Nvidia/AMD/VIA leave Sysmark company Bapco

Page 5 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Gikaseixas

Platinum Member
Jul 1, 2004
2,836
218
106
But sysmark is not intended to be the benchmark for the individual user.

That said, this is why AMD should have at the very least taken the time to cobble together some sort of coalition and beta-benchmark (heck even tesselation for the heaven benchmark was done on short notice and little leadtime) to coincide with their announcement yesterday.

It undermines them and their claims when they have nothing to present in terms of corroborating evidence of the sinister things in Sysmark2012 or the benefits of adopting their proposed alternative benchmark XYZ.

Instead we basically have AMD telling literally thousands of large businesses and government orgs that they should ignore the one benchmark that the industry has adopted as the de facto standard and use nothing in its place...uhm that is a non-starter for everyone, why even propose it if that is all you got?

If it's not for individual use, then Anandtech should remove it from it's tests. Also you're directing your criticism at AMD when in fact two other major players also left the boat. This alone tells you a lot (as already mentioned by other members here).
 
Last edited:

Gikaseixas

Platinum Member
Jul 1, 2004
2,836
218
106
I don't get why Idontcare caught so much flak in this thread... I thought most of his posts were very good.

and they're are indeed. His colaborations to this forums are very valuable but he's human and as such he is not right all the time.
 

jvroig

Platinum Member
Nov 4, 2009
2,394
1
81
If it's not for individual use, then Anandtech should remove it from it's tests.
1) It would be strange to not include an industry standard benchmark
2) Individual enthusiasts are not the only possible target of review sites

Also you're directing your criticism at AMD when in fact two other major players also left the boat. This alone tells you a lot (as already mentioned by other members here).
We can also possibly reasonably conclude that he is directing his criticisms towards (and only towards) AMD because NV and VIA have so far refused to do any finger-wagging PR war against BAPCo. As engadget reports:
engadget said:
On Monday, VIA and NVIDIA also joined the ranks of the recently defected, but refrained from any superfluous PR finger-wagging.

See, BAPCo themselves did the same thing. They fired back at AMD, not NV and VIA (yet). Why? Because the two haven't said anything yet. It's only AMD that has been vocal so far, so naturally only AMD gets an "answer".

When VIA and NVIDIA decide to air their own sets of dirty laundry, then we can have a more lively discussion involving all four parties, with criticisms directed squarely towards each of the four. Until then, it isn't surprising that only Intel and AMD are getting most of the attention here.
 

piesquared

Golden Member
Oct 16, 2006
1,651
473
136
What's the point of an industry benchmark if it isn't going to be representative of the hardware choices available on the market. If it leaves capabilities on the table, it's a major disservice to paying customers. And worse, customers are paying 100's if not 1000's of dollars to get that disservice. It's a gold mine for Integrate Electronics' and BAPCo's monopoly, and a con job for the paying customer.

From BSN, BAPCo Post Mortem: Is SYSmark actually INTELmark?



Uploaded with ImageShack.us
 

Gikaseixas

Platinum Member
Jul 1, 2004
2,836
218
106
We can also possibly reasonably conclude that he is directing his criticisms towards (and only towards) AMD because NV and VIA have so far refused to do any finger-wagging PR war against BAPCo

Valid points but...
Thread title clearly says "Nvidia/AMD/VIA leave Sysmark company Bapco". Let's discuss it all, not just part of the issue.
 

Arkadrel

Diamond Member
Oct 19, 2010
3,681
2
0
What's the point of an industry benchmark if it isn't going to be representative of the hardware choices available on the market. If it leaves capabilities on the table, it's a major disservice to paying customers. And worse, customers are paying 100's if not 1000's of dollars to get that disservice. It's a gold mine for Integrate Electronics' and BAPCo's monopoly, and a con job for the paying customer.

From BSN, BAPCo Post Mortem: Is SYSmark actually INTELmark?



Uploaded with ImageShack.us


*IF* they used newer versions of software, more of them would be GPU accelerated than your listing.


Basically most of the applications now / in future, will be GPU accelerated, its "odd" that they have 0 bareing on score.

