What do you think of AMD's Thuban

[DXcellence718]

You guys probably already heard the news, but now that amd's six core processor is coming out what are your thoughts. Personally im not really impressed, from what i've heard its only equivalent to an i7 quad but cost under $300. It is still the cheapest hex core a consumer can get, so i guess there still good for price/performance ratio(for a six core processor anyways)

 

[keithlm]

Now they are doing what they claimed was crap.

its funny to watch other companies eat their words.

The main reason AMD claimed MCM was garbage was because of how the cores communicated.

Unless you are claiming that AMD is now using that same method of core communication then AMD is not "eating their words".

Nice attempt but you failed. Please try again.

 

[fazers_on_stun]

Many Intel fans like to point out that Turbo generally always works and bumps the speed up one step even with all cores working.

So we have to ask: If there is rarely a time when the cores are running at base clock speed then what motivation would anybody ever have to claim base clock speed for any purpose other than trolling on forums or skewing benchmark results for marketing purposes?

Pointing out that simple fact is NOT nonsense as you are claiming. What is nonsense is people who know that base clock means nothing continually quoting it as a value that is somehow relevant in any conversation or benchmark. (Or even worse nonsense: Knowingly quoting base clock when you know the chip is running a single threaded application and would be running at a speed step even higher.)

Hmm, I haven't seen a large number of Intel fans claiming that, but then I'm not really looking for such a windmill to tilt at anyway 😛..

A blanket, one-size-fits-all-situations statement like "Turbo Boost is just a benchmarketing gimmick" is nonsense, as I correctly pointed out.

Interestingly enough, the "keithlm" posting over on AMDZone had this to say on the subject:

maduroutmb wrote:
This was a stupid marketing gimmick when Intel did it and that fact remains now that AMD has copied them. AMD Overdrive with Cool'n'Quiet are a much stronger set of tools for making the same thing happen, except that they do it in a far more controlled fashion and with predictable results. A version of AMD Overdrive for linux would be far more exciting to me than a "feature" that I am obligated to disable for the sake of stability.

by keithlm on Wed Mar 03, 2010 3:35 am

I concur, nice to see others share the opinion. But don't try to tell the Intel fans this... they'll go ballistic and their brains will explode.

The only real use that Intel had for the "turbo" feature was to allow forum posters ignore reality and pretend that their CPU's run at 2.66GHz. Thus they can claim that their 2.66Ghz chip was performing as well as the competition's chip running at 3.2Ghz. In truth for applications that use two cores or less their chips are really running at 3.2Ghz. But these fans really hate it if anyone points out that inconvenient fact. Especially since a great deal of benchmarks currently only use two cores.

If AMD does implement this trash into their chips and it happens to work better than Intel's gimmick, we'll see the current "must use it since it is enable by default" crowd change their mantra. At that point they will do an about face and want turbo disabled when doing comparative benchmarks. (Which I happen to agree with. But I will find it very humorous because they are currently soooooo adamant that it must remain enabled.)

So apparently that keithlm dismissed Turbo as "trash" and only useful for ignoring reality and pretending. It is that sort of dismissive logic that I was refuting as "nonsense", particularly so now that AMD has seen fit to include it with Thuban and Bulldozer.

 

[fazers_on_stun]

The main reason AMD claimed MCM was garbage was because of how the cores communicated.

Unless you are claiming that AMD is now using that same method of core communication then AMD is not "eating their words".

Nice attempt but you failed. Please try again.

No, it was AMD that failed to realize that MCM with the cores communicating over the FSB was more than adequate for the desktop, as shown by various & sundry benchmarks. And which fact let Intel beat AMD to market with a quad core by a year (or more if you discount Barcelona until the TLB bug fix version was released).

Being able to size up a market and deliver a product that is timely and eminently targeted to same is a skill that AMD heretofore has failed to demonstrate.

 

[Dekasav]

Being able to size up a market and deliver a product that is timely and eminently targeted to same is a skill that AMD heretofore has failed to demonstrate.

Wouldn't you say AMD has done this respectably well with a Phenom II X6 being released, Istanbul released early, and the entirety of the 45nm process? IMHO, AMD's change in leadership has been a very good thing. They're executing quite well (now).

 

[keithlm]

A blanket, one-size-fits-all-situations statement like "Turbo Boost is just a benchmarketing gimmick" is nonsense, as I correctly pointed out.

Actually you didn't "correctly point out" anything. You did state your opinion. Too bad that opinion doesn't make any sense when we look at the actual facts. (How often are people on this forum ridiculed for posting opinions that are not backed by facts? I guess you feel you can get away with it?)

You claim that Turbo-Boost is not a benchmarketing gimmick, and yet nobody can provide a reason why people willingly lie through their teeth by only quoting base clocks when they know full well it is blatantly false to do that. Okay, perhaps "lie" is a strong word. These people are merely taking advantage of the main benchmarketing benefits of Turbo-boost.

