a new type of shilling - game developer locks content to specific Intel i7 CPUs [updated]

[patrickjp93]

2 minutes ago, Dabombinable said:

http://www.dictionary.com/browse/opinion

http://www.dictionary.com/browse/fact

 

Don't make me break out the esults.

Did you read those definitions, because in my opinion, you seem to have used them in an incongruous way if you had.

[AlTech]

52 minutes ago, patrickjp93 said:

Not Intel's fault AMD's hardware sucks.

 

52 minutes ago, patrickjp93 said:

Not if that brand is the only one truly capable of supporting the program.

 

It's a pity you're last name isn't Schill/Shill or Schiller/Shiller. It would be a perfect match for you.

[MageTank]

1 minute ago, patrickjp93 said:

You speculated on the likelihood that they do. The variables at play to make the conclusion of likelihood are many.

 

Yes you can; exams do that all the time.

 

I do, simply watch ALL of those CPPCon videos.

 

I stand corrected, Prysin invited you.

 

No, the scenario is not impossible. Once again, selective reading and quoting is a nasty habit of yours. This is the difference between micro benchmarking and performance profiling. Obviously you will be running multiple calculations per vertex, and then entire vertex list will be in cache if you're smart about it.

 

It is a good idea. It gets a 10x speedup now, but it also should be used in conjunction with more complex physics for the best effect. Amdahl's Law and Gustafson's Law are still in effect. The memory may be the achilles heel to getting the entire list iterated over, but that doesn't mean you can't do more than just translate the vertex list. Scaling, sheering, rotation, and translation all happen in that exact order. You'd be stupid to iterate over the list 4 times rather than do the 4 ops to each vertex as it comes along. You increase the compute cycles and work done, but you don't increase the runtime.

 

It doesn't matter yet, because the code they write is still scalar.

Was it this blog? 

Or this one?

I agree with you Patrick, selective reading really is a bad habit. Can we skip to the part where you inevitably give up due to lack of evidence on your part to prove that an entire industry lacks the knowledge that you claim they lack? I have to be in bed soon. 

[patrickjp93]

1 minute ago, MageTank said:

Was it this blog? 

 

Or this o

to the part where you inevitably give up due to lack of evidence on your part to prove that an entire industry lacks the knowledge that you claim they lack? I have to be in bed soon. 

Yes, the first blog. And the evidence is there. You can reproduce the experiment. Now, that is evidence, but I was in the process of expanding on it and being thorough when interest died off. If you'd like me to continue, I can.

 

The scalar code is taken directly from Unreal Engine 4. The vector code is mine. Case closed

[patrickjp93]

4 minutes ago, AluminiumTech said:

 

 

It's a pity you're last name isn't Schill/Shill or Schiller/Shiller. It would be a perfect match for you.

What? Have you seen the SSE and AVX memory load/store times for Bulldozer and Piledriver? They're 4x what it takes Sandy Bridge to do the same. AMD's existing hardware sucks. The throughput, efficiency, and scalability are all vastly lower than Intel's. That is a fact. It is my opinion that it then warrants saying their hardware sucks.

[MageTank]

1 minute ago, patrickjp93 said:

Yes, the first blog. And the evidence is there. You can reproduce the experiment. Now, that is evidence, but I was in the process of expanding on it and being thorough when interest died off. If you'd like me to continue, I can.

 

The scalar code is taken directly from Unreal Engine 4. The vector code is mine. Case closed

So... do you stand corrected-corrected? Or how does that work? As for whether or not you should continue, i'd rather you never stop. You are a living perpetual motion machine, an infinite source of energy. 

[Dabombinable]

Just now, MageTank said:

So... do you stand corrected-corrected? Or how does that work? As for whether or not you should continue, i'd rather you never stop. You are a living perpetual motion machine, an infinite source of energy. 

*facepalms

[patrickjp93]

1 minute ago, MageTank said:

So... do you stand corrected-corrected? Or how does that work? As for whether or not you should continue, i'd rather you never stop. You are a living perpetual motion machine, an infinite source of energy. 

