AMD Phenom II X4 965 vs. Intel core i5-750

[dantoshi]

Which of these processors is the best one?

 

[andy5174]

Just an FYI for future reference, Anand's Far Cry 2 benchmark isn't gameplay. It's an artificial stress test using the FC2 engine that simulates how the game would play if each core was at 100%. Using the in-game benchmarking for actual gameplay, both an i5 and a 955 will net 90+ FPS with a GTX280.

Regardless, a 955, an i5, or an i7 will get you the same performance in gaming. As a matter of fact, P2's and 1156 CPus will be better in gaming than the 920 due to tighter framerates. Outside of games, an i5 will net you ~5-10% additional performance in non-gaming applications over the Phenom II, with the exception of a select few applications.

For games, the 955 and the i5 are excellent and will remain excellent for years. You would never notice a difference.

No, you are terribly wrong!

There is huge difference between i5 and 955.

The 955 is much better and will last for 10 years compared to merely 1 year on i5.

 

[andy5174]

Medium quality gaming benchmarks fail horribly, especially when the totals are in the 100's. In short, those anand gaming benchmarks are a joke.

On the other hand, if anybody cares to work out the lostcircuits benches?

Ah hell let me tell you how it ends up. The Phenom II 965 BE is 8% faster than the i5 750...that is with me being as generous as possible to the i5 by including the SSE4 V-Dub benchmarks which scores a 15% swing in the i5's advantage. Not to mention cinebench, yet another synthetic that favours intel.

Also, if I had worked out the average fps totals of the end gaming benchmarks and added them to the phenom II score, the 965 BE would probably have ended up 15% faster.

So yes I was wrong, it's not 5% its more like 15% in favour of the phenom II using real world apps.

Are you kidding?! It IS 45% in favour of the phenom II using real world apps.

 

[Atranox]

Keep in mind that the i5 platform has received numerous revisions since the time of benchmarking. At the time of most benchmarks, the AM3 platform was very solid, while the 1156 was using it's very first shipments.

For non-gaming applications, judging by most benchmarks out there (including lostcircuits), an i5 is about 5-10% better in non-gaming applications, which will shave off a few seconds to a few minutes, depending upon task size.

For gaming at normal resolutions, an i5 or a P2 will be exactly the same. Some games favor Intel architecture, and some favor the AM3 platform.

 

[andy5174]

And right about now someguy7 is going to confirm those results. Right someguy?

No need!

They are wrong as jennyh said they are wrong.

 

[jennyh]

No its about 10-15% in reality. That's the real reality not your warped reality andy.

 

[andy5174]

Keep in mind that the i5 platform has received numerous revisions since the time of benchmarking. At the time of most benchmarks, the AM3 platform was very solid, while the 1156 was using it's very first shipments.

For non-gaming applications, judging by most benchmarks out there (including lostcircuits), an i5 is about 5-10% better in non-gaming applications, which will shave off a few seconds to a few minutes, depending upon task size.

For gaming at normal resolutions, an i5 or a P2 will be exactly the same. Some games favor Intel architecture, and some favor the AM3 platform.

Rubbish!

i5 is about 45% WORSE in non-gaming applications.

 

[jennyh]

Keep in mind that the i5 platform has received numerous revisions since the time of benchmarking. At the time of most benchmarks, the AM3 platform was very solid, while the 1156 was using it's very first shipments.

For non-gaming applications, judging by most benchmarks out there (including lostcircuits), an i5 is about 5-10% better in non-gaming applications, which will shave off a few seconds to a few minutes, depending upon task size.

For gaming at normal resolutions, an i5 or a P2 will be exactly the same. Some games favor Intel architecture, and some favor the AM3 platform.

I just worked out the lostcircuits benchmarks and the 965 BE is 8% faster at the very least. If the synthetic cinebench and SSE4 v-dub benchmarks were removed, the 965 BE would be more like 15% faster overall.

 

[AMW1011]

Ah and just when we had Jenny speaking the truth...

The i5 750 and the Phenom II 965 trade blows in every benchmark, but the i5 750 comes out slightly ahead if you factor in that the i5 750 is faster clock for clock. At the end of the day both the Phenom II 965 and the i5 750 are totally equal with some imperceptible gains in the i5 750's favor.