Also that is just the TOP of the iceberg.


On top of that, the benchmarks are weighed in a way that supposedly favor Intel's offerings. The new SYSmark 2012 suite is said to contain 390 individual measurements, but only about 10 percent of those make a big impact on the overall score, that is then used to compare products.


There are two web browsers in the list, but both versions are now outdated.

So they wont be GPU accelerated, Intel wants it that way, so their CPUs score better.

Its odd to think that like 10 or more applications would be GPU accelerated, *if* they wanted (Intel doesnt, so it isnt), and yet non of these applications have any bareing on score.







Sysmark2012 = INTELmark.
 
Last edited:

Idontcare

Elite Member
Oct 10, 1999
21,110
64
91
If it's not for individual use, then Anandtech should remove it from it's tests. Also you're directing your criticism at AMD when in fact two other major players also left the boat. This alone tells you a lot (as already mentioned by other members here).

Valid points but...
Thread title clearly says "Nvidia/AMD/VIA leave Sysmark company Bapco". Let's discuss it all, not just part of the issue.

I've contributed a number of experience-based and knowledge-based posts detailing why this is a problem for AMD (since they are the one's making sure everyone knows it is a problem) and why it is a problem for the people who rely on Sysmark2012.

What tells us a lot is that rather than contribute something of comparable value to the thread topic you, and a select handful of others, have instead elected to make it personal and attempt to slime my character, imply that I am somehow intentionally hiding/spinning/etc something here in all this.

That's very telling. Don't like the message so you go after the messenger.

I've stuck with keeping my posts AMD/Intel centric because this issue impacts them to first-order, Via and NV are second-order in this situation. AMD made a big deal to ensure that their PR about this went to every one, NV and Via did not.

I did not create this delineation, these businesses did. Don't go after the messenger just because you don't like the sounds of his message.

And the thread title could be "Apple's/Pear's/Orange's are going up in price" and I could still restrict my posts to merely speaking about Apple's and it would be on-topic. Your innuendo, however, is not. It is a personal attack and I really hope the CPU mods do something about it because it is getting tiresome, not just for me but for the other members of the forum as well.
 

jvroig

Platinum Member
Nov 4, 2009
2,394
1
81
Thread title clearly says "Nvidia/AMD/VIA leave Sysmark company Bapco". Let's discuss it all, not just part of the issue.
I agree. But what's to discuss about (or what criticism can we levy against) NVIDIA and VIA given that they have not done any PR job on it yet? How can we take apart and examine NVIDIA and VIA's arguments and what they want to happen, when they haven't communicated anything yet?

AMD has. And so, as you can expect, we have talked about it. I said Intel is probably guilty of fixing the metric, which means I am inclined to believe Sysmark is more or less INTELmark. It just makes so much business sense. I also argued that leaving from BAPCo is an incomplete step - it would have had more positive reception (and an actual effect) if an alternative benchmark was made and presented by the three parties who left, instead of just having AMD (and so far only AMD) blog about it to complain to the world at large (to paraphrase Larry Ellison: "blogs just don't make up for good benchmarks").

This is why I see no conspiracy or fanboyism in the post/member that you were targeting. Three companies left. Only one "bitched" about it (so far). So BAPCo only responds to this one party. Idontcare similarly focused on this one company. It seems all above-board to me, because it still makes sense.

Now, when VIA and NV start complaining like AMD, and you see members still only hammering AMD for "bitching but not acting" (or whatever), then I would certainly be more inclined to believe there is some conspiracy / on-the-take / Intel payola thing going on here.

What I'm really trying to say here is that when we all exert a little more effort to be more logical, analytical, or more open to other possibilities (or all three, not really mutually exclusive), we can more often skip the accusations of "X is obviously being paid-off / fanboy", and instead try to discuss the actual topic. No different from what we (you and I) have done so here. You have not levied any accusations against me. I have not levied any accusations against you. We disagree on our conclusions, but in the process we have communicated what we are disagreeing on and why, and so far I believe we are doing a good job of communicating to each other the merits of our position. We do not have to end up agreeing with each other, but we also do not have to end up calling/labelling each other as fanboy of X.

