The Truth: i5 750 vs Phenom II 965 -Updated

[AMW1011]

Hello
There has been a lot of heated debate and arguing on these forums about which is better, a Phenom II 965 or an i5 750. On both sides, everyone has failed to back up their point and have posted one or two links at best which is inadequate to answer such a question. I am here to answer that question. The scientific method tells us to do everything in, at least, threes. Three sources, three experiments, three variables, ect. Well I have compiled 8 different sources. None were picked and chosen except for the Lost Circuits article since it keeps getting thrown around.

I must remind everyone that this is a thread to inform those who are trying to choose between the two, I'm not here to settle any vendettas or disputes, I don't give a rats ass about that.

Alright let us discuss Turbo mode. Turbo Mode is where, with 1-2 cores under load, the CPU clocks those cores to 3.2 GHz and with 3-4 core under load the CPU clocks to 2.8 GHz. With Turbo mode enabled, which it is in all the below benchmarks, the i5 750 clocks at 2.8 GHz for most of the app benchmarks, which is 600 MHz (2.8GHz vs 3.4 GHz) lower than the Phenom II 965. In most games the i5 750 is clocked at 3.2 GHz or 2.8 GHz which is 200 MHz or 600 MHz (3.2GHz/2.8GHz vs 3.4 GHz) lower than the Phenom II 965. A reader must take this into account with the below results.

Links:
http://www.techspot.com/review/193-intel-core-i5-750/
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3634
http://www.guru3d.com/article/core-i5-750-core-i7-860-870-processor-review-test/
http://www.overclockersclub.com/reviews/intel_corei5750_corei7870/
http://www.neoseeker.com/Articles/Hardware/Reviews/lynnfield/
http://www.bit-tech.net/hardware/cpus/2009/09/08/intel-core-i5-and-i7-lynnfield-cpu-review/1
http://www.tomshardware.com/reviews/intel-core-i5,2410.html
http://www.lostcircuits.com/mambo//index.php?option=com_content&task=view&id=63&Itemid=42

Note: No synthetics were listed here, though I think that they have some merit, the majority opinion is that they do not and therefor they are not included.

Here we go: (continued on next post)

2/24/2010 I have updated the pricing section and some facts. If I'm asked I will add some more articles that are more recent.

 

[randomizer]

Claiming that everything is neutered on AMD processors would be showing just as much ignorance as claiming that nothing is neutered and that all of this is a non-issue.

Saying the latter would be disagreeing with the FTC investigation (even if it is only "alleged'), and even independent statements from organisations like Pande Group.

 

[keithlm]

When you take two cpu's closely matched like the i5 750 and 965 BE, it becomes clear just how much that "10%" is really worth - if it can turn a losing cpu into a 'winning' one - because that is exactly what the i5 is compared to a 965.

I'm going to have to call you out on this:

The i5-750 is more appropriately matched with the PhII 940BE and not the 965BE. (since out-of-the-box they are more equal.)

The 965BE eats the i5 for breakfast and spits out the bones.

 

[AMW1011]

You need to get a better understanding of the situation. Not gonna say any more AMW, I have no problem with you and I wouldn't want your lack of understanding of the subject to change that.

I have done extensive reading on the subject Jenny, I appreciate the respect and apologize for not giving it equally.

I do not argue that the compiler disables SSE and SSE2 on non-Intel processors, I am just pointing out that this mostly effects synthetics and very few real-world apps. The only one I can find is LAME, which gives a substantial lead to Intel architecture, this does not change the whole picture.

You also know that I do not go around recommending the i5 750 because of some imperceptible performance advantage, but it is there even if it is irrelevant.

I invite you to give me proof that any one of the benchmarks that I have provided uses the Intel compiler in the test. I will then adjust accordingly.

I will also point out that the compiler is a moot point in games, where the i5 750 still has an advantage, albeit even more irrelevant.

