Delusion and General Stupidity

[Arkaign]

http://www.amdzone.com/phpbb3/...opic.php?f=52&t=135170

Read the first page of this thread and try not to LOL more than 5 times or shake your head in astonished sadness.

I mean, I know there are gullible, stupid people out there. But ye gads, the extent that these people bend and dodge and lie to try to prop themselves up is incredible. They should go into politics.

I'm neutral in the CPU maker deal. I could care less about what's in my system or my customer's system, provided it was a good deal at the time (price vs. getting the same performance level from the competition, etc) and does the job well enough.

Historically, I've more often than not recommended AMD processors, but of course these days, they only offer any value at the absolute low end of the spectrum. That may change, hopefully, but it's how things stand for now.

Man, read some of the posts and then sit back in amazement at how far out some of those people are. The OP is about the only sane poster in that thread.

 

[dmens]

lol. i trolled there a couple times. those people wouldn't know real cpu design even if it were right in front of their face. got banned because i said that running a cache slower than it could made no sense, hahaha.

 

[Idontcare]

Originally posted by: dmens
lol. i trolled there a couple times. those people wouldn't know real cpu design even if it were right in front of their face. got banned because i said that running a cache slower than it could made no sense, hahaha.

While I have no idea what the background info is behind the story you are alluding to, and I certainly know less about cache designs than you knew 10 years ago about them, but I was under the impression there could be some legitimate thermal budget reasons to run an asynchonous L3 cache as a trade-off to boosting the core clockspeed (and the thermal budget of the cores).

Kind of an IPC/watt type optimization, obviously software mix dependent.

Would love to see a thread opened on the topic so we could discuss and so I (and others) can be educated to a level that exceeds my simple opinion.

 

[taltamir]

I guess it really comes down to what applications you run on your machine. For the things I do on my PC, Phenom at 2.2GHz is >= Q6600, which was my only other option at the time I bought it ($535 was well above my budget yet I needed a Quad Core).

HAHAHA, I can't stop laughing!

 

[Idontcare]

Originally quoted by: Thomas Brackett Reed (American Politician, 1839-1902)
?They never open their mouths without subtracting from the sum of human knowledge.?

 

[SexyK]

Look at the URL you pointed to... and you wonder why it's crawling with rabid fanboys? Color me not surprised at the content.

 

[AVP]

I only read the first sentence "as big an AMD supporter as I have been over the years, it's weird for me to contemplate that Intel is a better choice at the moment" and I stopped. I just don't get it.

It appears to me that fanboyism and intelligence are mutually exclusive. At least consumer intelligence...where do these people get off thinking that their undying support is relevant to anyone but themselves?

 

[dmens]

Originally posted by: Idontcare
While I have no idea what the background info is behind the story you are alluding to, and I certainly know less about cache designs than you knew 10 years ago about them, but I was under the impression there could be some legitimate thermal budget reasons to run an asynchonous L3 cache as a trade-off to boosting the core clockspeed (and the thermal budget of the cores).

Kind of an IPC/watt type optimization, obviously software mix dependent.

Would love to see a thread opened on the topic so we could discuss and so I (and others) can be educated to a level that exceeds my simple opinion.

it was in i.r.t. to 65nm K8 cache delay and AMD's marketing line saying that the reason the cache delay got worse from 90nm to 65nm was because of their desire for "K8 scalability" which makes no zero sense. i said it was just process scaling but amdzone would have none of that, lol. deliberately running a L2 cache slower won't win much on power, and will kill performance, so what's the point.

also 10 years ago i was in high school still so yeah no chance of knowing anything about anything back then.

 

[Cogman]

If you knew what field Cat-Barf was in, you'd understand that CPU performance differences quickly become obvious to him (in his field). He has always been an avid AMD supporter. Also, I don't really like the implication that anyone who may point out any AMD shortcomings is either easily brainwashed, or just plain ignorant. Blind fanboy-ism is quite easy to spot, and that is not what is being demonstrated here. I'm one of those people who are waiting for Phenom to mature before I buy, and I see nothing wrong with that at all.

Holy COW! Lol, that one made my day. Its like the blind leading the blind. Seriously, why is it so hard for some people to see that AMD screwed up with Phenom? I even lol at the blatant ignoring of benchmarks and then claiming everyone else on the internet is payed by intel to give them favorable reviews. lol. Ok, I'm going back to reading more of this stuff. This is priceless.

My impression is that the Phenom/Spider architecture is far more important than generally recognized. There are a lot of Phenom reviews, and they have certain similarities. For the most part Phenom performs about 10% behind the Q6600 at the same clock speed. That's the impression most people are going home with. However, occasionally the Phenom scores are extremely high. You would think reviewers would focus on these odd results and try to develop an explanation.

Lol, is this guy serious. "Yeah, reviewers are unfair because they don't focus on strange and abnormal results in benchmark tests. I think that anything that has AMD ahead should get TONS of spotlight to make them look better!!!!1!"