It's simply a "net positive" all around when we do it like this, instead of when we do it the way you did when you first interacted with Idontcare.
 

Gikaseixas

Platinum Member
Jul 1, 2004
2,836
218
106
This is why I see no conspiracy or fanboyism in the post/member that you were targeting.

OMG i never implied anything close to that. Please do go back and read my initial input in this thread.

I'm of the opinion that by not having come up with an alternative benchmark on they way out, the parties involved are not at fault. We don't know what's behind it, what drove all 3 to get out at the same time. Don't you find it strange?

I understand where you coming from. AMD was very vocal by making it's point and therefore it's only natural to center the discussion around them. The direction i'm taking is to try and find out what's really behind it. The timimng, who got out, is there an alternative that we're not aware of??, etc.
 

lifeblood

Senior member
Oct 17, 2001
999
88
91
AMD had to leave the BAPCO, whether they have an alternate benchmark or not. AMD feels the weighting of the scores is deceptive. By staying in, the AMD trademark/emblem stayed on the box and marketing material making people believe AMD supported/agreed with the benchmark. By leaving, BAPCO must remove AMD's trademark.

Of course, once Intel comes up with a APU with with a decent GPU, BAPCO will change the weighting to lend more weight to GPU accelerated tasks.
 

piesquared

Golden Member
Oct 16, 2006
1,651
473
136
I've contributed a number of experience-based and knowledge-based posts detailing why this is a problem for AMD (since they are the one's making sure everyone knows it is a problem) and why it is a problem for the people who rely on Sysmark2012.

What tells us a lot is that rather than contribute something of comparable value to the thread topic you, and a select handful of others, have instead elected to make it personal and attempt to slime my character, imply that I am somehow intentionally hiding/spinning/etc something here in all this.

That's very telling. Don't like the message so you go after the messenger.

I've stuck with keeping my posts AMD/Intel centric because this issue impacts them to first-order, Via and NV are second-order in this situation. AMD made a big deal to ensure that their PR about this went to every one, NV and Via did not.

I did not create this delineation, these businesses did. Don't go after the messenger just because you don't like the sounds of his message.

And the thread title could be "Apple's/Pear's/Orange's are going up in price" and I could still restrict my posts to merely speaking about Apple's and it would be on-topic. Your innuendo, however, is not. It is a personal attack and I really hope the CPU mods do something about it because it is getting tiresome, not just for me but for the other members of the forum as well.

Your point about having an alternative prior to exposing BABCo is kind of irrelevant (the point is irrelevant, not you making the point). As AMD has a negligable marketshare in the targeted audience of this benchmark, what is there to lose? It seems they tried up until the 11th hour to have fair representation, and when that failed opted for the only alternative. I guess it was either that, or continue to be taken to the woodshed by BABCo and IntEl which does absolutely nothing for their image or prospective sales. Maybe now fewer customers will use Sysmark as a defacto tool for guaging purchasing decisions. Reviewers don't, but do the customers actually buying the products?

AMD had nothing to lose by this and everything to gain whether or not there is presently a one size fits all solution to analysing a products capabilities (which there isn't). They also had the most to lose out of the 3 that left, since one of their big target markets for Fusion is business so it's natural they would be the most vocal.

Lastly, there is an alternative to Sysmark. Phoronix Test Suite
 
Last edited:

Gikaseixas

Platinum Member
Jul 1, 2004
2,836
218
106
I've contributed a number of experience-based and knowledge-based posts detailing why this is a problem for AMD (since they are the one's making sure everyone knows it is a problem) and why it is a problem for the people who rely on Sysmark2012.

What tells us a lot is that rather than contribute something of comparable value to the thread topic you, and a select handful of others, have instead elected to make it personal and attempt to slime my character, imply that I am somehow intentionally hiding/spinning/etc something here in all this.

That's very telling. Don't like the message so you go after the messenger.