As for keithlm, just go away if you are just going to spew biased statements that are based on no facts and are irrelevant, this FYI is the definition of trolling.

 

[Raidur]

Did ya miss page 1 keith?

There should be a rule against spreading false information.

 

[keithlm]

Did you miss page 1 keith?

There should be a rule against spreading false information.

Good think that I'm only mentioning correct information then.

http://cid-4ff12cafc7aafd28.skydrive.live.com/self.aspx/.Public/merge-6106.pdf

If you ignore the memory and disk benchmarks the i5-750 doesn't have a chance against a 940. Oh well. (But then I'm sure some people would include the non-relevant memory and disk stuff so they can find a different overall outcome.)

As for keithlm, just go away if you are just going to spew biased statements that are based on no facts and are irrelevant, this FYI is the definition of trolling.

You would be correct IF I was mentioning something not based on facts. But since I am then you are not correct.

 

[AMW1011]

You would be correct IF I was mentioning something not based on facts. But since I am then you are not correct.

Provide real sources and quantifiable trends then we can talk about facts.

I can say that "the Pontiac G8 GXP EATS THE Nissan 370z FOR BREAKFAST!!"

I have provided no quanitifiable information or sources there.

And much like your comparison both cars are pretty evenly matched and both have their own distinct advantages, much like the 965 and i5 750.

 

[keithlm]

Provide real sources and quantifiable trends then we can talk about facts.

I can say that "the Pontiac G8 GXP EATS THE Nissan 370z FOR BREAKFAST!!"

I have provided no quanitifiable information or sources there.

And much like your comparison both cars are pretty evenly matched and both have their own distinct advantages, much like the 965 and i5 750.

Actually your right.. the link I gave above only shows the PhII 940 against the i7-750.

Here is the PhII965 at 3.4Ghz (And a few other results also.)

Sorry if it rains on your parade since it show the PhII965 being closer in results to an i7 clocked at 3.6Ghz than to a stock i5. But oh well.

http://cid-4ff12cafc7aafd28.skydrive.live.com/self.aspx/.Public/merge-8292.pdf



OH WAIT: I understand... I didn't include any game benchmarks and I only included benchmarks that people incorrectly assume that the i5/i7 always win at... that could be why some people are a bit confused. Oh well.

 

[AMW1011]

Actually your right.. the link I gave above only shows the PhII 940 against the i7-750.

Here is the PhII965 at 3.4Ghz (And a few other results also.)

Sorry if it rains on your parade since it show the PhII965 being closer in results to an i7 clocked at 3.6Ghz than to a stock i5. But oh well.

http://cid-4ff12cafc7aafd28.skydrive.live.com/self.aspx/.Public/merge-8292.pdf



OH WAIT: I understand... I didn't include any game benchmarks and I only included benchmarks that people incorrectly assume that the i5/i7 always win at... that could be why some people are a bit confused. Oh well.

That source is far from credible. It has obvious... holes, to say the least.

First the i5 750 will perform the same as the i7 920 in most of the benchmarks as the only difference is hyperthreading, which is almost totally unquantifiable in benchmarks.

Second, many of the benchmarks in there have very different results than the ones I've compiled, the ones by proffessional review sites.

Third, and my favorite, none of those benchmarks were taken on a windows based system but on mostly Linux-based systems and possibly a few OS X based systems:
http://www.phoronix-test-suite.com/

These results are completely irrelevant to the discussion, but based on them I will admit that AMD seems to beat Intel in a Linux environment, which is great for 1.02% of the market.

 

[bystander]

Actually at 3.2 Ghz and up the Phenom II is pretty much equal to or better than an i7 when it comes to games. Scaling for the i7 is also pretty bad as a 4Ghz PhII is just as fast/faster than a 4Ghz i7 in games. Proof:
http://www.legionhardware.com/articles_pages/cpu_scaling_with_the_radeon_hd_5970,1.html

You should check out this review...and all the benchmarks: http://www.xbitlabs.com/articles/video/display/radeon-hd5870-cpu-scaling.html

The thing is, most of these benchmarks are done on systems that are GPU bound. Look at what happens with 5870's in CF as they overclock the CPU's. Even then, you can see a lot of games end up GPU bound.

