Intel vs AMD: The Why

[amdfangirl]

Arguing things on forums has taught me to ignore people who insist they have a 37.6Ghz Core i7.

 

[SlitWeaver]

Because you can't continue scaling frequency. Other methods had to be found to increase performance.

Because I need my phone to have an i7-3770K in it ahaha...funny though, they make tablets with i7s in them...

 

[amdfangirl]

They are mostly rebranded desktop Core i3s that run with a lower clockspeed/voltage.

 

[PapaCrazy]

Funny that there's phones out there now with more cores than I have in my PC... that's kind of humiliating, actually. Damn smart phones.

 

[amdfangirl]

See, it's not that bad.

It's like PCs have mathematics professors whilst mobile phones have high school grads.

If you have the same number of maths professors and high school grads - you might have the same number of mathematicians, but it is clear which would be superior.

 

[SlitWeaver]

Funny that there's phones out there now with more cores than I have in my PC... that's kind of humiliating, actually. Damn smart phones.

:lol:

 

[PapaCrazy]

See, it's not that bad.

It's like PCs have mathematics professors whilst mobile phones have high school grads.

If you have the same number of maths professors and high school grads - you might have the same number of mathematicians, but it is clear which would be superior.

hah! awesome analogy. I can imagine the HS grads getting confused and giving up on certain problems... as has been my experience with phones.

 

[guanyu210379]

See, it's not that bad.
It's like PCs have mathematics professors whilst mobile phones have high school grads.
If you have the same number of maths professors and high school grads - you might have the same number of mathematicians, but it is clear which would be superior.

Watch out!
All professors were once high school grads.
LoL

 

[swordrage]

Correct me if I am wrong.. But as I understand with my consumer level knowledge, the architecture efficiency may work in the following way.

Consider two elementary students A and B.
"A" can do 5 calculations in a second, "B" can do 3 calculations per second. Thus "A" has a frequency of 5Hz, "B" has 3Hz.
"A" solves the problem 2+2+2 =? as "2+2=4, and 4+2=6". That is 2 separate steps of calculations. So to solve five such problems "A" will take 2 seconds (10 calculations, 2 for each problem)

While "B" has learnt multiplication and can do the same problem in one single step of calculation "3*2=6" and thus in 2 seconds he can solve 6 such problems. Thus a 3Hz brain can solve more problems than a 5Hz brain. And a 3.5GHz cpu can perform faster than a 4GHz processor.

But. when does "A" get the edge? If a problem like 2+3+5 comes. B CANNOT convert this to a multiplication problem and has to do 2 steps of addition and thus will fall behind "A" who is faster in addition problems. Similarly Intel can perform faster in tasks optimized for its architecture but sometimes lags behind AMD in tasks which are not optimized for it but requires brute force (more calculations per second) where more GHz matters.

Simply put, this should explain you WHY. I am not a Computer science guy and therefore am not aware of the microprocessor science and CPU architecture, but with my limited understanding I believe the above analogy.
Any suggestions, corrections are welcome.

And someone please tell me if I am correct. I will give myself a treat and offer a toast for you.
Cheers....

 

[8350rocks]

Far Cry 3, Crysis 3, Metro 2033 @ 1080p and 1440p Reviews:

http://teksyndicate.com/videos/amd-fx-8350-vs-intel-3570k-vs-3770k-vs-3820-gaming-and-xsplit-streaming-benchmarks

There's a gaming benchmark, on my way out the door, will add you some more links later.

 

[SlitWeaver]

Far Cry 3, Crysis 3, Metro 2033 @ 1080p and 1440p Reviews:

http://teksyndicate.com/videos/amd-fx-8350-vs-intel-3570k-vs-3770k-vs-3820-gaming-and-xsplit-streaming-benchmarks

There's a gaming benchmark, on my way out the door, will add you some more links later.

Thanks! So it would appear you'd need to select which CPU based on which games you're going to play; AMD some Intel others. Can't wait to see what other videos and benchmarks you link One thing I kniw though, my friend has the 8350 and it overheats pretty bad with the stock cooler. Also, more power consumption because of larger die

 

[iam2thecrowe]

Far Cry 3, Crysis 3, Metro 2033 @ 1080p and 1440p Reviews:

http://teksyndicate.com/videos/amd-fx-8350-vs-intel-3570k-vs-3770k-vs-3820-gaming-and-xsplit-streaming-benchmarks

There's a gaming benchmark, on my way out the door, will add you some more links later.

that is the most retarded review site i have ever seen.

 

[8350rocks]

Far Cry 3, Crysis 3, Metro 2033 @ 1080p and 1440p Reviews:

http://teksyndicate.com/videos/amd-fx-8350-vs-intel-3570k-vs-3770k-vs-3820-gaming-and-xsplit-streaming-benchmarks

There's a gaming benchmark, on my way out the door, will add you some more links later.