So apparently that keithlm dismissed Turbo as "trash" and only useful for ignoring reality and pretending. It is that sort of dismissive logic that I was refuting as "nonsense", particularly so now that AMD has seen fit to include it with Thuban and Bulldozer.

You are getting desperate; going to another forum to find opinions I posted over there in an attempt to negate something I posted here in this forum. Pathetic.

I should point out that I haven't recently posted that opinion in THIS forum because I know that many of the mindless drones on this forum would have their brains explode if I did post it here. You do realize that your posting of my opinion here doesn't negate anything I mentioned before?

You can't refute the minor base clock fact, so you attempt to pick on my personal opinion which is related to Turbo but does not actually pertain to the same thing.

You need to realize that the base clock quoting is fact. Would you like links to review websites so you can see them pretend the chips are running at their base clock speeds in their results? A completely separate issue is my opinion that Turbo-Boost from either company is a WORTHLESS benchmarketing tool. At least at the current time. Perhaps some day in the future Turbo might become a useful tool but at the current time it is only useful for doing one thing: Fooling people into ignoring reality; Most especially clueless Intel fanboys.

It is going to be fun to watch the circus on forums after AMD adds this benchmarketing gimmick to their own CPU and some of the people don't like what they see.


(EDIT: Please note I did add the word WORTHLESS I had originally omitted two paragraphs above... it changes the meaning of the paragraph. i.e., Being a benchmarketing tool is a fact, being a WORTHLESS benchmarketing tool is an opinion.)

 

[keithlm]

BTW: I also didn't address the unrelated concept you mentioned that it must not be a BenchMarketing Gimmick (tm) if AMD has now felt the need to add it to their CPU. (You and several others who have taken this up as the current Intel fanboy "talking point".)

DUH... Intel added something that fools people by giving results that can be conveniently interpreted incorrectly in their favor. Who would be stupid enough to actually think that AMD doesn't need to also add the same feature to skew results in the same manner to remain "competitive"?

 

[jennyh]

I agree with that last point. AMD had no option but to add their own turbo mode because of intels benchmarketing. We can argue about whether or not it's worth it, or if AMD 'copied' intel or whatever, simple fact is AMD were forced into it just so they can win a few benchmarks.

 

[fazers_on_stun]

Wouldn't you say AMD has done this respectably well with a Phenom II X6 being released, Istanbul released early, and the entirety of the 45nm process? IMHO, AMD's change in leadership has been a very good thing. They're executing quite well (now).

But they still haven't reclaimed the performance crown after 4 years of trying, plus the Thuban benches are not out yet so while I would agree they are performing much better than in the Barcelona days under Hector Ruiz, it remains to be seen whether they are "good enough" even now.

 

[Dekasav]

But they still haven't reclaimed the performance crown after 4 years of trying, plus the Thuban benches are not out yet so while I would agree they are performing much better than in the Barcelona days under Hector Ruiz, it remains to be seen whether they are "good enough" even now.

There is no reason AMD should EVER beat Intel. Ever. If Intel has its stuff together, they should always come out on top. There is a massive financial and marketshare gap between them. AMD even competing with Intel is amazing.

And I would definitely argue AMD is "good enough" even without Thuban. What's the share of desktop processors sold that are under $200 a pop? That'd be a really large share. Hence, AMD is good enough.

 

[jennyh]

You're forgetting the single most important thing Dekasav.

AMD has better engineers, always has done and always will. Intel's corruption just bought them time but the deal signed last year basically meant that was the end of that.

Intel is overconfident and has a short memory. Thuban will surprise just as many people as Magny Cours did.

 

[Dekasav]

You're forgetting the single most important thing Dekasav.

AMD has better engineers, always has done and always will. Intel's corruption just bought them time but the deal signed last year basically meant that was the end of that.

Intel is overconfident and has a short memory. Thuban will surprise just as many people as Magny Cours did.

Oh, I have no lack of faith in AMD's engineers, they're incredible. However, Intel can afford to design architectures with a much larger staff, and even a concurrent staff (two separate design teams pursue two different arches, drop the worse design after you figure out which will work better). Intel also has the money to make that 80-core processor just to do it, and find out the issues with multi/many-core arches.

AMD does an incredible job, but if Intel does as they should, Intel should always win.

 

[werxen]

Fazers + jennh = useless members.

Please get your own thread somewhere else or find each other on match.com
I'm pretty sure I'm not the only one that wants to watch you fanbois flamebait each other every other god damn thread.

 

[fazers_on_stun]

Actually you didn't "correctly point out" anything. You did state your opinion. Too bad that opinion doesn't make any sense when we look at the actual facts. (How often are people on this forum ridiculed for posting opinions that are not backed by facts? I guess you feel you can get away with it?)

