Encryption Projects as SU group

[sattechtips]

GTX 1080Ti run a single instance of CudaBISS at approximately 4 days entire range.
GTX 2080Ti runs a single instance of CudaBISS at approximately 30-32 hrs. entire range.
GTX 3090Ti run a single instance of CudaBISS at approximately 25-27 hrs. entire range.
GTX 4090 run a single instance of CudaBISS at approximately 11-12 hrs. entire range.
Maybe GTX 4090Ti run a single instance of CudaBISS at approximately 9-10 hrs. entire range.

 

[cayoenrique]

@moonbase

All if not most of the stuff I got involve( posting) is gone.

But I can still find @C0der, @kebien, @dale_para_bajo, @campag5242 and most important @K2TSET chitchats. For exaple the following forum links

CSA Brute Force
Why RBT, Cudabiss and K2TSET PGFA BF may fail!.
Brute Force ECM algo to hit CW = Key

@K2TSET did help me in the past getting hard to find keys using his FPGA.

Now CSA Brute Force hat such nice pictures and explanation on the 000001 search process and CSA operation. But all is gone.

Once one or a few show interest in CSA BruteForce I can post some programing samples. Clearly I will begin with simplest C programing. Then if people survive we try to do basic OpenCL in GPU.

 

[dvlajkovic]

"Golden Sample" is a series, maybe from RTX 3070 to 4090 + there with or not TI. It is not enough info.. It is not enough info.

You did not pay attention to details in story > link

@dvlajkovic is getting 934 Mkps from whatever Graphics card they are using, this is a slow speed, it is not fast compared to what an RTX 3090 or an RTX 4090 can achieve.
Based on my test resutls with various Graphics cards a speed of 934 MKps is from a card somewhere in the region of the 900 series or the 1000 series.

My personal tests with cards from these two series are listed below:
GTX 980Ti runs a single instance of CudaBISS at approximately 700 Mkps
GTX 1080Ti runs a single instance of CudaBISS at approximately 1060 MKps

I said it is 934,124 MKPS in just one instance, NOT 934 Mkps

This needs to be read as nine hundred thirty-four thousand one hundred twenty-four MKPS.
Resulting speed: -934124288 keys/s
In short, almost a billion keys per second.
That fast enough for you?

 

[cayoenrique]

OHHH! I will have to move to the Mediterranean Coast, I guess that apart from Heat and Fire, there are a lot of nice PC's....

Boy oh boy, no wonder so much Heat & Fire lately, there is a Dragon on the area...

 

[SatEze]

This could be elementary, but I have read the acronym "MKPs" mentioned here severally. How does one determine those values and where can it be found on the CUDA BISS tools? Would the "MKps" be the item circled in red in the attached screenshot?

 

[cayoenrique]

I can not run Cudabiss as I have no Nvidia GPU. But looking at a Youtube video

https://www.youtube.com/watch?v=YbcqdM39sjg

I can see that from time to time you get this

 

[moonbase]

This could be elementary, but I have read the acronym "MKPs" mentioned here severally. How does one determine those values and where can it be found on the CUDA BISS tools? Would the "MKps" be the item circled in red in the attached screenshot?

Your screenshot using an RTX 3060 graphics card presents the speed as kCW/sec.
This is thousands of keys per second.
MKps is millions of keys per second.

More recent generation graphics cards such as an RTX 3090 process CudaBISS searches in billions of keys per second and also allow the user to run multiple instances at the same time with decent speeds.

 

[cayoenrique]

Listen guys do not get mad with each other. For me a little less a little more is not the issue. As they do Full search in a day.

@dvlajkovic
Now just for the record lets look at the link you posted.