I've stuck with keeping my posts AMD/Intel centric because this issue impacts them to first-order, Via and NV are second-order in this situation. AMD made a big deal to ensure that their PR about this went to every one, NV and Via did not.

I did not create this delineation, these businesses did. Don't go after the messenger just because you don't like the sounds of his message.

And the thread title could be "Apple's/Pear's/Orange's are going up in price" and I could still restrict my posts to merely speaking about Apple's and it would be on-topic. Your innuendo, however, is not. It is a personal attack and I really hope the CPU mods do something about it because it is getting tiresome, not just for me but for the other members of the forum as well.

IDC do not feel targeted, specially by me cuz that's not the case. You might be feeling some pressure due to an increasing numbers of members not agreeing with your views but speaking for myself i just wanted to have a healthy discussion.

I don't agree with comments such as "not coming up with a replacement benchmark when leaving" so i have to challenge. I don't agree with you focusing at AMD even if they where the only ones to make a statement. This is a forum and by disagreeing we're not attacking you, targeting you, etc. Again at least i'm not.

I have the greatest respect from you, your knowledge only brings advantages to the forum but you're not perfect. Can't please everyone in the audience.
 

Madcatatlas

Golden Member
Feb 22, 2010
1,155
0
0
I've contributed a number of experience-based and knowledge-based posts detailing why this is a problem for AMD (since they are the one's making sure everyone knows it is a problem) and why it is a problem for the people who rely on Sysmark2012.

Well you`ve not really detailed, but made clear your opinion and interpretion. I disagree with your interpretation. I also object to what seems to be an effort by you and some other members, in this topic, to target AMD and AMD alone. It strikes me, and apparently others as well, as almost comical that one would to such an extent ignore the message or signal Nvidia and Via are broadcasting when they chose to leave WITH AMD.




What tells us a lot is that rather than contribute something of comparable value to the thread topic you, and a select handful of others, have instead elected to make it personal and attempt to slime my character, imply that I am somehow intentionally hiding/spinning/etc something here in all this.
I guess im on AMDs payroll or something now then. Selected even. I shouldnt bother with disclaimers and explanations for my choice of words i guess, after all they do so get ignored...
That's very telling. Don't like the message so you go after the messenger.
Thats right, i do not like the message as interpreted by you and i argument against the message as voiced by you. But its within your right to post your message, just as I believe it is within my right to post mine. Most people, including myself, have high respect for your character and integrity on this forum. Dont make that an issue. :)

I've stuck with keeping my posts AMD/Intel centric because this issue impacts them to first-order, Via and NV are second-order in this situation. AMD made a big deal to ensure that their PR about this went to every one, NV and Via did not.
That still does not explain why theres such a focus and insulting tone towards AMD and why theres such a lack of focus and attempt at getting a message from Nvidia and VIA. Bring them into the discussion and i promise you they will make a statement regarding their decision aswell.



Looking at the latest "info" on this, from BSN, i wonder why Nvidia declined from answering. Considering AMDs clear message that 0 of the 8 mentioned tests of GPU capabilities are weaved into the final score, what was it that Nvidia did not want to say?




AMD had to leave the BAPCO, whether they have an alternate benchmark or not. AMD feels the weighting of the scores is deceptive. By staying in, the AMD trademark/emblem stayed on the box and marketing material making people believe AMD supported/agreed with the benchmark. By leaving, BAPCO must remove AMD's trademark.

Of course, once Intel comes up with a APU with with a decent GPU, BAPCO will change the weighting to lend more weight to GPU accelerated tasks.


I am leaning towards this conclusion myself. Well put!
 
Last edited:

AnandThenMan

Diamond Member
Nov 11, 2004
3,991
627
126
And so, as you can expect, we have talked about it. I said Intel is probably guilty of fixing the metric, which means I am inclined to believe Sysmark is more or less INTELmark. It just makes so much business sense.
This "business sense" is also illegal. In fact, Intel is not permitted to use SYSmark as part of their sales pitch (neither is AMD, not they'd want to).

Let's not pretend otherwise here, SYSmark is heavily biased, and not a useful tool to gauge performance on a variety of platforms. If Anandtech continues to use it, then I would say the reviews here have lost all hope of being impartial.
 

podspi

Golden Member
Jan 11, 2011
1,982
102
106
This "business sense" is also illegal. In fact, Intel is not permitted to use SYSmark as part of their sales pitch (neither is AMD, not they'd want to).

Let's not pretend otherwise here, SYSmark is heavily biased, and not a useful tool to gauge performance on a variety of platforms. If Anandtech continues to use it, then I would say the reviews here have lost all hope of being impartial.


I've never heard anybody recommend SB over K10 because it was faster at sysmark. Usually it's Cinebench, or FPS in SCII, etc. The reason sysmark is so important is not because it makes reviews for enthusiasts biased one way or the other -- it is just one signal of performance in a sea of signals -- but because many organizations use sysmark as the end-all of benchmarks, and the sysmark scores are generated in such a way that % change in score != % change in performance. This skews things significantly when doing naive cost/benefit analysis, particularly when you are comparing AMD CPUs vs. Intel CPUs.

I still would love to know wtf Via left. I understand NV leaving (adding a discrete NV card doesn't really increase the sysmark, so goodbye any hope of getting into the business sector when the HD3000 is good enough), but why Via?! Maybe they didn't want to keep paying the membership fee? :D
 

Andres3605

Senior member
Nov 14, 2004
927
0
71
There has always been talks about intel influencing benchmarks in different ways to come on top,I remember an allegation about the use of crippled compilers in some benches as well last year and even in the sysmark 2002.


http://www.osnews.com/story/22683/I...quot_Cripple_AMD_quot_Function_from_Compiler_

"In fact, Fog points out that even benchmarking programs are affected by this, up to a point where benchmark results can differ greatly depending on how a processor identifies itself. Ars found out that by changing the CPUID of a VIA Nano processor to AuthenticAMD you could increase performance in PCMark 2005's memory subsystem test by 10% - changing it to GenuineIntel yields a 47.4% performance improvement! "
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
There has always been talks about intel influencing benchmarks in different ways to come on top,I remember an allegation about the use of crippled compilers in some benches as well last year and even in the sysmark 2002.


http://www.osnews.com/story/22683/I...quot_Cripple_AMD_quot_Function_from_Compiler_

"In fact, Fog points out that even benchmarking programs are affected by this, up to a point where benchmark results can differ greatly depending on how a processor identifies itself. Ars found out that by changing the CPUID of a VIA Nano processor to AuthenticAMD you could increase performance in PCMark 2005's memory subsystem test by 10% - changing it to GenuineIntel yields a 47.4% performance improvement! "


Why do people continue to use Intel's compiler? If I made a competing product, I'd make sure everyone knew that unlike Intel's compiler, which will cripple your software, ours is processor independent, and your program will run optimally on all systems. I'd produce examples of programs compiled on Intel's and on mine to demonstrate the superiority of my compiler. Do they pay off everyone? How do they get the industry to let them operate like this?
 

Lonyo

Lifer
Aug 10, 2002
21,938
6
81
Good news everyone! NV and VIA have both released statements too!

http://www.xbitlabs.com/news/other/..._Via_Technologies_Confirm_Quitting_BAPCo.html


VIA
"We have tendered our resignation to BAPCo. We strongly believe that the benchmarking applications tests developed for SYSmark 2012 and EEcoMark 2.0 do not accurately reflect real world PC usage scenarios and workloads and therefore feel we can no longer remain as a member of the organization"


NV
"We hope that the industry can adopt a much more open and transparent process for developing fair and objective benchmarks that accurately measure real-world PC performance and are committed to working with companies that share our vision"
 

jvroig

Platinum Member
Nov 4, 2009
2,394
1
81
This "business sense" is also illegal. In fact, Intel is not permitted to use SYSmark as part of their sales pitch (neither is AMD, not they'd want to). Let's not pretend otherwise here, SYSmark is heavily biased, and not a useful tool to gauge performance on a variety of platforms. If Anandtech continues to use it, then I would say the reviews here have lost all hope of being impartial.
Of course it is illegal, why else would I have kept saying Intel is "guilty" of it if it isn't some sort of crime or underhanded tactic?

It is nice that you have this view of sysmark. I would certainly agree with you, as an enthusiast. But the issue here is not about "enthusiasts" or the needs of enthusiasts. So as a person invovled in corporate IT, I would have to disagree - we need standards. They (Intel, AMD and friends) have been endorsing sysmark for the past decade, so we (industry) have used it for our own corporate and government purchasing purposes. It makes sense. That's why BAPCo was started in the first place, to provide such a standard we can use.

This resignation by AMD, NV and VIA will certainly have an effect - but only when something comes around to replace sysmark. Otherwise, despite their messages and complaints and protests about sysmark, government and corporate purchasers that have been using sysmark will almost assuredly continue to use sysmark, simply because despite the controversy, there is no alternative yet. (And yeah, these entities are known for being slow to adapt or change)

I have no idea why we are even arguing this. It is almost self-evident. I have not yet come across any serious government or corportate purchaser who does not need/want/use an industry standard benchmark to base or justify purchases on (the expected bribes notwithstanding; even after being bribed, there's still the matter of justification that is left, or at the very least, the matter of "covering thy ass").

As for Anandtech using sysmark, I do not know if they would adopt sysmark 2012, but they seem completely justified in adopting earlier versions, since it still has AMD's stamp of approval, and it is the de facto standard. No matter your feelings, they can't leave that out. The good news is that it is just one benchmark out of the 2 or 3 dozen that they use! Given the wealth of benchmarks, and how easy they have made it through CPU Bench to access and filter this info, I see no problem. Most of the other sites do the same thing anyway - they have a full suite of tests, because of the reality that no single benchmark will realistically capture the "difference between processors" when this difference greatly depends upon use case.

For example, for my own purposes, I completely skip most benches and just look at Cinebench single and multi-threaded, Excel, archiving, and game benchmarks. Those are the only things that matter to me personally, and upon which I base my "decisions" or musings. That's something you generally can't do as a gov't / corporate purchaser though. You can't put your justification as "Anandtech finds this a 25% better computing platform based on the Cinebench, Excel, archiving, and gaming benchmarks" (actually, that may fly in some companies, since we aren't talking of cookie-cutter types all over the world, but that's more the exception instead of the expected; I could definitely use that in our company). That's because Anandtech isn't an industry-recognized and industry-accepted benchmarking entity, neither is Tom's, [H], Xbit, Hardware Canucks, or whoever else you can think of. But BAPCo is (it was formed for that very purpose), and so their product is the industry standard for a lot of huge organizations and government.

I know what you are wishing for - a fairer metric that doesn't cheat AMD, and NV, and VIA. That's what I want as well. Why else would I have said "yep, Intel is guilty of fixing this bench, most likely"?

But just because I want a fairer metric doesn't mean I can just blindly believe (and declare to all) that suddenly sysmark is now "nothing" due to AMD and friends leaving. I'd be surprised if 3 years down the road these agencies and corporations using sysmark will have already "switched". For one thing, government and big business just don't move that fast. And for another thing, switch to what? Nobody has offered an alternative (even NV and VIA have not hinted at what companies should be using as a replacement), so nobody in those agencies and corporations are even thinking of switching (switching to what?).
 
Last edited:

jvroig

Platinum Member
Nov 4, 2009
2,394
1
81
Why do people continue to use Intel's compiler? If I made a competing product, I'd make sure everyone knew that unlike Intel's compiler, which will cripple your software, ours is processor independent, and your program will run optimally on all systems. I'd produce examples of programs compiled on Intel's and on mine to demonstrate the superiority of my compiler. Do they pay off everyone? How do they get the industry to let them operate like this?
The answer to that is the answer to why even AMD's own math library ends up screwing up AMD's own processors in some things (they use some intel libs): Intel made a very good compiler (well, actually, compiler + libraries). Programmers have different preferences, and some of them chose the intel compiler, probably some of them needed a specific performance characteristic it had, and a good number of them just don't care about this hardware issue (believe it or not, being a programmer does not immediately mean one is also a hardware enthusiast) in the same way that most companies don't care about supporting linux as a platform.

Personally, I prefer GCC, as it is the best cross-platform compiler in my experience (and the experience of my best friend as well, who uses compilers far more than me). But I know that GCC isn't the "best" in terms of performance in all cases (but it's pretty great in some, and it is consistently good even if it isn't the best in everything).

As to why someone hasn't taken advantage of the gold mine that is the "make a compiler better than Intel's compiler and market the heck out of it", it's because it is so much hard work (can you imagine the resources Intel hurled at that?), and it is something that fewer and fewer programmers are passionate enough about to make it a profitable market. There are tons of compilers around, that's for sure. It's just that not many of them actually end up to be on par with Intel's.
 
Last edited:

Keysplayr

Elite Member
Jan 16, 2003
21,219
56
91
Why do people continue to use Intel's compiler? If I made a competing product, I'd make sure everyone knew that unlike Intel's compiler, which will cripple your software, ours is processor independent, and your program will run optimally on all systems. I'd produce examples of programs compiled on Intel's and on mine to demonstrate the superiority of my compiler. Do they pay off everyone? How do they get the industry to let them operate like this?

This might be exactly why I think you wouldn't do anything of the sort.
You would produce examples to demonstrate your compilers superiority. Would you also allow samples to be released that shows your compilers inferiority? So with respect, your statement seems full of it and hard to swallow.. :D
 
Last edited:

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
This might be exactly why I think you wouldn't do anything of the sort.
You would produce examples to demonstrate your compilers superiority. Would you also allow samples to be released that shows your compilers inferiority? So with respect, your statement seems full of it and hard to swallow.. :D

Exactly how would I stop it? Assuming there were such examples.

You do realize when I say "me" I'm speaking figuratively? Personally, I couldn't produce a compiler if my life depended on it. Those who do though should have no problems demonstrating the bias, with Intel's. Would you use a compiler that you knew was crippling your program if it were running on something other than an Intel processor?
 

Idontcare

Elite Member
Oct 10, 1999
21,110
64
91
Why do people continue to use Intel's compiler? If I made a competing product, I'd make sure everyone knew that unlike Intel's compiler, which will cripple your software, ours is processor independent, and your program will run optimally on all systems. I'd produce examples of programs compiled on Intel's and on mine to demonstrate the superiority of my compiler. Do they pay off everyone? How do they get the industry to let them operate like this?

I'm not particularly sure why, may have to do with practical things on the business side of the equation such as site-wide licensing fees, speed of the compilers themselves, the utility of the rest of the compiler suite (diagnostics, tracing errors, feedback on dependencies, etc).

Personally I've never actually used Intel's compilers. I used GCC, Microsoft compilers, and Portland Group.

Portland Group is my favorite but only because it produces the fastest running code for my specific app of interest. Intel might produce even faster running code, but my app was running on AMD CPU's at the time and I simply wanted to avoid all the potential compiler funny business so I went with Portland Group.

If you are compiling x86 apps for Mac's then there is probably no pragmatic reason to avoid Intel's compilers.

I'm guessing that the truth is that people and businesses make their compiler choices based on mundane and pragmatic reasoning's, hence the truth of the situation is not very juicy and cannot be spun into a sexy catchy story, hence no one has bothered to publish one detailing the steamy backroom deals that go on to cajole people/businesses into using Intel's compilers (i.e. there probably are no steamy backroom deals...otherwise someone somewhere would be squealing about it to Theo and Fudo and Charlie).
 

podspi

Golden Member
Jan 11, 2011
1,982
102
106
Yea, people seem to forget how much better things are now then they were say, 20ish years ago.

I just found a manual from ~ '93 that states the program is ONLY compatible w/ Intel CPUs. Now, those types of incompatibility issues are a thing of the past (as far as I am aware of) but if most/all of your customers use Intel CPUs, and then Intel compiler creates the fastest code... It's a no-brainer :D