How big is the performance difference between AMD & Intel?

[OLpal]

I'm also hoping for the 790FX to be given the new SB750, but if not the 790GX isn't a bad alternative.. Really brings the X4 chips to OC life !!

Here is a link for you to see Call of Duty Intel vs. AMD X4
http://www.legitreviews.com/article/682/11/
also 3d-mark http://www.legitreviews.com/article/682/12/
PC MARK http://www.hothardware.com/Art...50_B3_Revision/?page=4
Gaming http://www.pcper.com/article.p...=539&type=expert&pid=8
As you can see $$ for $$ the AMD X4 matches with intel well , now with SB750 motherboards the OC will be more competitive for OC'n.

After having Q9300 & E8400 you would not be happy with X2 5000 BE speed of processor !!

Ol'Pal

Originally posted by: v3rax
Thanks for that.... So the SB750 SB isnt available on the 790 FX?

I do not need integrated graphics as I am running crossfire.

Is there any other difference between the 790GX and the 790FX other than the integrated graphics?

It seems that the 790FX and SB750 would be a great combo.. if that is going to happen.

 

[imported_nightfox]

Would the AMD tri-cores also OC better with SB750?

 

[Philippart]

most phenoms got benchmarked with ddr2 800 memory, however intel systems get mostly benchmarked with the faster ram. the phenoms develop their whole potential when they're run with ddr2 1066 memory (since they have an integrated memory controller)!

try to find benchmarks on the internet, you'll see the difference!

 

[OCGuy]

Originally posted by: Philippart
most phenoms got benchmarked with ddr2 800 memory, however intel systems get mostly benchmarked with the faster ram. the phenoms develop their whole potential when they're run with ddr2 1066 memory (since they have an integrated memory controller)!

try to find benchmarks on the internet, you'll see the difference!

Thats a new one

 

[Fox5]

Originally posted by: nightfox
Would the AMD tri-cores also OC better with SB750?

They should.
I'd imagine overclocked tri cores with the SB750 would come within striking distance of overclocked intel 6MB dual cores (within 10-25%) while having a whole extra core to work with. That extra core can easily account for 30%-40% more performance in apps that support a 3rd core.

 

[aigomorla]

Originally posted by: Fox5

Originally posted by: nightfox
Would the AMD tri-cores also OC better with SB750?

They should.
I'd imagine overclocked tri cores with the SB750 would come within striking distance of overclocked intel 6MB dual cores (within 10-25%) while having a whole extra core to work with. That extra core can easily account for 30%-40% more performance in apps that support a 3rd core.

wasnt the SB750 more aimed at phenom? i remember someone told us exactly what it was and i forgot. :\

i know im a bad mod to forget something this important! :[

 

[Tempered81]

intel > amd stock

intel > amd overclocked

this is "penryn > phenom" not "athlon > netburst"

 

[Fox5]

Originally posted by: aigomorla

Originally posted by: Fox5

Originally posted by: nightfox
Would the AMD tri-cores also OC better with SB750?

They should.
I'd imagine overclocked tri cores with the SB750 would come within striking distance of overclocked intel 6MB dual cores (within 10-25%) while having a whole extra core to work with. That extra core can easily account for 30%-40% more performance in apps that support a 3rd core.

wasnt the SB750 more aimed at phenom? i remember someone told us exactly what it was and i forgot. :\

i know im a bad mod to forget something this important! :[

SB750 controls some magical settings in phenom (probably related to timings) that allow for better overclockers at lower voltages.

Anyhow, the only AMD cpus I'd consider at the moment would be:
The cheapest X2, solely for max price/performance ratio, even after overclocking.
And the cheapest X3, for being able to come close to the single threaded performance of core 2 duos after overclocking on both, while having an extra core.

 

[dmens]

Originally posted by: Fox5
SB750 controls some magical settings in phenom (probably related to timings) that allow for better overclockers at lower voltages.

Anyhow, the only AMD cpus I'd consider at the moment would be:
The cheapest X2, solely for max price/performance ratio, even after overclocking.
And the cheapest X3, for being able to come close to the single threaded performance of core 2 duos after overclocking on both, while having an extra core.

nothing magical about it, most likely adjusting the PLL duty cycle. there's a reason the hack allows higher overclock percentages on low end parts than high end parts.

i personally think it is hilarious AMD exposed that setting to the public. it stinks of desperation. i particularly like the "documentation"... just call it a crapshoot already. +12/-12? LOL.

