speed same ,a bit slower with phoenix
GPUHoarder
Member
**
Offline Offline

Activity: 84
Merit: 22


View Profile
March 01, 2018, 05:58:44 PM
 #800

I may look into it, just to check. While they have only a Win version, where I'm less practiced, I don't need nor want to go through the logic itself - location and extraction of the binary GPU kernel will tell me. Even if they based it on yours and made modifications - certain things in it stand out, so it shouldn't be ambiguous at all.
While I *know* it's quite possible to do better than Claymore's kernels - recently, he's made improvements resulting in them being quite a bit better, which cuts some of my own advantage. The optimizations in regards to core computations I've done result in me requiring far lower core clocks for the same hashrates, so I'll take the power savings. Despite this, I don't see Claymore blatantly lying about something so easily proved.

It would be nice if you can confirm my words. No modifications at all, the kernels are the same, taken from v10. I know you dumped my binaries for v10 so you can compare them easily.

It is also quite wrong of you to assume these things:
  • That you can't be fooled by our anti-reversing measures (really do you think that we can't detect dll-injection attacks?).
  • And that, if you do succeed in extracting our real/current kernels and leak them in the open, we can't do the same with yours. "We know" a guy that was part of the driver development team of AMD and he is itching to test its kernel-level emulator by trying to extract your current kernels. Heck, if he's successful, we can even put your kernels in our miner and let the users select them explicitly and see for themselves how they compare with ours.  Grin

1. Yes you cannot detect my runtime attacks, at least not all, and 100% you cannot detect driver level attacks. In fact, your protection is really weak, I spent some more time and now have three different ways to get kernels and they all return same binary. Of course you will state that your super-protection even detects my system driver, it's ridiculous.
2. My dumped kernels are on this forum already in public, so I don't care about your "kernel guy", "a guy from AMD" etc. The only thing I care is to prove that you are the liar.



I am also very active in the low level side of this, and have also dumped both kernels at a mock hardware level and can absolutely confirm Claymore’s post. Private kernels developed to match specific hardware can be developed that are faster than Claymore’s, but significant work has gone into those kernels. They are absolutely rights  of Claymore and stolen in this case. Never mind all originally stole from open source...
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 [40] 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 ... 107 »
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!