The i7 does out perform the 965, it just that it takes a lot of GPU muscle to make it apparent.

 

[Raidur]

Are you implying that we ignore the countless benchmarks from page 1 keith?

 

[randomizer]

If anyone here has a Phenom II system and Core i5 system with comparable hardware, they could simply download the source code for the open source programs, compile with GCC/G++ without Intel libraries, and then benchmark them. Then we'd have an unbiased set of results and there would be less posts dismissing benchmarks and more posts talking about the benchmarks.

 

[smithereen]

You should check out this review...and all the benchmarks: http://www.xbitlabs.com/articles/video/display/radeon-hd5870-cpu-scaling.html

The thing is, most of these benchmarks are done on systems that are GPU bound. Look at what happens with 5870's in CF as they overclock the CPU's. Even then, you can see a lot of games end up GPU bound.

The i7 does out perform the 965, it just that it takes a lot of GPU muscle to make it apparent.

I'd say the i5s/i7s outperform the 965 more consistently when there is a lot of GPU muscle available.

 

[Raidur]

@bystander

Sadly that benchmark has am2+/ddr2/8x8x, which (according to the amd fanboys) makes it unusable.

I however only see those things hurting performance by 0-5% (10 at MOST). So to me they work, but not to some people I guess.

We'll just have to wait and see. I'd bet money that we'd see near the same results though.

 

[yannifb]

You should check out this review...and all the benchmarks: http://www.xbitlabs.com/articles/video/display/radeon-hd5870-cpu-scaling.html

The thing is, most of these benchmarks are done on systems that are GPU bound. Look at what happens with 5870's in CF as they overclock the CPU's. Even then, you can see a lot of games end up GPU bound.

The i7 does out perform the 965, it just that it takes a lot of GPU muscle to make it apparent.

Yeah i see that but remember that catalyst driver update that greatly helped AMD cpu's with crossfire configs? That actually brought them a lot closer to the i7s. But yes Im not denying the i7 is better, but still not by too much (except for in far cry and HAWX) especially in 1920 by 1080 and up. But the only thing that annoys me is when people say no one plays at resolutions like 1200p and up... if you spend money on 5870 cfx or 5970's why the hell wouldn't you play at high resolutions?

 

[JAYDEEJOHN]

This is why Im waiting for the FTC.
Theres no way anyone can "guess" as to what the FTC may do to Intel here, or what theyll find, or how theyll go about it all etc.
It may be nothng, it may be huge, and like I said, we dont know the numbers, and Im betting the FTC will have better numbers than we will ever have, and the higher the numbers, the worse itll be for Intel, lower, the opposite, or better itll be.
Point to understand is, thru SW, Intel has made their cpus better, not thru HW only, and whether this approach is legal is for the FTC to rule on

 

[keithlm]

That source is far from credible. It has obvious... holes, to say the least.

First the i5 750 will perform the same as the i7 920 in most of the benchmarks as the only difference is hyperthreading, which is almost totally unquantifiable in benchmarks.

Second, many of the benchmarks in there have very different results than the ones I've compiled, the ones by proffessional review sites.

Third, and my favorite, none of those benchmarks were taken on a windows based system but on mostly Linux-based systems and possibly a few OS X based systems:
http://www.phoronix-test-suite.com/

These results are completely irrelevant to the discussion, but based on them I will admit that AMD seems to beat Intel in a Linux environment, which is great for 1.02% of the market.

Let me address your points:

First: Not really imporant or relevant. Not sure why you even mention it.

Second: Do I care? So because the results don't agree with what you believe you want to just ignore them.