 

[zsdersw]

Originally posted by: perdomot
The forums aren't the only place that problems got reported. When I was looking to get a P35 mobo, I also checked the reviews at Newegg and saw many of the same problems reported there that I saw in the forums. Abit's board was one of the most complained about boards because of the booting issues although once it booted up, it seemed to perform fine. Go back a few pages or do a search under booting issues and you'll see quite a bit of discussion. Haven't seen much in the way of complaints about the P45 so far, so it might be a good reason to try one of those mobos. I'm running a 780G mobo right now but am tempted by the new 790GX series just released. Even though I wont need the raid 5 or ACC benefits for OCing, it still feels tempting and there are plenty of people who give in to temptation.

What you cannot determine from a cursory glance at forums or Newegg reviews is the idiocy of the person who posted a comment or review or how overly-ambitious they were trying to be. Everyone has an opinion, but that doesn't mean their opinion is relevant or significant.

Car forums are the same way. What you're describing is a Ford man (a guy who prefers Ford products) glancing at the Chevrolet forums, seeing a lot of posts about various problems, and then reinforcing his own conclusion that Ford must be better or that Chevrolet vehicles must particularly suck. What he doesn't realize is that all car forums have lots of problems listed on them.

 

[perdomot]

While its true that you cannot know how ambitious someone was trying to be, you can make note of the fact that people who bought a mobo where experiencing the same problems reported on the forums. If enough people are reporting problems, its illogical to ignore it just because some people are not experiencing the same problem. What really caught my eye the most when AMD started their Spider platform was their focus on how it wasn't just the individual parts that matter but how they work together that really makes or breaks the experience. Apple has been pushing their "it just works" philosophy for years and after all the issues I experienced, I have begun to see their point. What good is it to have a great cpu is the mobo it plugs into gives you grief. I tried two different mobos with two different chipsets from two different manufacturers and had issues. I haven't had problems with the AMD platform and I can't just pretend that it doesn't matter. How many different mobos does a guy have to go through before saying "enough"? The new P45 chipset seems to be a better product so perhaps I'll give it a try in the future.

 

[Markfw]

And how do you explain to 20 or so P35 motherboards from different mfg that I have used, and or setup, all with no problems ? Luck, but all these other people have problems ? whether its here or newegg, only the problems will be posted, not the successes.

Do you have any FACTS to support your position ?

 

[perdomot]

We seem to be getting away from the point of this thread which was about the difference in performance between intel and amd. As someone who has used both, my opinion is that they are equivilent if OCing is not going to be done. I also said that I had had problems with a C2D rig even when not OC'd and others had posted this as well in both forums and reviews. I have had no such problems on the AMD rig and believe it is worthwhile to say so. I understand others have had no problems with their rigs and am happy for them. My experience was not so good. Plain and simple. At Newegg, their reviews section has both good and bad reviews so I can't accept the idea that only problems are posted about.

 

[Idontcare]

Originally posted by: zsdersw
What you cannot determine from a cursory glance at forums or Newegg reviews is the idiocy of the person who posted a comment or review or how overly-ambitious they were trying to be. Everyone has an opinion, but that doesn't mean their opinion is relevant or significant.

Car forums are the same way. What you're describing is a Ford man (a guy who prefers Ford products) glancing at the Chevrolet forums, seeing a lot of posts about various problems, and then reinforcing his own conclusion that Ford must be better or that Chevrolet vehicles must particularly suck. What he doesn't realize is that all car forums have lots of problems listed on them.

This needs to be captured as part of a forum guideline stickied to the top...a statement of just simple common sense that only a minor percentage of vocal posters need gentle reminding of.

Sampling public opinion (be it by canvasing forum posts or Gallup polls) is never a substitute for actually generating data in a controlled environment. You get what you pay for, and freely accessible public opinion is nearly as worthless as purchased public opinions.

(disclaimer: the opinion expressed above is a freely given opinion )

Originally posted by: dmens
i personally think it is hilarious AMD exposed that setting to the public. it stinks of desperation. i particularly like the "documentation"... just call it a crapshoot already. +12/-12? LOL.

I have no doubt that such a move offends the engineering-driven side of your personality but surely you can agree that from an AMD shareholder wealth creation side of the financial world there is no such thing as stench of desperation when it comes to leveraging engineering tricks to increase the likelihood of increasing shareholder value.

You see this as an engineering trick of desperation...others will see it as a move (marketing driven or management driven) to do something that can be done to improve shareholder value. Nothing more.