So prove my claim and evidence are both wrong if you're so inclined. If little old me is 10 times the programmer you can find at Epic Games, Blizzard, EA, Ubisoft, Naughty Dog, CDPR, and IDTech, then the industry has a damn problem lacking expertise.

[MageTank]

2 minutes ago, Dabombinable said:

*facepalms

What else was I to do? It'd be even more incorrect to call him an infinite source of power. That would imply he had power.

1 minute ago, patrickjp93 said:

So prove my claim and evidence are both wrong if you're so inclined. If little old me is 10 times the programmer you can find at Epic Games, Blizzard, EA, Ubisoft, Naughty Dog, CDPR, and IDTech, then the industry has a damn problem lacking expertise.

That selective reading kicked in again. I was poking fun at the fact that you said it was @Prysin that invited me, when I proved that it was you. The joke came from the fact that you said "I stand corrected" when making that claim. As for you being 10x the programmer as those employed by the studios you mentioned, that is yet to be seen. Your work is on paper, let me know when you have something tangible, that is created, that we can see with our own eyes out-performing other games. Don't tell me to go compile that junk either, my overclocks are only pseudostable. #LaysXMP

[patrickjp93]

3 minutes ago, MageTank said:

So... do you stand corrected-corrected? Or how does that work? As for whether or not you should continue, i'd rather you never stop. You are a living perpetual motion machine, an infinite source of energy. 

Still Prysin's suggestion, but my first claim was right. I did invite you, and that experiment does prove my point in part at the very least.

[patrickjp93]

20 minutes ago, MageTank said:

What else was I to do? It'd be even more incorrect to call him an infinite source of power. That would imply he had power.

That selective reading kicked in again. I was poking fun at the fact that you said it was @Prysin that invited me, when I proved that it was you. The joke came from the fact that you said "I stand corrected" when making that claim. As for you being 10x the programmer as those employed by the studios you mentioned, that is yet to be seen. Your work is on paper, let me know when you have something tangible, that is created, that we can see with our own eyes out-performing other games. Don't tell me to go compile that junk either, my overclocks are only pseudostable. #LaysXMP

See now you're the one backing down from the scientific method. Anyone on here with Sandy Bridge and above can run the test. Or you could be a good sport and either disable your overclock or just not be a wuss and run the program since it's nowhere near as stressful on the system as AIDA64, Prime95, or Intel Burn Test. It's just vectorized addition. No fourier transforms, no hashing, no division, just addition.

 

It isn't just on paper. It's something you can run.

 

@Prysin @Dabombinable @AluminiumTech Come on, someone step up to the plate. Validate or invalidate my experiment! Someone have the balls to actually give me a fair shake.

 

 

[MageTank]

1 minute ago, patrickjp93 said:

See now you're the one backing down from the scientific method. Anyone on here with Sandy Bridge and above can run the test. Or you could be a good sport and either disable your overclock or just not be a wuss and run the program since it's nowhere near as stressful on the system as AIDA64, Prime95, or Intel Burn Test. It's just vectorized addition. No fourier transforms, no hashing, no division, just addition.

 

It isn't just on paper. It's something you can run.

That'd work, if I knew even a tiny little bit about compiling code, let alone actually running it. I am no expert, nor do I claim to be. If you could make it into a fancy little executable, i'd run it with and without my overclocks. I'd even let you know whether or not more memory bandwidth helps. It's not me backing down, it's me asking you to do some of the work for me, there is a difference, lol. 

 

I do want to play a little bit with the context of your words though. "Now you're the one backing down". This implies someone else was backing down. Does this mean you were the one backing down before? That'd be an admittance of defeat, from the man that never loses, would it not?

[patrickjp93]

8 minutes ago, MageTank said:

That'd work, if I knew even a tiny little bit about compiling code, let alone actually running it. I am no expert, nor do I claim to be. If you could make it into a fancy little executable, i'd run it with and without my overclocks. I'd even let you know whether or not more memory bandwidth helps. It's not me backing down, it's me asking you to do some of the work for me, there is a difference, lol. 

 

I do want to play a little bit with the context of your words though. "Now you're the one backing down". This implies someone else was backing down. Does this mean you were the one backing down before? That'd be an admittance of defeat, from the man that never loses, would it not?