Third: The least important one of your points. So because it is not windows and is Linux we are supposed to ignore the results. Somehow if it is Linux and compiled using GCC individually on each test system it must be invalid in some way?

You appear to be grasping at anything because these results do not agree with your opinion.

You know... you're doing the same thing that people were ridiculed for if they didn't agree with any of your posted results. You'll go in and start picking at anything possible. Guess you don't like it when the tables are turned. But rest assured that I won't treat you the same way that people tht didn't agree with your links were treated.

Nevertheless the i5/i7 does not perform as well on non-game applications as the mythology propagated on various forums mistakenly claims.


EDIT: I know many people ignore 3dMark Vantage but the GPU score for the PhII is between 1500 and 3000 points HIGHER than the i5/i7 at the same clock speeds for Crossfired systems showing that the PhII at those speeds performs much better than the i5/i7 systems. But since Vantage is synthetic we have to ignore any results from it.

 

[AMW1011]

Let me address your points:

First: Not really imporant or relevant. Not sure why you even mention it.

Second: Do I care? So because the results don't agree with what you believe you want to just ignore them.

Third: The least important one of your points. So because it is not windows and is Linux we are supposed to ignore the results. Somehow if it is Linux and compiled using GCC individually on each test system it must be invalid in some way?

You appear to be grasping at anything because these results do not agree with your opinion.

You know... you're doing the same thing that people were ridiculed for if they didn't agree with any of your posted results. You'll go in and start picking at anything possible. Guess you don't like it when the tables are turned. But rest assured that I won't treat you the same way that people tht didn't agree with your links were treated.

Nevertheless the i5/i7 does not perform as well on non-game applications as the mythology propagated on various forums mistakenly claims.


EDIT: I know many people ignore 3dMark Vantage but the GPU score for the PhII is between 1500 and 3000 points HIGHER than the i5/i7 at the same clock speeds for Crossfired systems showing that the PhII at those speeds performs much better than the i5/i7 systems. But since Vantage is synthetic we have to ignore any results from it.

First: This is VERY important as it shows that something is off with the review.

Second: They fly in the face of many much more reliable sources, if these sources came out with similar results then I would have no problem with them.

Third: This is the most important, comparisons need controls. A good test will only change a single variable. When that isn't possible it is best to limit variability. Changing the operating system is a huge variable to change. That would be comparing a Ferrari and a Corvette on a dirt rode to see which is the absolute fastest, it changes everything.

Fourth: I don't care anymore Keith, I tried to be reasonable if you don't want to be reasonable then go to another forum because apparently we are all biased Intel shills.

 

[keithlm]

if these sources came out with similar results then I would have no problem with them.

Read that sentence to yourself about 5 times. Maybe 10 times. Then ask yourself what is wrong with what your wrote.

Let me rephrase what you said: You are saying that because the results do not agree with what you believe to be true you have a problem with them. But if they did completely agreed with what you believe you would have no problem with them.

You see nothing wrong with that? You don't find that biased in any way?

BTW: Your rant about it being Linux versus windows is not really important in any way since all of those tests were run under Linux. Your point would only have value if some tests were under Linux and some ran under Windows. Since that is not the case then your point is not valid.

From what you mentioned in your last post if someone doesn't agree with you and shows you a reason why they do not agree then they are being "unreasonable" if you decide that you don't agree with them.

 

[Raidur]

I thought he said your sources are shady and unreliable.

 

[JAYDEEJOHN]

Ive seen the Linux scores before, it shows the largest amount of corruption.
If youre running linux, this really suxors

 

[keithlm]

I thought he said your sources are shady and unreliable.

Apparently on this forum any results that do not agree with the popular consensus are shady and unreliable. Luckily that doesn't make the results wrong; it only makes them unacceptable to the people that want to put their heads in the sand and ignore the results because they don't like what they see.