Thanks! So it would appear you'd need to select which CPU based on which games you're going to play; AMD some Intel others. Can't wait to see what other videos and benchmarks you link One thing I kniw though, my friend has the 8350 and it overheats pretty bad with the stock cooler. Also, more power consumption because of larger die

Yes, stock cooler is garbage...get a 212 EVO

http://www.overclock.net/t/1333027/amd-fx-8350-vs-i5-3570k-delidded-single-gpu-and-crossfire-gpu

That link shows some benches using FX8350 @ 4.8 GHz vs i5-3570k @ 4.8 GHz.

8350 wins 3 benches intel wins 3 benches.

@iam2thecrowe:

His method is sound and his tests are accurate...what else do you want from a benchmark? Someone who spews BS about Intel > AMD no matter the result?

 

[amuffin]

Far Cry 3, Crysis 3, Metro 2033 @ 1080p and 1440p Reviews:

http://teksyndicate.com/videos/amd-fx-8350-vs-intel-3570k-vs-3770k-vs-3820-gaming-and-xsplit-streaming-benchmarks

There's a gaming benchmark, on my way out the door, will add you some more links later.

that is the most retarded review site i have ever seen.

Um.

No.

It's the best review site out there! Look at how they list out the data and everything, makes anandtech, tom's, and everybody else look stupid.

 

[8350rocks]

I detect sarcasm, if you want a pretty spreadsheet and graph, he posted it in the comments on page 2. The test was run with a GTX 670 a few weeks later and the results were similar.

https://docs.google.com/spreadsheet/ccc?key=0AgHEgqqaasTkdGJmNmpNMGowa0hqX3VUNS1ZZEpaUlE

In case anyone was too lazy to fish for the more "professional" looking results.

 

[juanrga]

Introduction
The first thing you're thinking is that this is another "Intel is better than AMD" or "AMD is better than Intel" thread. Well, thank god it's not! I have a number of curiosities about the differences between Intel and AMD that I would love to have satisfied (as I'm sure many others do as well). I'm not sure where to begin, so we'll just jump right on in and hope for the best.

Notes before we begin...
Please don't post anything along the lines of "Intel is better because it is not AMD" or any cr*p like that. I'm looking for intelligible responses and factual/proven information.

Which is for who?
A lot of times I have seen AMD CPUs recommended to people who do things like Adobe and the like and Intel recommended to gamers (of course AMD is recommended to gamers too). I'm curious as to the reasoning behind these suggestions to PC builders. Can anyone link to some (recent) tests proving which is better in which case?

Intel is stronger. Why?
Now I know AMD users reading this section are already screaming out in outrage that I could say such a horrible thing; however, it's been proven with hundreds of thousands of hours worth of testing and benchmarking that Intel CPUs are more powerful than AMD CPUs (at least in gaming). Why is it that an the i7-3770K outperforms the FX-8350? Based just on the specs, the 8350 should be much more powerful than the 3770k; it has more cores and a much higher clock. Why, then, does the 3770k outperform the 8350? Does Intel use better materials? Do they place components in a more efficient way inside the CPU?

Benchmarks
-----
Source: Passmarks
3770K= 9,639
8350= 9,163
Difference= 476
-----
Source: CPUBOSS
Cinebench (all cores)
3770K= 25,703
8350= 32,437
Cinebench (single core)
3770K= 6,862
8350= 4,319
3D Mark 11
3770K= 8,470
8350= 6,980

1) AMD is good enough for gamers. If I was to simply to a minimum I would say that AMD FX is better for multitasking and modern parallel code, whereas Intel i7-3770k is better for older single threaded code.

2) There are several benchmarks showing how the FX-8350 beats an i7-3770k.

3) AMD chips are a victim of the Criple_AMD function found in Intel compilers which generate suboptimal code when detect a AMD chip. Cinebench is not optimized for AMD chips for instance.

4) AMD FX-8350 holds the world-wide record of overclocking, showing the excellent design and materials used.

5) PassMark scores are obtained from averaging users submissions. They do not compare chip to chip but an average-users-built to another average-users-built. Since the FX is a cheaper chip it is likely that most submission from FX owners are using cheaper parts (HDD, SDD, RAM, MOBO...) than Intel owners. Moreover, the difference in the score is of about a 5%, which is within the limits of error of the score.

 

[e56imfg]

Don't bring Logan into this. His testing methodology is beyond questionable and so are his results. He doesn't know what he is doing so please do not include him into this argument.

 

[amuffin]

Don't bring Logan into this. His testing methodology is beyond questionable and so are his results. He doesn't know what he is doing so please do not include him into this argument.

No, he has the best testing methodology. I don't know what you're talking about!

 

[8350rocks]

Don't bring Logan into this. His testing methodology is beyond questionable and so are his results. He doesn't know what he is doing so please do not include him into this argument.

But if intel had won those benchmarks it would be ok to include them right? Because then intel would have won...and you would say..."Well, it goes along with all the intel biased sites I read so, why not?"

Even Tom's shows the FX8350 is VERY close to the intel chips in MANY games...you can't argue over 3-5 FPS when the margin for error is 10% of 100-120 FPS...you're well below margin for error no matter how you slice it.

So, please...stop with your "holier than thou" attitude about reviews that do not specifically favor intel.