 

[jennyh]

Truespace 5.1

i5 = 55
p2 = 63

i5 is 13% faster

Truespace 7.5

i5 = 157
p2 = 144

p2 is 9% faster

Cinebench 2003

i5 = 1935
p2 = 1708

i5 is 12% faster

Cinebench r10

i5 = 11320
p2 = 11096

i5 is 2% faster

MainConcept H.264

i5 = 50.2
p2 = 43.68

p2 is 13% faster

Dvdshrink 3.2

i5 = 167
p2 = 159

p2 is 5% faster

Nero 9 recode

i5 = 422
p2 = 402

p2 is 5% faster

V-dub non sse4

i5 = 16
p2 = 15

p2 is 6% faster

v-dub sse4

i5 = 17
p2 = 20

i5 is 15% faster

DIEP chess

i5 = 863563.4
p2 = 948604.1

p2 is 9% faster

Crysis

i5 = 170.5
p2 = 177.8

p2 is 4% faster

UT3

i5 = 147.0
p2 = 139.0

i5 is 5% faster


Overall

i5 wins =

13
12
2
15
5

= 47

p2 wins =

9
13
5
5
6
9
4

= 51

Phenom II 965 BE is 8% faster than the i5 750 on those alone.

That is being as generous to the i5 as I possibly can be. Btw I stopped before the end gaming benchmarks which (taken on average) would have scored the phenom II 965 BE even higher.

By being generous to the i5 I mean I included the cinebench benchmarks (synthetic favouring the intel), and the SSE4 V-Dub benchmark which frankly is a joke but there you go.

Argue with those numbers, anybody.

 

[randomizer]

Just stop arguing, now. You know it's going to end up nowhere.

 

[andy5174]

@jennyh

No need to argue. You rule the world!

And no PC benchmark needed as the results given by your imagination/make up is the best.

 

[Atranox]

You can't base your conclusions off of one review. While I'm sure lostcircuits is very accurate, there are numerous variables that go in to benchmarking. You can have 3 different people run the same benchmarks with identical setups, and I can almost assure you that each would get different results. For every site that has the 955/965 winning, there are two that has the i5 winning. You can safely assume from that, that the i5 is slightly faster than a Phenom II for non-gaming.

 

[andy5174]

You can't base your conclusions off of one review. While I'm sure lostcircuits is very accurate, there are numerous variables that go in to benchmarking. You can have 3 different people run the same benchmarks with identical setups, and I can almost assure you that each would get different results. For every site that has the 955/965 winning, there are two that has the i5 winning. You can safely assume from that, that the i5 is slightly faster than a Phenom II for non-gaming.

NO!

jennyh > anything

 

[jennyh]

Atranox the point which seems to be lost on a great many people is, the i5 is winning the *synthetic* benchmarks not real world apps.

If Cinebench was removed the Phenom II would be a country mile ahead. When looking at benchmarks you have to look at what is being benched and when it's an actual program the Phenom II wins more than it loses.

Look at my numbers and you'll see a huge 'win' in cinebench for the i5. It's not a real program it's a bloody benchmark! Then check out the SSE4 V-Dub result and read what they say about that on lostcircuits.

If it wasn't for those the phenom II would be an absolute mile ahead. It is a better cpu and the benchmarks prove it if you know what you are looking at.

 

[jennyh]

Andy what exactly are you contributing to this thread except utter garbage?

 

[andy5174]

I made immense contribution as I agree with you.

 

[jennyh]

You know my phenom II 940 at 3.8ghz is faster than your i5 at 3.6ghz right?

 

[ElMoIsEviL]

Actually I took my numbers from the owner of Lost Circuits, who incidentally has benchmarked both systems a number of times.

Allright...

How about you get off your lazy backside and look for them yourself in the forum section.

Naw you're the one making the positive claim. You ought to substantiate your claim with evidence.

Didn't you know that if you don't post a link that contradicts the majority belief on this forum then you are automatically wrong? Even if you later post it then it doesn't matter, you'll still be wrong. (Even if you are quoting a review that is actually on this site.)

A single link?
The "beliefs", if you can call them that, on this forum are based on a rational and reasonable interpretation of the evidence. If you have 10 Objective Websites telling you one thing and a single 1 telling you something else... who would you be more prone to believe?

No. This user has proven that she is terrible at math or just makes up numbers. And thats whats goes on over at the zone. Except they ban you.

I agree with your statement.

Yep I'm aware of that...however my sheer determination will eventually break them down and they'll come to see reason.

I'm gonna start with someguy7, who surely will want to get to the truth?