Link 1
I read

Processing time: 12329469.000000 (ms)
Total amount of keys searched: 41696649805824.000000
Resulting speed: -913098240 keys/s

This from start do not makes sense. Let me show you. Lets assume "." is for decimal and "," every thousands. Be aware I see a negative "-" previous of the 9

Processing time: 12329469.000000 (ms) = 12329469 / 1000 = 12329.469 Seconds
Keys / sec = 41696649805824 / 12329.469 = 3381869065.55538 = 3,381,869,065 = 3.38*10^9 = 3.38 billions = 3,391 MKPS

But it shows

Resulting speed: -913098240 keys/s = 913,098,240 = 9.13/10^8 = 0.913 billions = 913 MKPS

Lets compare with moonbase numbers even if he sis 4 instances. He did 3376 MKS on 10,496 cuda cores. If I ignore all other factors in the PC and speeds of the cores I can proximate yours.

your 4 instance speed = his 4 instance speed x ( your # cores * your core speed ) / ( his # cores * his core speed )
= 3376 x ( 16384 x 2235 ) / ( 10496 x 1395 ) = 8443.09887227905 MKPS

This is what moonbase is trying to say. I suspect that your mother board could influence that speed. Or on fact you are doing 3,391 MKPS as I point out a few line up.

 

[SatEze]

I can not run Cudabiss as I have no Nvidia GPU. But looking at a Youtube video

https://www.youtube.com/watch?v=YbcqdM39sjg

I can see that from time to time you get this

Thanks for the clarification.

 

[C0der]

-913098240 = 0xC9933A00 = 3381869056 = 3381M

 

[moonbase]

-913098240 = 0xC9933A00 = 3381869056 = 3381M

Thank you for pointing this out.

I think there has been some confusion in this topic about the search speeds achieved by @dvlajkovic
It was not obvious to me that the figure of -913098240 had to be converted to a normal number.

If this conversion that you illustrated has to be applied then the speed of 3381 MKps from a single instance is fast, quite a jump up from an RTX 3090 perfomance with CudaBISS.

 

[dvlajkovic]

Ok, we have established that I've got the slowest card here, much slower than 1080ti. That's fine

It surely takes almost 24hrs for full range with cudabiss, one instance, using my slow dragon.
Of course, this can be shorter in a case the cw starts with a lower figure, i.e. 29 or 75 or AB ...
But it takes a little less than 24hrs when cw goes like this: FF 52 86 D7 C9 EA 41 F4.
That's tested many times and the results were posted on SU (sometimes without the resulting speed 'cause everyone wanted a cw, not the technicality how fast that cw was found).
Now when we brought all this out, can we move fwd with our 'league of biss warriors' and make a plan what can be done to speed up the whole BF search thing (without attack list)?
Thank you

 

[moonbase]

Now when we brought all this out, can we move fwd with our 'league of biss warriors' and make a plan what can be done to speed up the whole BF search thing (without attack list)?
Thank you

If CudaBISS was re-coded to make use of modern PC architecture and GPU's it could probably be made to run faster.
To me, that seems the best approach. However, the source code is not available as far as I am aware?

There might be other BF search apps that have been coded by developers that are faster than CudaBISS?
However, I do not think there are any in the public domain

 

[cayoenrique]

@C0der nice to see you.

@dvlajkovic from what I read on past threads, C0der is one of the masters to ask for help he been discussing this How to speed up CSA brute-force for quite some time.
From me. I do not have on my hand, a complete tool similar to Cudabiss that I can give you. I can tell you I have try in the past and I have pieces where I test it but fail to get improvements. In general my stream decypher is to slow. I can ONLY offer to help discuss the Issue here and learn on the way. I started this thread about 2 days ago. I ask people to join a leaning course in CSA brute-force under OpenCl a day ago. And up to now no one said it is interested to join....

I just want to clarify to you something about your comment on

(without attack list)?

You need to work as hard to do attack list than consecutive search from 0 to 0xFFFFFFFFFFFF. It will use same program with same final speed. At best, what you may mean is forget CPU and do it in GPU.

Now when we brought all this out, can we move fwd with our 'league of biss warriors' and make a plan what can be done to speed up the whole BF search thing (without attack list)?

So who can move fwd and help @dvlajkovic?