Can we have an intelligent discussion about merits and deficiencies without the intel crowd causing a raucous over the 2 being FAR closer in performance than they initially thought?

 

[amuffin]

Like you'll get anywhere with the username "8350rocks"

Anyway, I have a feeling this thread is going to turn to a flamewar.

 

[e56imfg]

Don't bring Logan into this. His testing methodology is beyond questionable and so are his results. He doesn't know what he is doing so please do not include him into this argument.

But if intel had won those benchmarks it would be ok to include them right? Because then intel would have won...and you would say..."Well, it goes along with all the intel biased sites I read so, why not?"

Even Tom's shows the FX8350 is VERY close to the intel chips in MANY games...you can't argue over 3-5 FPS when the margin for error is 10% of 100-120 FPS...you're well below margin for error no matter how you slice it.

So, please...stop with your "holier than thou" attitude about reviews that do not specifically favor intel.

Can we have an intelligent discussion about merits and deficiencies without the intel crowd causing a raucous over the 2 being FAR closer in performance than they initially thought?

No it's not about whoever is winning. His results are constantly inconsistent and there's no way in hell that the 8350 doubles the performance compared to the 3570K. Arma and Far Cry 3. I know the 8350 plays better with Far Cry 3 but everything on the web shows that the 3570K is only a few frames from the 8350. Give me some benchmarks that agree with Logan's ludicrous results. Unless Logan is the only one who is right and every other tech review site is wrong...

Where in this thread say that I favor Intel? Both CPUs should've been consistently neck-to-neck in all benchmarks. Logan's results are all over the place. I'm bashing on Logan, not AMD.

You can't have an intelligent argument with illogical proof. Give me one well known Intel biased site, I dare you.

 

[8350rocks]

Don't bring Logan into this. His testing methodology is beyond questionable and so are his results. He doesn't know what he is doing so please do not include him into this argument.

But if intel had won those benchmarks it would be ok to include them right? Because then intel would have won...and you would say..."Well, it goes along with all the intel biased sites I read so, why not?"

Even Tom's shows the FX8350 is VERY close to the intel chips in MANY games...you can't argue over 3-5 FPS when the margin for error is 10% of 100-120 FPS...you're well below margin for error no matter how you slice it.

So, please...stop with your "holier than thou" attitude about reviews that do not specifically favor intel.

Can we have an intelligent discussion about merits and deficiencies without the intel crowd causing a raucous over the 2 being FAR closer in performance than they initially thought?

No it's not about whoever is winning. His results are constantly inconsistent and there's no way in hell that the 8350 doubles the performance compared to the 3570K. Arma and Far Cry 3. I know the 8350 plays better with Far Cry 3 but everything on the web shows that the 3570K is only a few frames from the 8350. Give me some benchmarks that agree with Logan's ludicrous results. Unless Logan is the only one who is right and every other tech review site is wrong...

Where in this thread say that I favor Intel? Both CPUs should've been consistently neck-to-neck in all benchmarks. Logan's results are all over the place. I'm bashing on Logan, not AMD.

You can't have an intelligent argument with illogical proof. Give me one well known Intel biased site, I dare you.

Anandtech is intel biased...read their reviews...if you actually read the way they talk about AMD, you pick it up. They dismiss AMD like they were always going to be an "also ran" before it even started, and they downplay the strengths.

Second, Far Cry 3, in his review was 3 FPS difference, Crysis 3 was the only game where a dramatic difference existed, Metro 2033 was a maximum difference of 5-7 FPS in the FX8350s favor.

Where were those results skewed?

When he ran it with the GTX 670 the FPS in Far Cry 3 was a difference of 2 FPS in favor of intel being ahead.

Who is biased toward AMD now, or with poor scientific methods? Your results you cited were how close? A few FPS? Hmm...remarkably...they are similar! I wonder why that would be...?

 

[e56imfg]

Sorry but Anandtech is not Intel biased. That's ludicrous.

Logan didn't even bench Crysis 3... Did you even look at your own proof lol? Crysis 3 wasn't even released when he made these videos. I think you're referring to Warhead.
http://www.youtube.com/watch?v=4et7kDGSRfc&feature=player_detailpage#t=443s
http://www.youtube.com/watch?v=4et7kDGSRfc&feature=player_detailpage#t=433s

I'm not saying all of his results are wrong but some are just laughable at best. I would be just as suspicious if he said the 3570K had double or triple the FPS against the 8350.

I think someone should close this thread. This isn't even on topic.

 

[juanrga]

Sorry but Anandtech is not Intel biased. That's ludicrous.

I read many reviews from them and if it is not bias then it is plain ignorance plus lack of professionalism favouring certain brands.

Moreover, what 8350rocks is saying is well-known to google search:

 

[aonor]

AMD uses half core technology but they market the half cores as whole. So FX-8350 really just has the same amount as i7-3770k.

Source:
http://en.wikipedia.org/wiki/Ivy_Bridge_(microarchitecture)#Desktop_processors
http://en.wikipedia.org/wiki/Piledriver_(microarchitecture)#Desktop