Were Intel in a similar position and I were an Intel shareholder I would be more than a little irritated to learn that marketing's or management's hands were tied and prevented from releasing "features" to the market place because a few die-hard engineers felt it sullied their pride and self-respect to do so.

If you personally have a choice of swallowing your pride and supporting marketing's decision to release a "feature" versus your company electing to lay you off so as to increase shareholder value I suspect your sense of what is desperation versus straight forward business sense will decidedly shift in favor of preserving your personal revenue stream Mine would and has at times in the past.

 

[dmens]

Originally posted by: Idontcare
I have no doubt that such a move offends the engineering-driven side of your personality but surely you can agree that from an AMD shareholder wealth creation side of the financial world there is no such thing as stench of desperation when it comes to leveraging engineering tricks to increase the likelihood of increasing shareholder value.

You see this as an engineering trick of desperation...others will see it as a move (marketing driven or management driven) to do something that can be done to improve shareholder value. Nothing more.

Were Intel in a similar position and I were an Intel shareholder I would be more than a little irritated to learn that marketing's or management's hands were tied and prevented from releasing "features" to the market place because a few die-hard engineers felt it sullied their pride and self-respect to do so.

If you personally have a choice of swallowing your pride and supporting marketing's decision to release a "feature" versus your company electing to lay you off so as to increase shareholder value I suspect your sense of what is desperation versus straight forward business sense will decidedly shift in favor of preserving your personal revenue stream Mine would and has at times in the past.

actually i dont think it'll make a damn difference at all. the hack in question really only benefits overclockers who buy low-bin chips. those chips have low margins and the number of people who will actually buy a phenom because of such a feature is minuscule. so the only thing they really achieved is publicly display their desperation.

perhaps that is an overly cynical interpretation, but then again, it's not like i have not seen similar behavior from my pov, and i came to the same cynical conclusion as well.

 

[daveybrat]

Originally posted by: aigomorla

Originally posted by: Fox5

Originally posted by: nightfox
Would the AMD tri-cores also OC better with SB750?

They should.
I'd imagine overclocked tri cores with the SB750 would come within striking distance of overclocked intel 6MB dual cores (within 10-25%) while having a whole extra core to work with. That extra core can easily account for 30%-40% more performance in apps that support a 3rd core.

wasnt the SB750 more aimed at phenom? i remember someone told us exactly what it was and i forgot. :\

i know im a bad mod to forget something this important! :[

You should be able to overclock Tri-Cores just as well as Quad's in theory. The Tri-Core Amd processors are just new Phenom B3 cores with one core disabled.

 

[Idontcare]

Originally posted by: dmens
actually i dont think it'll make a damn difference at all. the hack in question really only benefits overclockers who buy low-bin chips. those chips have low margins and the number of people who will actually buy a phenom because of such a feature is minuscule. so the only thing they really achieved is publicly display their desperation.

perhaps that is an overly cynical interpretation, but then again, it's not like i have not seen similar behavior from my pov, and i came to the same cynical conclusion as well.

I would argue to not couple revenue with shareholder value in such tight fashion...anything that raises stockprice increases shareholder value but a rise in shareholder value does not hinge on revenue increasing.

If the feature in question does nothing more than generate a 1% increase in the positive buzz surrounding AMD then it will have done something to increase shareholder value (or when it all net nets out then maybe it just slows down the losses in shareholder value) and for AMD shareholders that is better than AMD management doing nothing while Rome is burning.

 

[OLpal]

As a CATCH 22
I own quite a bit of AMD stock of which intel has given us a terrible beating on !!
As a enthusiest & builder i'm using Core 2 Duo E8400 & E8600 for builds !!
When AMD brings the heat down on their CPU & it performs a little better !! [the SB 750 was a good start] i'll return to AMD once agin !!!

Originally posted by: Idontcare

Originally posted by: dmens
actually i dont think it'll make a damn difference at all. the hack in question really only benefits overclockers who buy low-bin chips. those chips have low margins and the number of people who will actually buy a phenom because of such a feature is minuscule. so the only thing they really achieved is publicly display their desperation.

perhaps that is an overly cynical interpretation, but then again, it's not like i have not seen similar behavior from my pov, and i came to the same cynical conclusion as well.

I would argue to not couple revenue with shareholder value in such tight fashion...anything that raises stockprice increases shareholder value but a rise in shareholder value does not hinge on revenue increasing.

If the feature in question does nothing more than generate a 1% increase in the positive buzz surrounding AMD then it will have done something to increase shareholder value (or when it all net nets out then maybe it just slows down the losses in shareholder value) and for AMD shareholders that is better than AMD management doing nothing while Rome is burning.