Start here someguy - http://www.lostcircuits.com/mambo//index.php?option=com_content&task=view&id=63&Itemid=42&limit=1&limitstart=12

Take every one of those benchmarks and tell us how much faster or slower the 965 BE is compared to the i5 750. For a maths wizard, that should be easy.

I have noticed that since I have arrived here at Toms Hardware, that you have begun to use my wording. "Reasoning and Rationality" were never part of your vocabulary prior to my return. An interesting observation. Here is a test to see if you truly are "reasonable".

Your data: http://www.lostcircuits.com/mambo//index.php?option=com_content&task=view&id=63&Itemid=42

Vs. Contradicting data:
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3634
http://www.tomshardware.com/reviews/intel-core-i5,2410.html
http://www.hardocp.com/article/2009/09/07/intel_lynnfield_core_i5_i7_processors/
http://www.tweaktown.com/reviews/2909/intel_lynnfield_core_i5_750_and_core_i7_870_performance_testing/index.html
http://www.xbitlabs.com/articles/cpu/display/core-i7-870.html
And MANY MANY more.

So reason with me why it is that I should ignore all of the other sites and instead rely on the data from Lost Circuit?
What have all these other sites done which somehow invalidates their data?

That my dear is called Reasoning.

 

[andy5174]

Allright...


Naw you're the one making the positive claim. You ought to substantiate your claim with evidence.


A single link?
The "beliefs", if you can call them that, on this forum are based on a rational and reasonable interpretation of the evidence. If you have 10 Objective Websites telling you one thing and a single 1 telling you something else... who would you be more prone to believe?


I agree with your statement.


I have noticed that since I have arrived here at Toms Hardware, that you have begun to use my wording. "Reasoning and Rationality" were never part of your vocabulary prior to my return. An interesting observation. Here is a test to see if you truly are "reasonable".

Your data: http://www.lostcircuits.com/mambo//index.php?option=com_content&task=view&id=63&Itemid=42

Vs. Contradicting data:
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3634
http://www.tomshardware.com/reviews/intel-core-i5,2410.html
http://www.hardocp.com/article/2009/09/07/intel_lynnfield_core_i5_i7_processors/
http://www.tweaktown.com/reviews/2909/intel_lynnfield_core_i5_750_and_core_i7_870_performance_testing/index.html
http://www.xbitlabs.com/articles/cpu/display/core-i7-870.html
And MANY MANY more.

So reason with me why it is that I should ignore all of the other sites and instead rely on the data from Lost Circuit?
What have all these other sites done which somehow invalidates their data?

That my dear is called Reasoning.

jennyh > And MANY MANY more

 

[BadTrip]

Just stop arguing, now. You know it's going to end up nowhere.

Thank God I am not the only one sick of all this nonsense.

@OP The numbers are here. Study them and make up your own mind.

 

[andy5174]

There is not a single benchmark that is reliable except the famous "jennyh" benchmark.

 

[andy5174]

Thank God I am not the only one sick of all this nonsense.

@OP The numbers are here. Study them and make up your own mind.

I didn't want to join the debate too.

However, I can't tolerate seeing other peoples saying the so right jennyh is wrong. Hence, I joined it once again.

 

[notty22]

http://www.lostcircuits.com/
Hanging on one no name web site = fail
His testing methods : For comparison purposes, all benchmarks were run in WinXP 32
Mentions: Intel Turbo is cheating= final fail
Compare to Guru3d, Anand, THG , Tweaktown, PC World FGS !

What the i5 750 did to AMD pricing in 2 months.

 

[jennyh]

Your data: http://www.lostcircuits.com/mambo//index.php?option=com_content&task=view&id=63&Itemid=42

Vs. Contradicting data:
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3634
http://www.tomshardware.com/reviews/intel-core-i5,2410.html
http://www.hardocp.com/article/2009/09/07/intel_lynnfield_core_i5_i7_processors/
http://www.tweaktown.com/reviews/2909/intel_lynnfield_core_i5_750_and_core_i7_870_performance_testing/index.html
http://www.xbitlabs.com/articles/cpu/display/core-i7-870.html
And MANY MANY more.

So reason with me why it is that I should ignore all of the other sites and instead rely on the data from Lost Circuit?
What have all these other sites done which somehow invalidates their data?

That my dear is called Reasoning.

The xbitlabs article doesn't have an i5 750. The tweaktown article doesn't have a 965 BE. The hardocp article doesn't have a 965 BE.

 

[jennyh]

And please, everybody knows THG and Anand are intel shills.