It's really easy. Heck, I provided the hot-to in the blog post. Get a Linux machine (I don't know the flags to get intrinsic functions working with Microsoft's underperforming crap compiler), go to the command line, navigate to the directory with the code file and type the following:

 

g++ -std=c++14 -march=native -O3 -o give_your_executable_a_name

or install clang and type:

clang++ ...

or install the intel compiler and type:

icpc ...

 

To then run the executable, type the following:

./your_executable_name

 

You constantly claim I'm the one running from the arguments, except I never do, and this time it's in plain public view that you are a coward.

 

No, it's not an admittance of defeat. It's pure contempt for you and your bravado.

[MageTank]

1 minute ago, patrickjp93 said:

You constantly claim I'm the one running from the arguments, except I never do, and this time it's in plain public view that you are a coward.

 

No, it's not an admittance of defeat. It's pure contempt for you and your bravado.

Now THAT is Ad Hominem. 

 

1 hour ago, patrickjp93 said:

The Ad Hominem is the last weapon of a desperate man who's already lost the argument.

 Thank you past-Patrick for always having my back.

 

In other news: asking someone to provide evidence to you, only for them to say "my work is evidence", and then admitting you do not know how to use their work, makes you a coward. More at 11. 

[patrickjp93]

2 minutes ago, MageTank said:

Now THAT is Ad Hominem. 

 

 Thank you past-Patrick for always having my back.

 

In other news: asking someone to provide evidence to you, only for them to say "my work is evidence", and then admitting you do not know how to use their work, makes you a coward. More at 11. 

No, it's a statement of my opinion, and that opinion is based on the fact you said you wouldn't validate the experiment because of a "pseudo stable overclock" which isn't an excuse when you can undo the overclock or just try to run it anyway.

 

I provided you how to use my work in the blog you cowardly liar.

[MageTank]

Just now, patrickjp93 said:

No, it's a statement of my opinion, and that opinion is based on the fact you said you wouldn't validate the experiment because of a "pseudo stable overclock" which isn't an excuse when you can undo the overclock or just try to run it anyway.

 

I provided you how to use my work you liar.

The pseudostable overclock was a joke towards @Lays, hence the "#LaysXMP". I was mocking his overclock. I can understand how a man as sophisticated as yourself might fail to grasp the concept of a joke. Also, I never said I wouldn't validate the experiment, I simply said "don't tell me to compile that junk either". Which, you totally did. Rude of you to not listen like that. Not only that, if you were to be offended by anything, it should of been me saying "junk" in reference to your code. Australia has you all upside down.

 

Somebody is gonna get that. 

 

Besides, they don't make tests strong enough to make my overclocks fail. I like how you see me as this destructive force, out to destroy your reputation, when you do a far better job at it than I ever will. I'm just an idiot that sets out to prove other idiots wrong. Don't fault me for that. 

 

[patrickjp93]

9 minutes ago, MageTank said:

The pseudostable overclock was a joke towards @Lays, hence the "#LaysXMP". I was mocking his overclock. I can understand how a man as sophisticated as yourself might fail to grasp the concept of a joke. Also, I never said I wouldn't validate the experiment, I simply said "don't tell me to compile that junk either". Which, you totally did. Rude of you to not listen like that. Not only that, if you were to be offended by anything, it should of been me saying "junk" in reference to your code. Australia has you all upside down.

 

Somebody is gonna get that. 

 

Besides, they don't make tests strong enough to make my overclocks fail. I like how you see me as this destructive force, out to destroy your reputation, when you do a far better job at it than I ever will. I'm just an idiot that sets out to prove other idiots wrong. Don't fault me for that. 

 

Don't know him, had no context for the joke, missed it, and now we're moving on.

 

I'm going to tell you to compile it, because otherwise if I provide you an executable file, you'll tell claim I rigged it, and you'd be fair to claim that since there's no way for you to tell what the source code was (without also knowing how to disassemble and decompile AND detangle an executable).

 

Now all you need is a down under line and you can make a run at my masculinity so you can keep dodging the actual evidence in front of your face and scoring brownie points with @Dabombinable who likes all of your posts despite the fact they have no substance.