Why so angry, Keith?? After all, you are the one posting "personal opinion" as fact over on AMDZone, which fact you are just now belatedly admitting to. Of course, over on AMDZone, any Intel-bashing post is required to be treated as factual; otherwise Ghost will perma-ban a challenger for failing to imbibe sufficient Koolaid.

OK, here are the facts as evidenced in these links. Just remember, in order to refute your "personal opinion", all that is needed is one fact to show that yes, turbo is not merely "trash" or a "benchmarketing gimmick" as you allege:

- Hardware Canucks shows anywhere from 6% to 23% performance boost with turbo on vs. turbo off for the i7-870.

- THG shows a 21% performance boost on the i5-750 on iTunes with turbo.

There's more but I'll leave that to you as an exercise in Internet searching.

Anyway, to summarize in case anybody here suffers from short-term memory dementia, the benchmarks disagree with you. People who have actually bought & used the technology as opposed to tooting mere personal opinions disagree with you. AMD itself also disagrees with you. I suppose we could say you are a disagreeable person .

You claim that Turbo-Boost is not a benchmarketing gimmick, and yet nobody can provide a reason why people willingly lie through their teeth by only quoting base clocks when they know full well it is blatantly false to do that. Okay, perhaps "lie" is a strong word. These people are merely taking advantage of the main benchmarketing benefits of Turbo-boost.

LOL - who cares?? Apparently you got up such a head of steam that your spectacles fogged over and missed my previous statement that I don't care about base clocks vs. turbo clocks. Go argue with somebody who does. You're merely tilting at some imaginary windmill here, Mr. Don Keith-ote 😀.

You are getting desperate; going to another forum to find opinions I posted over there in an attempt to negate something I posted here in this forum. Pathetic.

Oh c'mon & man up to take ownership of what you said in public. And as noted above, now they are "opinions" rather than fact. You are backpedaling faster than a unicyclist going downhill on a 90 degree slope .

I should point out that I haven't recently posted that opinion in THIS forum because I know that many of the mindless drones on this forum would have their brains explode if I did post it here. You do realize that your posting of my opinion here doesn't negate anything I mentioned before?

And also proves nothing. Hmm, "mindless drones" - sorta sounds like 'brainless sheep'. Now where have I heard that before...

You can't refute the minor base clock fact, so you attempt to pick on my personal opinion which is related to Turbo but does not actually pertain to the same thing.

O cast ye eyes a short distance above, Don Keith-ote. Yours truly, Sancho.

You need to realize that the base clock quoting is fact. Would you like links to review websites so you can see them pretend the chips are running at their base clock speeds in their results? A completely separate issue is my opinion that Turbo-Boost from either company is a WORTHLESS benchmarketing tool. At least at the current time. Perhaps some day in the future Turbo might become a useful tool but at the current time it is only useful for doing one thing: Fooling people into ignoring reality; Most especially clueless Intel fanboys.

Unfortunately for you, the facts speak otherwise. I guess we could ask JF-AMD for his opinion, seeing as how he is an AMD marketing director. Oh wait - you AMDZone drones, mindless or otherwise, drove him off your website...

It is going to be fun to watch the circus on forums after AMD adds this benchmarketing gimmick to their own CPU and some of the people don't like what they see.

Not at all - AMD merely realized that Intel innovated something quite useful, and proceeded to ape a rather primitive version of same.

 

[jennyh]

Fazers + jennh = useless members.

Please get your own thread somewhere else or find each other on match.com
I'm pretty sure I'm not the only one that wants to watch you fanbois flamebait each other every other god damn thread.

Who the hell cares what you think? Get a quad and get into the 2000's already mr backward-tech.

 

[werxen]

Who the hell cares what you think? Get a quad and get into the 2000's already mr backward-tech.

Who the hell cares what you think? Get an Intel processor and get into the 2000's already miss backward-AMD-tech.

 

[jennyh]

Not at all - AMD merely realized that Intel innovated something quite useful, and proceeded to ape a rather primitive version of same.

Useful? Useful in terms of benchmarketing maybe. No smart person would celebrate this feature as being something truly useful.

As for primitive version, LOL. 133mhz is good yes? :lol: , what does the i5 do, 400mhz best case? Yet Thuban does 500mhz best case and it's 'primitive' or 'less elegant'. Even though it does a lot more than intels laughable 'turbo', it's primitive? 😀

 

[Chad Boga]

Who the hell cares what you think? Get a quad and get into the 2000's already mr backward-tech.

Who the hell cares what you think? Get an Intel processor and get into the 2000's already miss backward-AMD-tech.