 

[Me2019H]

I ask people to join a leaning course in CSA brute-force under OpenCl a day ago. And up to now no one said it is interested to join....

Hello cayoenrique,
I have been waiting for the lessons you will give to beginners to get started, and I follow all of your topics daily
I am in the process of buying a graphics card. Can't we start the tutorial and experiment on the .TS file and the explanation without using the graphics card?

 

[dvlajkovic]

So who can move fwd and help @dvlajkovic?

Huh, it's not just me... there are a lot of powerful nvidia VGA cards around.
The cudabiss has been developed when cuda cores were just introduced to graphics market, much before GTX 1080 and the others.

Nowadays many of us have beasts like RTX 2xxx, RTX 3xxx, etc. and would appreciate a tool that can BF biss encrypted video much faster than an old cudabiss.

Any OpenCL based tool would be also welcome if it can surpass cudabiss.
I'd like just to point out that a very few people among us here are using linux, that's why we would really like a windows version, too -if possible.

 

[cayoenrique]

Me2019H I knew when you show up you will say sign me in.

For newbies:
Cuda is a programing method use only by Nvidia.
Opencl is the standard, it is designed to run across multiple platforms. From Nvidia to none Nvidia GPUs, CPU, FPGA, Android Phones, Super Computers or even in your pull wagon (wheelbarrows ) you had when you where a child....
In general Opencl do as equal as Cuda. Now there are a few instructions that makes Cuda faster in certain application but in other Opencl will run faster.

What we are talking about?
GPGPU General-purpose computing on graphics processing units

Why?
Because in general most GPU have at most 4 cores. So at best you could split the CPU and make them run 4 time the same work load.

Lets Ignore that a PC has many other elements that affect speed.
And lets take cayoenrique's sample CPU Intel 4 Core @ 2.00GHz .
Computing Power = 4 x 2 Ghz = 8

Now lets compare with dvlajkovic GPU it has 16,384 cores @ 2.235 Ghz
Computing Power = 16,384 X 2.235 = 36,618.24

If some how I could use GPU to make simple calculations the I could have an improve!!

The improve from my CPU to dvlajkovic GPU is that in the same time I do 8 computations he do 36,618.24. I am so shame on my computer....

Any OpenCL based tool would be also welcome if it can surpass cudabiss.

I do not have it. But you guys can try to build one!!! Nor I can not provide assurances that your final build in Opencl will be faster than Cudabis. Only you can prove that to me.

I'd like just to point out that a very few people among us here are using linux, that's why we would really like a windows version, too -if possible.

I learn from my last Tutorial in Encryption. People did not got involve, even when I did post a thread how to run the lLinux instructions in your W10. So now all my sources have a makefile that can build either windows exe or Linux posix programs.
But you are required to install necessary drivers to compile Opencl.

 

[sattechtips]

Ok, we have established that I've got the slowest card here, much slower than 1080ti. That's fine

It surely takes almost 24hrs for full range with cudabiss, one instance, using my slow dragon.
Of course, this can be shorter in a case the cw starts with a lower figure, i.e. 29 or 75 or AB ...
But it takes a little less than 24hrs when cw goes like this: FF 52 86 D7 C9 EA 41 F4.
That's tested many times and the results were posted on SU (sometimes without the resulting speed 'cause everyone wanted a cw, not the technicality how fast that cw was found).
Now when we brought all this out, can we move fwd with our 'league of biss warriors' and make a plan what can be done to speed up the whole BF search thing (without attack list)?
Thank you

Look like old shit cudabiss version.

 

[dvlajkovic]

you ever found unknown cw that starts with F using cudabiss that covers full range 000000000000 - FFFFFF000000 ?

 

[cayoenrique]

I can not I do not have Cudabis.

But I wonder;
1rst you say "But it takes a little less than 24hrs when cw goes like this: FF 52 86 D7 C9 EA 41 F4." Like if FF*** where the easiest...
Now you kind of insinuate that there is something about cw starting with FF****