Really, What respectable reviewer would spend all his time trying to explain why something is odd when everything else says that it is more or less a fluke? If intel preforms 10% in 90% of the applications that people run, guess what, that means that for just about everyone intel will come out faster all the time. lol.

 

[zach0624]

where do these people get their numbers? scaling from 2 to 4 cores at over 20percent? i know where multiple cpus are involved barcelona is supposed to scale slightly better than conroe but i dont understand how they believe some of the crap that comes out of their mouths. Im a big amd fan and want phenom to do well but wow.

 

[sutahz]

Originally posted by: AussieFX from AMDZone
Phenom does have a price advantage 9500vs Q6600, then take into consideration the price of intel motherboards, DDR3 etc...

Yeah, because you have to buy DDR3 to user a Q6600 right? Retard(s). I've put together computers for people (in the newegg basket) to show them how much it would cost. Intel still had AMD beat on the most budget system I could build ($250). Intel motherboards are not that much more expensive if at all.

 

[Idontcare]

Originally posted by: dmens

Originally posted by: Idontcare
While I have no idea what the background info is behind the story you are alluding to, and I certainly know less about cache designs than you knew 10 years ago about them, but I was under the impression there could be some legitimate thermal budget reasons to run an asynchonous L3 cache as a trade-off to boosting the core clockspeed (and the thermal budget of the cores).

Kind of an IPC/watt type optimization, obviously software mix dependent.

Would love to see a thread opened on the topic so we could discuss and so I (and others) can be educated to a level that exceeds my simple opinion.

it was in i.r.t. to 65nm K8 cache delay and AMD's marketing line saying that the reason the cache delay got worse from 90nm to 65nm was because of their desire for "K8 scalability" which makes no zero sense. i said it was just process scaling but amdzone would have none of that, lol. deliberately running a L2 cache slower won't win much on power, and will kill performance, so what's the point.

also 10 years ago i was in high school still so yeah no chance of knowing anything about anything back then.

Ah, I'm with you now. Yeah the 65nm K8 L2$ thing had to be a resource allocation (the financial kind) versus deliverables timeline intersection (the marketing kind) type tradeoff.

I know I am not saying anything new to you, but in a perfect world (for AMD) AMD would have had the resources they needed to fund their shrink team above the minimum threshold necessary to complete the shrink in the timeline desired by marketing without foregoing performance deliverables such as L2$ latency.

 

[Aluvus]

Originally posted by: dmens
it was in i.r.t. to 65nm K8 cache delay and AMD's marketing line saying that the reason the cache delay got worse from 90nm to 65nm was because of their desire for "K8 scalability" which makes no zero sense. i said it was just process scaling but amdzone would have none of that, lol. deliberately running a L2 cache slower won't win much on power, and will kill performance, so what's the point.

also 10 years ago i was in high school still so yeah no chance of knowing anything about anything back then.

For those playing at home, the official explanation was that it was to give them breathing room to introduce chips with much larger L2 caches if needed. Whether that is true or not is of course open to debate.

http://www.anandtech.com/cpuch...howdoc.aspx?i=2893&p=3

They never did release Barcelona chips with drastically larger caches. So if that genuinely was the reasoning, it ultimately didn't do them any good.

 

[Idontcare]

Originally posted by: Aluvus

Originally posted by: dmens
it was in i.r.t. to 65nm K8 cache delay and AMD's marketing line saying that the reason the cache delay got worse from 90nm to 65nm was because of their desire for "K8 scalability" which makes no zero sense. i said it was just process scaling but amdzone would have none of that, lol. deliberately running a L2 cache slower won't win much on power, and will kill performance, so what's the point.

also 10 years ago i was in high school still so yeah no chance of knowing anything about anything back then.

For those playing at home, the official explanation was that it was to give them breathing room to introduce chips with much larger L2 caches if needed. Whether that is true or not is of course open to debate.

http://www.anandtech.com/cpuch...howdoc.aspx?i=2893&p=3

They never did release Barcelona chips with drastically larger caches. So if that genuinely was the reasoning, it ultimately didn't do them any good.

Kind of like the original arguments for K7 Mustang having 4MB L2 cache but then the product line rolled over into standard cachesize Palomino SKU's.

 

[heyheybooboo]

The IMC & cache issues have royally screwed AMD - first with AM2 and now Phenom. What was once an advantage went all to junk. For the benefit of everyone let's hope the new steppings create a more competitive environment ...

 

[The Boston Dangler]

my favorite:
AussieFX
"Where I live it's possible to get 50deg C ambient temps at this time of year and I have no trouble with heat and I'm oc'ed to 2.4 using 1.3v and a 6000+ cooler. If you want hot by all means get a C2D..."

for those of us in the civilized world, 50C equals 122F. throw another shrimp on the barbie, ya chucklehead.

 

[harpoon84]

I'm surprised Viditor isn't registered on those forums.