 

I'm going to fault you when your methods suck and your effort is even more worthless.

 

You have the code, you know how to compile it, and you know how to run it. You have no right to say my work is invalid if you can't put in the minimal effort to run the experiment yourself. You're being cowardly, plain and simple.

[Dabombinable]

5 minutes ago, patrickjp93 said:

Don't know him, had no context for the joke, missed it, and now we're moving on.

 

I'm going to tell you to compile it, because otherwise if I provide you an executable file, you'll tell claim I rigged it, and you'd be fair to claim that since there's no way for you to tell what the source code was (without also knowing how to disassemble and decompile AND detangle an executable).

 

Now all you need is a down under line and you can make a run at my masculinity so you can keep dodging the actual evidence in front of your face and scoring brownie points with @Dabombinable who likes all of your posts despite the fact they have no substance.

 

I'm going to fault you when your methods suck and your effort is even more worthless.

 

You have the code, you know how to compile it, and you know how to run it. You have no right to say my work is invalid if you can't put in the minimal effort to run the experiment yourself. You're being cowardly, plain and simple.

Compile it yourself, so we don't have you denying the results with "you changed something".

[patrickjp93]

Now, I'm going to cook a gourmet dinner and eat it. Take some time to install a Linux VM (I prefer Fedora, but Ubuntu will be fine) or a dual boot, update GCC using sudo apt-get update or sudo dnf update, copy the code into TextEdit, save the file as a .cpp file, compile it like I showed above, run it like I showed above, and come back to me with a result.

[MageTank]

1 minute ago, patrickjp93 said:

Don't know him, had no context for the joke, missed it, and now we're moving on.

 

I'm going to tell you to compile it, because otherwise if I provide you an executable file, you'll tell claim I rigged it, and you'd be fair to claim that since there's no way for you to tell what the source code was (without also knowing how to disassemble and decompile AND detangle an executable).

 

I'm going to fault you when your methods suck and your effort is even more worthless.

 

You have the code, you know how to compile it, and you know how to run it. You have no right to say my work is invalid if you can't put in the minimal effort to run the experiment yourself.

I wouldn't accuse you of rigging an executable. What could you possibly rig about it that would help (or diminish) your cause? If it performs better than comparable code, then your words have merit. Though, I am curious. Which methods of mine suck, and exactly what effort am I putting in to this? So far, it's been a very simple pattern between us. Someone makes a statement, you make a baseless claim about said statement, and I show up to ask you to prove it. That's been the formula since you and I first met, and has never changed. Your only evidence in this entire thread, are CppCon videos and names of random people (at least, random to me) that are not even referring to the gaming industry. The sources you've provided have done absolutely nothing for your argument. 

 

As for your code, I'll get my brother to compile it for me, and I'll run it. I still don't get what it will prove, but i'll do it for you. Unless your code can encompass every aspect of a video game program, and not be bottlenecked by something else entirely (Amdahl's law, as you are aware), I don't see it as being entirely useful. 

[patrickjp93]

Just now, Dabombinable said:

Compile it yourself, so we don't have you can't deny the results with "you changed something".

That's a Byzantine Generals problem. I at least offered to trust you. It seems I am the only gentleman around here. I gave you the exact methodology. Use it. You could also make a video of you going through the process if you think I'd have the gall to accuse you.

[Dabombinable]

2 minutes ago, patrickjp93 said:

That's a Byzantine Generals problem. I at least offered to trust you. It seems I am the only gentleman around here. I gave you the exact methodology. Use it. You could also make a video of you going through the process if you think I'd have the gall to accuse you.

Your version of providing evidence includes text books that are expensive and hard to get+encouraging others to breach copyright.....why should I trust you again?

[patrickjp93]

Just now, MageTank said:

I wouldn't accuse you of rigging an executable. What could you possibly rig about it that would help (or diminish) your cause? If it performs better than comparable code, then your words have merit. Though, I am curious. Which methods of mine suck, and exactly what effort am I putting in to this? So far, it's been a very simple pattern between us. Someone makes a statement, you make a baseless claim about said statement, and I show up to ask you to prove it. That's been the formula since you and I first met, and has never changed. Your only evidence in this entire thread, are CppCon videos and names of random people (at least, random to me) that are not even referring to the gaming industry. The sources you've provided have done absolutely nothing for your argument. 