NOTE: The Phoronix people are getting ready to release a new "LIVE" DVD that allows you to insert a DVD, boot a machine, and run the tests. (All without installing anything.) The LIVE DVD removes ambiguity by forcing the hardware to be compared on a "level" playing field without artificially handicapping any of the hardware. As such the testing will be repeatable.

When it is becomes that easy to run it is likely that some reviewers may start using it. The testers that used to run Linux tests only stopped because it was too much of a pain to install Linux and run the tests.

I wonder what excuse you guys will come up with when that starts to happen. I predict an attempt at distraction. i.e., "Linux is useless and nobody uses it... you should look at these Cinebench or Adobe Photoshop scores instead." (Because everybody runs Cinebench and Photoshop.)

EDIT: 2/25/2010 AM:: Phoronix has released their newest 2010.1 version of their PTS Desktop LIVE test DVD. I guess I spoke too soon last night.

You don't like what you saw in my Phoronix test results? Run the tests yourself and post the results.

It's easy:
1. Download and burn DVD.
2. Boot DVD and run tests. Probably overnight.
3. Upload results.
4. Optionally create a PDF file with comparisons like I did.

 

[Raidur]

You are the one ignoring results keith.

Most of us would love to see the cheaper phenom ii beat out i5/i7. It would mean a lot for the industry and for us the consumer especially.

But... i5 and i7 are superior in 'our' programs in 'our' environment.

How do you expect us to ignore the countless reviews and benchmarks that point to i5?

You need to realize we're all not intel fanboys. Hell as a company id prefer amd. We just know how to read results, and many many results.

As amw said phenom ii will have its wins, and linux appears to be one of them. Just because it is better at a couple things doesn't mean we are to ignore the rest and assume phenom ii is better. I think YOU are the biased one.

We (most of us) show love to both companies and you are the one picking sides.

We do not ignore results. Coming from a troll we tend to think twice.

Imagine if this was all flipped. You'd be laughing in out faces when we tried to convince you to ignore 80% of the information about the subject and to go off what our single benchmark says and the performance in a different operating system.

 

[JAYDEEJOHN]

But its also important NOT to ignore these findings as well.If youre playing at home in a sporting event, you usually have an advantage. I think some of this applies here, tho who knows how much and how severe.
Id just remind you, Intel has been accused of this, and have also been ruled against for doing some shady things. Is this another piece? Time will tell

 

[AMW1011]

Keith it doesn't matter how the processors compare in an obscure OS or without an OS because 98% of people will only benefit from the performance in windows.

Also, you totally read what you wanted into my statement. It is not that the results don't agree with my "opinions", but that they don't agree with the results of all the proffesional sites.

 

[keithlm]

You are the one ignoring results keith.

Most of us would love to see the cheaper phenom ii beat out i5/i7. It would mean a lot for the industry and for us the consumer especially.

But... i5 and i7 are superior in 'our' programs in 'our' environment.

How do you expect us to ignore the countless reviews and benchmarks that point to i5?


You need to realize we're all not intel fanboys. Hell as a company id prefer amd. We just know how to read results, and many many results.

Please expound on how my posting Linux benchmarks is in any way "ignoring results". I'm sorry but you won't be able to do that since those two actions have nothing to do with each other.

You seem to want to put your head in the sand and pretend that Linux results do not reveal anything and that Windows is the only important operating system. This is because the Linux results do not mirror your opinion. Do you understand that your opinion does not allow you to completely denigrate and ignore anything you do not like? When you do that you actually define the term "fanboy".

Intel has been accused of this

If Intel is forced to fix their libraries and play on a level playing field... I will be nourished by the tears of the whining fanboys.

it doesn't matter how the processors compare in an obscure OS

Linux is not "obscure" as you seem to think.

Regardless: I thought the goal was to actively investigate comparative performance between brands. Ignoring results you don't like is not a good tactic if you seriously have that goal. If I told you to ignore some of your results you would have an aneurism. But somehow you find it acceptable to do the same in reverse and don't consider that to be seriously fanboyish.