 

[BLaber]

Originally posted by: NTB
Just wondering...My brother asked me to price out a system for him today, and I haven't kept up much with hardware since I built my own C2D system last year. It doesn't make any difference to me - or him, I don't think - which one he gets; I'm just wondering if, given an Intel and an AMD system running at the same CPU speed, there would be a significant difference between the two.

Nathan

You really wana know diffrence at same speed then check these two links :

Intel Core2Quad 9450 C1 @ 2700MHz Vs AMD Phenom 9850 Black Edition B3 @ 2700 MHz

&

In this review we compare AMD´s latest AM2+ based Phenom X4 9850 to Intel´s mainstream Q9300 Processor

Hope this help you

 

[OLpal]

Yo Blaber, i like that Mad Shrimps test, kind of proves what i've been telling my friends !!

If they could just get their power usage & heat down... wow what a chip, & that wasn't even with the new SB 750 boards.... Good Job, Blaber !!

Didn't understand their cost analysis though !!!!

Originally posted by: BLaber

Originally posted by: NTB
Just wondering...My brother asked me to price out a system for him today, and I haven't kept up much with hardware since I built my own C2D system last year. It doesn't make any difference to me - or him, I don't think - which one he gets; I'm just wondering if, given an Intel and an AMD system running at the same CPU speed, there would be a significant difference between the two.

Nathan

You really wana know diffrence at same speed then check these two links :

Intel Core2Quad 9450 C1 @ 2700MHz Vs AMD Phenom 9850 Black Edition B3 @ 2700 MHz

&

In this review we compare AMD´s latest AM2+ based Phenom X4 9850 to Intel´s mainstream Q9300 Processor

Hope this help you

 

[BLaber]

Yeah dude , I agree cost analysis was done without any logic that one could understand.

If they(AMD) could just get their power usage & heat down...

OLpal , I agree 100% with you , if they can solve the overclocking and high wattage problems , heat is alright I think , then they are looking good.

By the way I expect AMD Deneb to be a huge improvement in terms of overclocking and wattage....

 

[BLaber]

Hey OLpal check this link , It ill make you SMILE , that is if you look at it from AMD's eyes

http://www.fudzilla.com/index....=view&id=9082&Itemid=1

 

[SlowSpyder]

Originally posted by: BLaber

Originally posted by: NTB
Just wondering...My brother asked me to price out a system for him today, and I haven't kept up much with hardware since I built my own C2D system last year. It doesn't make any difference to me - or him, I don't think - which one he gets; I'm just wondering if, given an Intel and an AMD system running at the same CPU speed, there would be a significant difference between the two.

Nathan

You really wana know diffrence at same speed then check these two links :

Intel Core2Quad 9450 C1 @ 2700MHz Vs AMD Phenom 9850 Black Edition B3 @ 2700 MHz

&

In this review we compare AMD´s latest AM2+ based Phenom X4 9850 to Intel´s mainstream Q9300 Processor

Hope this help you

Wow, I never saw that Madshrimps review before. The Phenom looks to pretty easily sit with Yorkfield for the most part, even clock for clock. I do wish they would have tested more than two games though. But, as I said in a different thread, it still doesn't matter, I still think AMD is screwed with anything branded 'Phenom'. Most people will see a review like that and just shrug it off. In so many peoples minds, C2D/Q = good, Phenom = bad regardless of any benches that may show otherwise. But, the one thing I do have to give Intel is that I bet a Q9300 can probably pretty routinely hit more the 3.0GHz, the Phenom 9850 at 3.0GHz is probably nearly as high as it'll go, generally speaking.

 

[harpoon84]

The problem with the Madshrimps review is that theres hardly any 'real world' apps being tested, its mainly just synthetics. No offence to the reviewer but SuperPi, wPrime, Everest, Sandra and 3DMark are totally useless benchmarks for evaluating real world performance. PCMark and Cinebench are more useful but are still synthetics at the end of the day. The only 'real' apps they ran was the x264 encoding and the 2 games. I'm sorry but thats a pretty poor review to be drawing any conclusions from.

Extremetech's review is more thorough, and they've tested 6 games, not 2. http://www.extremetech.com/art.../0,2845,2282395,00.asp

Tech-Reports CPU roundup is the best though, includes basically every existing Phenom and C2Q, and a few C2Ds and X2s as well. http://techreport.com/articles.x/14756