That AzmountAryl guy is quite funny, apparently whatever review site that doesn't show Phenom in a positive light is 'paid off' by Intel. Hehe...

 

[taltamir]

well... i better start a review site and rake in some of those big bucks intel gives out... since the phenom is obviously better and I would LOVE to be paid off for something...

 

[Idontcare]

Originally posted by: harpoon84
I'm surprised Viditor isn't registered on those forums.

Ah now that ain't necessary.

 

[Idontcare]

Originally posted by: The Boston Dangler
my favorite:
AussieFX
"Where I live it's possible to get 50deg C ambient temps at this time of year and I have no trouble with heat and I'm oc'ed to 2.4 using 1.3v and a 6000+ cooler. If you want hot by all means get a C2D..."

for those of us in the civilized world, 50C equals 122F. throw another shrimp on the barbie, ya chucklehead.

Somebody needs to buy these AMD chips, otherwise Nehalem is gonna be costing us $1k when they finally release it in 2010. So a few fanboys are delusional, its OK for folks to buy AMD chips and not have buyers remorse by whatever means they convince themselves they got a super deal. It's OK.

 

[Tempered81]

Originally posted by: Idontcare

Originally posted by: The Boston Dangler
my favorite:
AussieFX
"Where I live it's possible to get 50deg C ambient temps at this time of year and I have no trouble with heat and I'm oc'ed to 2.4 using 1.3v and a 6000+ cooler. If you want hot by all means get a C2D..."

for those of us in the civilized world, 50C equals 122F. throw another shrimp on the barbie, ya chucklehead.

Somebody needs to buy these AMD chips, otherwise Nehalem is gonna be costing us $1k when they finally release it in 2010. So a few fanboys are delusional, its OK for folks to buy AMD chips and not have buyers remorse by whatever means they convince themselves they got a super deal. It's OK.

yeah i've trolled amdzone for some good laughs in the past. (since c2d release)

@idontcare: anything new on the roadmap for core2 chips? Like die shrinks or ipc improvements? Is penryn the last stop before the first nahelem?

You guys think we will get a new penryn stepping?

 

[Angerisagift]

I'm with Idontcare on this, one, someone needs to buy AMD's chips so that Intel has to price competitively, just as long as it's not me.

though I'd be just as floored if AMD released a higher end offering that could compete with Penryn and Nehalem

 

[Idontcare]

Originally posted by: Angerisagift
I'm with Idontcare on this, one, someone needs to buy AMD's chips so that Intel has to price competitively, just as long as it's not me.

though I'd be just as floored if AMD released a higher end offering that could compete with Penryn and Nehalem

Even if Intel was a mediocre company without its intense focus and drive on engineering and market domination you'd still be asking for a company with 1/4 the resources to someone magically create a product with superior performance.

They did that once with the Athlon, they did it a second time with the K8. How many magic beans you expect the boys in green to have left in that bag?

I'm impressed with AMD's ability to do what they've done. To get anywhere close to 80% the results of Intel with 25% the investment is pretty impressive.

But that doesn't help us consumers when its a two-man field and one man is 4X bigger than the other. This isn't a contest. Score 1 for the Intel shareholder, score 0 for the consumer in 2010.

 

[Idontcare]

Originally posted by: jaredpace
anything new on the roadmap for core2 chips? Like die shrinks or ipc improvements? Is penryn the last stop before the first nahelem?

You guys think we will get a new penryn stepping?

You know thats a darn good question. Will they "Cedar Mill" the Core2 architecture into a 32nm "dumb shrink" and sell them at ridiculous cheap prices (<$60) like they did at 65nm?

I can see the financial rationalizations for such, no doubt Intel has already contemplated their options and downselected with far more information and insight than any of us will ever have.

Let me dispell any perceptions that I know anything more than anyone else on these internet forums. There are a few among us who do know, they know a lot more than they will ever let on, but I am not one of them insofar as Intel's product roadmaps are concerned.

Now will Intel respin another stepping of Penryn (i.e. 45nm)? I am not sure why they would, but typically the common reasons for respinning a stepping are (1) fixing gating errata, and (2) fixing yield (parametric and speed-bin) limiting issues.

So the question to ask ourselves is are there any emberrasing errata on current Penryn steppings or is there an unacceptable yield ceiling on current Penryn parts? We common folk won't know about yield issues.

G0 stepping at 65nm surely improved yields in that the reduced thermal profiles allowed more chips to have viable schmoo plots thru the sweet-spot of higher speed-bins.

Can we rationalize a thermal power argument for limiting current Penryn yields such that we could expect a respin with lower power consumption to boost the likelihood of higher speedbin chips? For 45nm I just don't see it from the current power-consumption plots floating around the web.

 

[Scottae]

Was a big fan of AMD but man after X2 came out ... nothing but SHEET really even for the price... just my 2cents and damn fanboys are sad... I would like nothing more than have AMD and Intel to be more on par in performance but they just arent