AMD running Intel specific code.

Now I am more educated, and understand that Intel isn't necessarily being evil (in this case).
 
Originally posted by pantherdan
wwwwwwwwww if you want to enlighten someone who is not "in the know" that is fine. Coming off as some kind of superior being actually makes you look like a punk. Thanks for the divine enlightenment. You da man.
Click to expand...

I'm really amazed you decided to quote me after making such an asinine post. I really wish you had not, because I feel obligated to respond.

Coming from someone who wrote a 3 paragraph rant on Intel with such interesting comments as:

Was this code written by Intel that attaches itself to programs to make Intel products look better, or is this someething contained in some OSes, or is this something that is included in third party software?

I am sure I know why this is being done (because Intel is above the law)

the illegal monopoly style crimes that are being committed

I am thinking this is an Intel- Microsoft crime IMHO. It would be typical since both companies think they are above the law and have always attempted to crush the competition in any illegal manner they see fit.
Click to expand...

You obviously didn't even bother to read the link. You don't understand even the very basics of the situation. Yet, you still pulled out the jump to conclusions mat and leapt off the edge.

If you are not an Intel fanboy, don't act like one.
Click to expand...

Out off all the idiotic comments you've made, that deserves an honorable mention for duplicity.
 
Drop the personality confrontations out of this thread or I will close it and start removing the problem(s) from the forum.
 
just look at the freezing problems Folding@Home has been having with AMD CPUs.
Click to expand...

I never had a single problem with Folding@Home (with -forceasm flag) until they released the new version. If you run Folding@Home on an older core (not processor core, but program core) with the -forceasm flag on, it works correctly. This isn't an AMD problem, but rather something someone did at Folding@Home.

As for all the compilier hoopla, who cares. AMD should hire some damn people and write their own compilier. Shove the Intel compilier so far up Intel's ass that....well you add the rest!
 
Originally posted by Jason711
thats sort of what i have noticed...

it runs on amd chips (only k8?) the exact same (correct me if i am wrong)

and.. intel has put something in the code to make it only run on "pentium" based cpu's, which in turn, gives it a large boost...

therefore.. we throw up the bullshit flag.

now.. i guess it is their compiler, so they can do what they want.

or is it that simple?
Click to expand...

In this one specific test, that is the case.
however, compile with /QxP, and it will most likely crash on K8, because the compiler will attempt to use SSE3 op codes which K8 of course does not support.

But to generalize and say intel is doing this to hurt AMD is simply wrong. It is a saftey mechanism for CPU specific code. Possibly an over zealous one, but not a devious one.

Run a /Qx[W|N|B|P] compiled program on a P3, and it won't run either. That's because the CPUID check is to prevent the code from being run on any CPU, amd, cyrix or even Intel, that the code wasn't intended for.

"pentium" based cpu's, which in turn, gives it a large boost...
Click to expand...

There's no evidence that the CPU flag itself does anything to speed.
/arch:SSE_ is off by default.
the orginal article says he used the default x86 CPU flag (but gives the flag for Willamette QxW), which would not enable SSE instructions
and the Northwood flag, which should enable SSE instructions by default.
Then claimed intel was giving it's cpus a boost.

where a competant comparison would have been between /arch:SSE2 for K8 and /QxN for Northwood.
to see if the CPU flag means anything or if it's just easy way to default the appropriate CPU specific instructions (and optimal libraries) to ON.
 
Originally posted by Plaristocrates
For starters, this is an Intel compiler...why anyone would expect Intel to thoroughly test their compiler optimizations on AMD processors is beyond me.
Click to expand...

Why anyone would think that neither Intel or AMD has tested the crap out of their competitor's design is beyond me. Especially if that design offers some real performance advantages. Of course each company tests the others product(s), they test the crap out of them to see how they perform under a variety of conditions so they can understand what the level of competition is. Wouldn't you ?

Stuff happens, even on CPUs who should run basic x86 code identically...just look at the freezing problems Folding@Home has been having with AMD CPUs.
Click to expand...

This is an issue I have been personally following closely and it appears they may have a work around for it now. However that "work around" appears to be related to a modification of the Gromacs core code, meaning it's in Gromac's/Stanford's code and not an error in the later AMD Tbred and Hammer cores themselves. But until they have finished testing with the newer core, I'm sure we will not hear where the problem was. Even then I'm kinda doubting they will come clean publically and tell us what the deal was. But we'll see.
 
Someone is going to banned here...I feel it :(

As for the compiler issue, I don't see why it is creating such a hoopla. It is Intel's compiler and they can put any detection mechanisms they want in it. Besides, all the mechanism does is check if the right cpu is in place. All this just shows AMD cpu's can perform just as well as a P4 (except like SSE3 or something other instruction sets AMD doesn't support). AMD should really get around to making their own compiler :eek:
 
People, look at it another way:

Originally posted by Alf Alpha:

Put yourself in the shoes of a developer: you can either use this flag and add the increased complexity of managing and debugging multiple cpu-specific binaries or you can choose to ignore this flag and have a single binary.

This assures that -QxN will get rarely used. Intel must have some better reason to do this than pissing on AMD...
Click to expand...
 
Originally posted by blvdKing
Why not use the Microsoft of gcc compilier instead of the Intel compiler?
Click to expand...
If your target market has a CPU that sucks on non-optimized code, are you going to make your product look bad by compiling it with gcc? MSVC++ or VB is better, but ICC is still the best for optimizing to the P4 platform.
 
Originally posted by CIWS
Why anyone would think that neither Intel or AMD has tested the crap out of their competitor's design is beyond me. Especially if that design offers some real performance advantages. Of course each company tests the others product(s), they test the crap out of them to see how they perform under a variety of conditions so they can understand what the level of competition is. Wouldn't you ?
Click to expand...
I don't disagree with you...I didn't mean to imply that neither Intel nor AMD tests each others cores; of course they do. What I meant was that Intel probably doesn't test particular compiler optimizations to make sure that they work with AMD's cores, at least not with the kind of quality assurance they do on their own. If Intel were to have their compiler run the special code on AMD systems, they [Intel] would have to answer to their customers if the code turns out to break under certain conditions. I doubt that Intel wants to spend support time and money fixing issues with their competitor's products.

As for the Folding@Home thing, I have to confess that I only know the surface details, not the internals. I used it as an example of a case where code that was probably intended to run the same on all CPUs didn't. If I was mistaken I apologize.

DaveX: why would someone get banned? I'm pretty sure that the argument has more or less run its course, and the personal attacks were limited to just two people who are also over it now (or if they weren't, mod intervention no doubt fixed that :D ).
 
You must log in or register to reply here.
Back
Top