Well played to both of you. :lol: :lol: :lol:

 

[jennyh]

Who the hell cares what you think? Get an Intel processor and get into the 2000's already miss backward-AMD-tech.

This is my 3rd pc.

http://valid.canardpc.com/show_oc.php?id=1125246

 

[jennyh]

For some reason the board here as stuffed up my quotes, so the above makes no sense.

Nobody will notice any difference from your normal posts Chad.

 

[Chad Boga]

Nobody will notice any difference from your normal posts Chad.

:lol: :lol: :lol:

No one lacking a third digit in their IQ. 😀

 

[amdfangirl]

Fazers + jennh = useless members.

Yeh, well that's not as bad as amdfangirl going off-topic in every post. There's nobody more annoying than amdfangirl! Particularly if she starts talking about food!

 

[BadTrip]

Intel is in control

http://www.huffingtonpost.com/2010/04/08/mind-reading-brainscan-so_n_530009.html

 

[someguy7]

True AMD put 2 more cores with their quads on the same process node but Intel put s more cores at 65nm within the same TPD envelope with its C2Qs. So its not that amazing. And if I remember correctly, I don't think Intel used a new low-k dielectric to do that.

Either way, what they did is not what needs to impress. Its the performance that does.

This is not true jimmy. When Intel released the core2 quad it did not keep the same TDP. All of the dual core chips for rated for 65w. The quads where in excess of 100w. Even with the 45nm shrink the TDP went down to 95w.


And jennyh please just be quiet with that BS about AMD adding turbo to win some benchmarks. It is done to improve performance. That is it. It is so simple its flat out comical the crap you just said.

Let me make as clear and simple as possible. AMD's new chip will have 6 cores. Not many apps are going to actually load 6 cores. When 3 of those 6 cores are not active the other 3 will get boosted 400-500mhz. This is not a benchmark gimmick. It is there to make a task that is going to load 1,2 or 3 cores of the 6 500mhz faster. So you can have a very fast cpu for workloads that only use 1-3 cores and have the fastest cpu that AMD makes for apps that can use all the cores. Now the flagship chip will not get beat by the cheaper dual cores with the higher base clock.

Sure it will look better in benchmarks. Because it is better in the real world.

 

[Chad Boga]

And jennyh please just be quiet with that BS about AMD adding turbo to win some benchmarks. It is done to improve performance. That is it. It is so simple its flat out comical the crap you just said.
.
.
Sure it will look better in benchmarks. Because it is better in the real world.

Ever since Conroe came out and benchmarks didn't matter any more, the AMD faithful have been so traumatised by benchmarks, they just can't think straight about them any more.

 

[jimmysmitty]

I agree with that last point. AMD had no option but to add their own turbo mode because of intels benchmarketing. We can argue about whether or not it's worth it, or if AMD 'copied' intel or whatever, simple fact is AMD were forced into it just so they can win a few benchmarks.

You just follow whatever anyone says good about AMD, huh?

You're forgetting the single most important thing Dekasav.

AMD has better engineers, always has done and always will. Intel's corruption just bought them time but the deal signed last year basically meant that was the end of that.

Intel is overconfident and has a short memory. Thuban will surprise just as many people as Magny Cours did.

Better engineers? If that were true, Phenom 1 would have killed Core 2 Quad like they said, it didn't, HTT 3.0 would be faster than QPI, its not, and they would have had SMT and Turbo before Intel, they didn't.

MC didn't suprise anyone. Well no I am wrong, it did. People were amazed how 12 real cores didn't pound 12 fake cores. They barely beat them.

This is not true jimmy. When Intel released the core2 quad it did not keep the same TDP. All of the dual core chips for rated for 65w. The quads where in excess of 100w. Even with the 45nm shrink the TDP went down to 95w.


And jennyh please just be quiet with that BS about AMD adding turbo to win some benchmarks. It is done to improve performance. That is it. It is so simple its flat out comical the crap you just said.

Let me make as clear and simple as possible. AMD's new chip will have 6 cores. Not many apps are going to actually load 6 cores. When 3 of those 6 cores are not active the other 3 will get boosted 400-500mhz. This is not a benchmark gimmick. It is there to make a task that is going to load 1,2 or 3 cores of the 6 500mhz faster. So you can have a very fast cpu for workloads that only use 1-3 cores and have the fastest cpu that AMD makes for apps that can use all the cores. Now the flagship chip will not get beat by the cheaper dual cores with the higher base clock.

Sure it will look better in benchmarks. Because it is better in the real world.

Yea I forgot. 65w TDP because Pentium D was the first 65nm CPU. But 45nm was not the first 95w TDP quad. The Q6600 G0 was.

As for the benchmarks, if they make Intel look good, its benchmarketing. If they make AMD look good, its legit. Its a lose/lose vs Keith and jenny.......