 

As for your code, I'll get my brother to compile it for me, and I'll run it. I still don't get what it will prove, but i'll do it for you. Unless your code can encompass every aspect of a video game program, and not be bottlenecked by something else entirely (Amdahl's law, as you are aware), I don't see it as being entirely useful. 

Change the loop number count for one of them, super easy.

 

You use emotional pleas and insults to diminish your opponent rather than actually debate the facts and evidence. Very little, because you slink away from the argument with every pass.

 

Herb Sutter:https://herbsutter.com Principal Software Architect microsoft and chief of the ISOCPP Concurrency Standard Working Group. He is in charge of building concurrency standards into ISO C++ and implementing them in the Microsoft Visual C++ standard library.

 

Mike Acton: https://www.linkedin.com/in/mikeacton
Senior Engine Developer at Insomniac Games

 

Scott Meyers: former Chief Software Architect for Nokia and the founder of the Code::Dive conference. Nowadays he is a professional instructor and consultant who teaches various software development and programming skills.

 

Jason Turner: http://www.oreilly.com/pub/au/6951

Independent Contractor and Major Contributor to the LLVM C++ Standard Libraries. He's been programming in C++ professionally for 15 years.

 

These are all very big names in the world of C++. 

 

You don't see it as useful even though vertex manipulation is one of the core functions of a game engine that is done for every frame... Vectorized addition is also used in lighting. It's a core piece of every part of the rendering pipeline. You speed that up by 10x, you get a significant boost in the entire pipeline. 

[patrickjp93]

8 minutes ago, Dabombinable said:

Your version of providing evidence includes text books that are expensive and hard to get+encouraging others to breach copyright.....why should I trust you again?

Just because you aren't willing to do the work to get my publicly available evidence doesn't make the evidence invalid.

 

You shouldn't, but I'm giving you something tangible you can trust, so stop making this about me and make it about the facts.

[MageTank]

1 minute ago, patrickjp93 said:

Change the loop number count for one of them, super easy.

 

You use emotional pleas and insults to diminish your opponent rather than actually debate the facts and evidence. Very little, because you slink away from the argument with every pass.

 

Herb Sutter:https://herbsutter.com Principal Software Architect microsoft and chief of the ISOCPP Concurrency Standard Working Group. He is in charge of building concurrency standards into ISO C++ and implementing them in the Microsoft Visual C++ standard library.

 

Mike Acton: https://www.linkedin.com/in/mikeacton
Senior Engine Developer at Insomniac Games

 

Scott Meyers: former Chief Software Architect for Nokia and the founder of the Code::Dive conference. Nowadays he is a professional instructor and consultant who teaches various software development and programming skills.

 

Jason Turner: http://www.oreilly.com/pub/au/6951

Independent Contractor and Major Contributor to the LLVM C++ Standard Libraries. He's been programming in C++ professionally for 15 years.

 

These are all very big names in the world of C++. 

 

You don't see it as useful even though vertex manipulation is one of the core functions of a game engine that is done for every frame... Vectorized addition is also used in lighting. It's a core piece of every part of the rendering pipeline. You speed that up by 10x, you get a significant boost in the entire pipeline. 

I don't slink away, I just dig through the rubbish, find the opening, and exploit it. Take this for example:

Quote

Mike Acton: https://www.linkedin.com/in/mikeacton
Senior Engine Developer at Insomniac Games

If he is a developer at Insomniac Games, doesn't that mean he is part of the gaming industry? If he is part of the gaming industry, and knows how to optimize for victim cache vs pure hierarchy, wouldn't that refute your own claim? Don't you even dare say Insomniac Games doesn't count as a AAA studio, because Ratchet and Clank was my jam, and was not only a platform headliner, it's also a movie now. Can't get more AAA than that.

 

Also, you keep saying I use insults, but when have I insulted you?