You can try my software as many times as you want (without paying anything) and you only decide to pay (or not) after speedup achieved with the tuned compiler sequence (i.e. sequence of LLVM optimizations) over -O3 has been reported to you by email.
So... No risk for you!
Topic: LLVM pass selector using AI for 0.01 BTC each try (up 3x performance over -O3) (Read 731 times)
August 14, 2014, 01:42:36 PM
August 13, 2014, 09:46:53 AM
This is state of the art technology!
You can improve the performance of your applications by over 2 times!
You just need to compile part of your C code (the part that you send using the online form, should be the part that is taking most CPU cycles and therefore needs to be optimized more) with Clang 3.5 with the specific compiler sequence my software derives specifically for your code!
If you are a software developed my software can give you a HUGE advantage for as little as 0.01 BTC per online request.
I'm only asking such a low price because I want a lot of people to use it.
The price may be much higher in future!
You can improve the performance of your applications by over 2 times!
You just need to compile part of your C code (the part that you send using the online form, should be the part that is taking most CPU cycles and therefore needs to be optimized more) with Clang 3.5 with the specific compiler sequence my software derives specifically for your code!
If you are a software developed my software can give you a HUGE advantage for as little as 0.01 BTC per online request.
I'm only asking such a low price because I want a lot of people to use it.
The price may be much higher in future!
August 12, 2014, 08:49:41 AM
If you do now know what the hell I'm talking about I can answer your deep questions...lol
August 11, 2014, 09:49:58 PM
This is a service that I created based on research I've been doing, so it is very close to state of the art.
http://80.240.135.146
It automatically provides LLVM OPT optimization sequences that will result in better performance for your C (other languages will be added in future) applications.
Can be very usefull if you use Clang/LLVM and are trying to squeeze the maximum performance from a certain part of your source code (usually the function(s) consuming more CPU cycles).
The instructions to use the service are in the main page of the website.
You will first receive an email stating the performance improvement (i.e. speedup) over Clang -O3 that was achieved using the tailored LLVM OPT sequence, and only then you will decide if you want to pay 0.005 Bitcoin (address to sent to is in the email) for receiving the compiler sequence that achieved the said speedup.
The service is in beta/alpha, but it is working very well.
That is why I'm only asking for 0.01 BTC per compiler sequence.
If you have any problem or want to ask something please send me a PM, or write in this thread.
This is really state of the art stuff! It uses very well tuned Artificial Intelligent algorithms!
You can try my software as many times as you want (without paying anything) and you only decide to pay (or not) after speedup achieved with the tuned compiler sequence (i.e. sequence of LLVM optimizations) over -O3 has been reported to you by email.
So... No risk for you!
Enjoy
http://80.240.135.146
It automatically provides LLVM OPT optimization sequences that will result in better performance for your C (other languages will be added in future) applications.
Can be very usefull if you use Clang/LLVM and are trying to squeeze the maximum performance from a certain part of your source code (usually the function(s) consuming more CPU cycles).
The instructions to use the service are in the main page of the website.
You will first receive an email stating the performance improvement (i.e. speedup) over Clang -O3 that was achieved using the tailored LLVM OPT sequence, and only then you will decide if you want to pay 0.005 Bitcoin (address to sent to is in the email) for receiving the compiler sequence that achieved the said speedup.
The service is in beta/alpha, but it is working very well.
That is why I'm only asking for 0.01 BTC per compiler sequence.
If you have any problem or want to ask something please send me a PM, or write in this thread.
This is really state of the art stuff! It uses very well tuned Artificial Intelligent algorithms!
You can try my software as many times as you want (without paying anything) and you only decide to pay (or not) after speedup achieved with the tuned compiler sequence (i.e. sequence of LLVM optimizations) over -O3 has been reported to you by email.
So... No risk for you!
Enjoy
Jump to: