Ars: AMD may be irrelevant

[frozentundra123456]

...to maximize product exposure by drawing out reviews of it across a several day period? Do you seriously not comprehend the logic in that strategy?

Well, maybe they could release one game benchmark per day and draw it out over a week. That is the "logical" extension of that strategy. It may be "logical" to an AMD fan, but it is manipulative in my mind. I could even understand it more if they had done it a few months before full release, but to do it now just makes no sense to me.

And Ironically, if Toms reviews are correct, the CPU performance is actually decent. It is the GPU performance that dissuades me from buying the chip, since it is barely adequate for less demanding modern games at low resolutions and med quality settings. And if you dont game, any SB, IvB, or Llano igp is adequate.

 

[alexruiz]

Then give me one logical reason to release GPU benchmarks and withhold CPU benchmarks.

My suspicion is that they will focus the "CPU" side review on thew whole platform, and talk more about the Hudson-D4 chipset (A85x) than the cPU itself. 8 SATA ports 3 native surely sounds tempting. But, that is only a suspicion. It is hard to come with a good logical reason.

And why wouldn't Tom's reviews be correct? Lately, he has been the one whose reviews have been closer to the target (GTX660Ti for example) Or are you going to cling to the stone age and cite bias if well threaded benchmarks are used? If any, WE need better threading in the software.

 

[KingFatty]

And Ironically, if Toms reviews are correct, the CPU performance is actually decent. It is the GPU performance that dissuades me from buying the chip, since it is barely adequate for less demanding modern games at low resolutions and med quality settings. And if you dont game, any SB, IvB, or Llano igp is adequate.

Yes, this is what puzzles me.

The selective release of the NDA by AMD, should we judge it any differently based on the results? What I mean is, if they were "manipulating," to supposedly try to delay the release of "bad" information, would that still be true if they were delaying the release of "good" information (that made them look good), and releasing the "bad" information first?

In other words, is the anger with AMD over the selective release of the NDA, is that based on the actual information itself?

What if AMD had made the decision to release half the data early, and tossed a coin to decide whether to release the GPU or CPU data first. Is that tossing of a coin OK, does it excuse the behavior of selective early release of information, instead of hiding all the information until later?

Also, when you are talking about selectively releasing information, you know that all the information will get out eventually. So, let's say they are trying to hide something bad, won't that be even worse than the actual bad information? I've repeatedly heard that if there is something bad (e.g., a blowjob in the white house), it's much worse to try to lie about it or cover it up, that will blow up in your face (har har get it)

 

[Keysplayr]

Yes, this is what puzzles me.

The selective release of the NDA by AMD, should we judge it any differently based on the results? What I mean is, if they were "manipulating," to supposedly try to delay the release of "bad" information, would that still be true if they were delaying the release of "good" information (that made them look good), and releasing the "bad" information first?

In other words, is the anger with AMD over the selective release of the NDA, is that based on the actual information itself?

What if AMD had made the decision to release half the data early, and tossed a coin to decide whether to release the GPU or CPU data first. Is that tossing of a coin OK, does it excuse the behavior of selective early release of information, instead of hiding all the information until later?

Also, when you are talking about selectively releasing information, you know that all the information will get out eventually. So, let's say they are trying to hide something bad, won't that be even worse than the actual bad information? I've repeatedly heard that if there is something bad (e.g., a blowjob in the white house), it's much worse to try to lie about it or cover it up, that will blow up in your face (har har get it)

And you can ask "What if" all day long. It has been made clear the problem people have with this manipulation of reviews. Its' not like the GPU and CPU are separate entities. Same product. You can only report on half of it.
Seems legit.

 

[CHADBOGA]

Can you stop promoting AMD incessantly?

It is hard to break the habit of a lifetime. :awe:

 

[frozentundra123456]

My suspicion is that they will focus the "CPU" side review on thew whole platform, and talk more about the Hudson-D4 chipset (A85x) than the cPU itself. 8 SATA ports 3 native surely sounds tempting. But, that is only a suspicion. It is hard to come with a good logical reason.

And why wouldn't Tom's reviews be correct? Lately, he has been the one whose reviews have been closer to the target (GTX660Ti for example) Or are you going to cling to the stone age and cite bias if well threaded benchmarks are used? If any, WE need better threading in the software.

First, you may be correct about the platform, I am not that familiar with it. Can the same platform be used with Vishera and Steamroller? If so that is a plus. For Llano, since it is a mainstream to low end part, I dont know if all the extra features are a major plus for most users.

Secondly, I did not mean to imply that TH data was incorrect or biased. I only qualified it because it has not been verified by another source. I work in medical research and we do not take any data as valid until it has been verified by a second source. I think his data is correct, actually.

 

[frozentundra123456]

Yes, this is what puzzles me.

The selective release of the NDA by AMD, should we judge it any differently based on the results? What I mean is, if they were "manipulating," to supposedly try to delay the release of "bad" information, would that still be true if they were delaying the release of "good" information (that made them look good), and releasing the "bad" information first?

In other words, is the anger with AMD over the selective release of the NDA, is that based on the actual information itself?

What if AMD had made the decision to release half the data early, and tossed a coin to decide whether to release the GPU or CPU data first. Is that tossing of a coin OK, does it excuse the behavior of selective early release of information, instead of hiding all the information until later?

Also, when you are talking about selectively releasing information, you know that all the information will get out eventually. So, let's say they are trying to hide something bad, won't that be even worse than the actual bad information? I've repeatedly heard that if there is something bad (e.g., a blowjob in the white house), it's much worse to try to lie about it or cover it up, that will blow up in your face (har har get it)

My problem is with the process in general, that is allowing release of only part of the information. It doesnt matter whether the information they held back was good or bad. It is that the manufacturer controlled what the review sites published.

 

[Makaveli]

Just so we're clear guys, this isn't a proxy discussion on Andrew Cunningham's educational background or on the value of degrees in today's society. This is a technical forum so we ask that you please stick to the subject.

-Thanks
ViRGE

hehe's that's to bad I actually found that more interesting than what this thread is about

Here's hoping 2013 is a better year for AMD and brings more choices for us the consumer.

 

[KingFatty]

And you can ask "What if" all day long. It has been made clear the problem people have with this manipulation of reviews. Its' not like the GPU and CPU are separate entities. Same product. You can only report on half of it.
Seems legit.

So maybe that's where I'm a bit uninformed, how we equate manipulation of early selective release of information that is under NDA, with manipulation of the actual review?

I'm very opposed to manipulating the review, but that's like on one end of the spectrum. I tend to think it's OK to have an NDA where both parties agree to share a secret.

So I'm still having trouble where these two ends of the spectrum meet - whether it's OK to fiddle with the NDA, does that cross the line into actual tampering with results and/or dictating how a review can be conducted?

Sorry if this is getting into philosophical stuff or whatever, I don't mean to derail the thread, but these things are frustrating to think about because I'm still undecided and want more info/facts to help me decide.

 

[paperfist]

Still holding, I'll probably be dumb enough to hold all the way to zero. Yes, it's completely illogical, but at this point I can only lose a few grand on paper.

If they manage any kind of pop I'll be out though. Forget making a profit, they get back to $5 and I'm out.

What did you buy in at? Just wondering as I'm in the same boat.

 

[AtenRa]

Sorry if this is getting into philosophical stuff or whatever, I don't mean to derail the thread, but these things are frustrating to think about because I'm still undecided and want more info/facts to help me decide.

It is very simple,

People that dont want to see any AMD product having a big performance advantage against an Intel product are the ones that having a problem with the Trinity iGPU Previews.
Intel is the first of them all that doesnt like AMDs Trinity appear beating them like that in iGPU performance, they cant be the second best

 

[piesquared]

What did you buy in at? Just wondering as I'm in the same boat.

He/she forgot to mention that it is a short position.

 

[Phynaz]

Trinity appear beating them

That's the point. AMD is trying to appear like they are beating Intel by using BS marketing tactics.

I'll make you a bet, that in a wide variety of computing applications a quad core Trinity will not beat a quad core i5. I'll also say that it will have a hard time beating an i3 in a lot of applications. And it will consistently lose to an i3 in single / lightly threaded applications (not games).

What do you say, want to take me up on it? Let's even include power consumption, since you keep posting that over, and over, and over.

 

[inf64]

It's easy,just look up the intel chips in the same price category and you will find Trinity's product placement on the market.

Hint : i5 is not in 5800K price range.

 

[pelov]

If they haven't had low power designs and an SoC brewing for the past 3 years I'm inclined to say that yes, they're screwed.

 

[AtenRa]

That's the point. AMD is trying to appear like they are beating Intel by using BS marketing tactics.

I'll make you a bet, that in a wide variety of computing applications a quad core Trinity will not beat a quad core i5. I'll also say that it will have a hard time beating an i3 in a lot of applications. And it will consistently lose to an i3 in single / lightly threaded applications (not games).

What do you say, want to take me up on it? Let's even include power consumption, since you keep posting that over, and over, and over.

At default A10-5800K vs Core i3 Ivy it could be a draw, but Trinity has the advantage of OverClocking.

Core i5 is not in the same price segment.

 

[The Alias]

That's the point. AMD is trying to appear like they are beating Intel by using BS marketing tactics.

I'll make you a bet, that in a wide variety of computing applications a quad core Trinity will not beat a quad core i5. I'll also say that it will have a hard time beating an i3 in a lot of applications. And it will consistently lose to an i3 in single / lightly threaded applications (not games).

What do you say, want to take me up on it? Let's even include power consumption, since you keep posting that over, and over, and over.

http://www.tomshardware.com/reviews/a10-5800k-a8-5600k-trinity-apu,3241.html

didn't have a hard time beating the i3 that's for sure

 

[Phynaz]

http://www.tomshardware.com/reviews/a10-5800k-a8-5600k-trinity-apu,3241.html

didn't have a hard time beating the i3 that's for sure

Look at that, the i3 beats the A10 in VS compiles, Adobe Acrobat, Adobe Premier and iTunes. I stopped looking after that.

Really, if you are going to try to make a point, at least read your link before posting it. You made this just too easy.

 

[alexruiz]

Look at that, the i3 beats the A10 in VS compiles, Adobe Acrobat, Adobe Premier and iTunes. I stopped looking after that.

Really, if you are going to try to make a point, at least read your link before posting it. You made this just too easy.

The i3 loses in mainconcept, fritz, OCR, photoshop, 7-zip, solidworks, handbrake, 3dsmax, after effects... so?

Here it is for a better view
http://forums.anandtech.com/showpost.php?p=34037961&postcount=251

Edit: Why the insistence on "single threaded applications"? Are you still clinging to the stone age? If any, we need better threaded software, get all those cores to good use. Even if you use Intel, any application that doesn't take advantage of the full CPU capabilities should be avoided in real life.

 

[Phynaz]

Are we really arguing if AMD's brand new top of the line CPU beats Intel's last generation sub-mainstream CPU? Really? Some people here need to think a little bit harder.

AMD - "We price our products according to the value they provide". So even AMD admits their top of the line chip is low value.

 

[Phynaz]

Why the insistence on "single threaded applications"? Are you still clinging to the stone age?

Because many (most?) applications today aren't heavily multi-threaded (Duh!). You sound just like the AMD contingent when Derpdozer came out, "the future is multithreaded, then it will beat Intel".

Guess what? The future is here and Derpdozer is still a bad cpu. But yeah, let's cut half the cores off it and clock it a little faster, all of a sudden it's the savior AMD.

But go ahead, buy one and save a few percent off your system price. I'll continue to buy i5/i7 CPU's and use less power and get my work done faster. Performance is worth $ to most people.

 

[RampantAndroid]

The i3 loses in mainconcept, fritz, OCR, photoshop, 7-zip, solidworks, handbrake, 3dsmax, after effects... so?

Here it is for a better view
http://forums.anandtech.com/showpost.php?p=34037961&postcount=251

Edit: Why the insistence on "single threaded applications"? Are you still clinging to the stone age? If any, we need better threaded software, get all those cores to good use. Even if you use Intel, any application that doesn't take advantage of the full CPU capabilities should be avoided in real life.

The majority of apps today are still single threaded.

 

[pelov]

Are we really arguing if AMD's brand new top of the line CPU beats Intel's last generation sub-mainstream CPU? Really? Some people here need to think a little bit harder.

That would apply to you as well...

Since when is the FM2 line AMD's "top of the line" product? They're priced to compete with the i3's for a reason. You can't magically add $100 to the price tag to benefit your argument.

Furthermore, I think it'd be silly to buy an Ivy i3 over an A10-5800K at the same price. The advantage that Intel holds here isn't in the CPU performance -- which is actually in single and dual-threaded scenarios -- or the GPU performance (where it loses pretty badly) but rather in the platform and what it offers as a whole; whether that be Intel SRT or the option of buying a high end i7 upgrade. AMD has only offered moar SATA3 ports and moar USB 3.0 ports. Chipset wise, AMD has been really weak whereas Intel stepped it up with 1155.

Despite the on-die GPU being stuck in the awkward teenager stage, where it's not quite good enough to drive 1080p at good/high settings, it's still their better product with more potential (minus Bobcat which is apparently still selling like hotcakes).

AMD is destined to failure if they attempt to please you or me. Intel has stopped racing for IPC and clock speed while AMD chased moar coarz. Neither one of those is going to make them much money going forward, regardless of how many chips enthusiasts on this site buy.

 

[Phynaz]

Since when is the FM2 line AMD's "top of the line" product?

What models are AMD positioning higher?

 

[pelov]

Vishera? You've got 2-module to 4-module parts on AM3+. Not that I think it's going to be good processor. In fact I think it's going to underperform just like Bulldozer (minus the horrendous power consumption which Vishera should improve upon), but that doesn't matter.

I think a lot of you need to realize that neither Intel nor AMD are making enthusiast processors anymore. This notion that if either one of them fail on the desktop and that somehow that's going to run them out of business is quite funny. How many people buy desktops nowadays? And how many of them update their hardware frequently? How is that small percentage of users going to impact their sales, exactly? AMD still holds 30% market share on the desktop and they haven't been able to compete since AMD64 X2. That should tell you something...

Whether AMD can make it out of their slump depends on whether they can improve upon Bobcat and produce ULV chips in the 17W and under range. It has nothing to do with IPC, clock speed or cores.