Is an i5 worth it for mid end cards?

[zappian]

Oh another thing.

You say this is common sense but i see prebuilts with an i7 and a r9 270.

I see people IRL pairing a gtx 650 and an i7 .

Just because its obvious for you doesn't mean everyone knows it , not everyone is a teçhie like you.

Some respect would be great.

People are dumb / misinformed  , its what benchmarks are there for ,to help them.

You might know that there's no difference between i3 and i5 in mid end gpus but some people dont know and want proof.

[Laputacake]

nothing to see here

[vSyNd1c4t3]

you can check my youtube channel i've tested this already with my GTX 780...the pentium is a stuttery mess at the settings the GTX 780 can handle fine, the core i3 does limit the performance in many games but of course not in all of them...and in some games even the i5-4460 will limit a GTX 780 (ACU for example).

https://www.youtube.com/user/YansaG/videos

i even have videos on BF4 with the old AMD FX

 

Awesome, I'll come check them out, help when I make recommendations.

[zappian]

The point of the thread is to help people buying new systems showing you can pair an i3 with a popular card like the r9 280 ( my 270x is actually about on par with that card with the overclock)

You don't need an i5.

Just because you dont need the info doesnt mean it shouldnt be there for who needs it.

The point is the thread is JUST proving that the i3 is just fine for mid end cards unlike what some people think

So much hate

[david cassar]

Great tests but the only thing that is flawed IMO is that the games/benchmarks are GPU dependant. You could put in an athlon or a pentium and you would probably get similar frames in those benchmarks. Are there any benchmarks of games like battlefield multiplayer which are known to depend on cpu performance?

[zappian]

Great tests but the only thing that is flawed IMO is that the games/benchmarks are GPU dependant. Are there any benchmarks of games like battlefield multiplayer which are known to depend on cpu performance?

 

I didnt benchmark bf 4 with the i3 I am afraid.

Metro 2033 is very cpu heavy .

I have dragon age inquistion benchmakrs coming

Also read this:

http://linustechtips.com/main/topic/345931-intel-core-i3-vs-i5-benchmarks/?p=4725423

[david cassar]

I didnt benchmark bf 4 with the i3 I am afraid.

Metro 2033 is very cpu heavy .

I have dragon age inquistion benchmakrs coming

Also read this:

http://linustechtips.com/main/topic/345931-intel-core-i3-vs-i5-benchmarks/?p=4725423

Yes I know I have read that haha all I'm saying is that you could have proven your point better if you varied the benchmarks a bit more.

[zappian]

Yes I know I have read that haha all I'm saying is that you could have proven your point better if you varied the benchmarks a bit more.

 

I have more benchmarks coming

[david cassar]

I have more benchmarks coming

Looking forward to them

[zappian]

Thread updated 

 

-Added Dragon Age Inquisition Benchmarks

 

http://linustechtips.com/main/topic/345931-intel-core-i3-vs-i5-benchmarks/?p=4711167

Results are within margin of error again.

[Oktyabr]

Would love to see Arma3 used as a benchmark.  I know the game on ultra crushes just about everything but it's the only game I'm considering a new build for so...

[ADZ_123_!"£]

Just a question in BF4 multi-player how bigger fps difference would there be between an I3 and an I5?

 

Thanks.

[Blackie Sheen]

I dont really get the point of this testing since pretty much everybody knows that games dont use more than 2 cores. Having a quad core is just better when you are gaming and still want to have intensive tasks in the background.

 

Ok, because many people will think that the quad core would be better

[ADZ_123_!"£]

I dont really get the point of this testing since pretty much everybody knows that games dont use more than 2 cores. Having a quad core is just better when you are gaming and still want to have intensive tasks in the background.

 

Ok, because many people will think that the quad core would be better

But then why is that FC4 wont work on dual cores. (Other than the fact that ubisosoft are idiots...)

[Blackie Sheen]

But then why is that FC4 wont work on dual cores. (Other than the fact that ubisosoft are idiots...)

Its one of the few games that do currently (I know I should of mentionned ''most games'' and not just ''games''. With the advent of directx12, there will be a lot more games that uses more than 2 cores but I dont think that saying that ''quad core is necessary for gaming'' is really a good representation of the current necessity for gaming.

 

And yes, Ubisoft are dorks because it is just stupid to be locked to using 4 cores for example. The game should allow for using less than 4 cores.

[D3MON]

I know you say you are testing at the setting you use, but benchmarking at 800x600 would have ensured a CPU-bound scenario, and would have shown you how well each CPU can potentially do without the GPU as a variable.

if you wanted cpu benchmarks use cinebench, this is a gaming test.

-edit, thanks zappian for the testing, i know how much of a pain and effort goes into it.

then idiots come along and tell you that you know nothing and ignore your work.

[othertomperson]

I mean it speaks volumes of the community that crappy memes get tons of likes but people trying to post possibly relevant information , taking time of their day to make a presentable thread gets flamed to high hell.

I had the same reaction when I posted the intel mantle thread.

It kinda makes weary of posting any further benchmarks I do.

 

If you're going to take critical analysis of your testing methodology as "flaming" then that's pretty immature. If you are going to post benchmarks then people should comment on whether or not conclusions can actually been drawn from it.

 

 

if you wanted cpu benchmarks use cinebench, this is a gaming test.

 

This doesn't test anything, that's the problem. There is literally no information to be discerned from this.

 

Like literally, if not testing the CPU what, exactly is a "i3 vs i5" test even for?

[D3MON]

Its one of the few games that do currently (I know I should of mentionned ''most games'' and not just ''games''. With the advent of directx12, there will be a lot more games that uses more than 2 cores but I dont think that saying that ''quad core is necessary for gaming'' is really a good representation of the current necessity for gaming.

 

And yes, Ubisoft are dorks because it is just stupid to be locked to using 4 cores for example. The game should allow for using less than 4 cores.

actually i tested bf3 and itll run okay on a dual core with both pinned at 99% but minimums suck, its okay as a tri core however with high cpu usage, quad core is great and anything higher just allows more cpu overhead and lower cpu usage (so lower temps), bf4 and other frostbite games are the same, in fact most multi threaded AAA titles are (crysis 3 for instance) thats not so say all multi threaded games need it as you said, swtor uses multiple threads really well (stupidly well considering its an online rpg...) but you could run that on a sempron 

[D3MON]

 

If you're going to take critical analysis of your testing methodology as "flaming" then that's pretty immature. If you are going to post benchmarks then people should comment on whether or not conclusions can actually been drawn from it.

 

 

 

This doesn't test anything, that's the problem. There is literally no information to be discerned from this.

 

Like literally, if not testing the CPU what, exactly is a "i3 vs i5" test even for?

 

you didnt get anything from this!?

how about learning which games benefit from faster cores, how about which games are legitimately demanding and require more processing power and which ones dont?..

okay all of them performed pretty much the same ergo the gpu was the limiting factor except metro which has a worse minimum, that said the minimums on both were so low its likely to be a game problem and not a cpu problem, eitehr way hes done allot of work to give us reference points for the future, what have you done for us lately?

[FuzzyYellow]

No, but you are also benchmarking the CPU, not the entire system. The games you are benchmarking are quite graphically intensive. It's not outside the realm of possibility to get 50 fps on both CPUs but only because your 270X is bottlenecking you. Your results should apply to someone with a 250 as much as they apply to someone with Titan Xs in SLI, and you do that by removing the GPU from the equation.

 

 

 

Testing CPU performance with a mid-low end GPU isn't. To compare CPUs you either go way overboard on the GPU (Titan X, since you don't want SLI/CF to affect the result either), or you find other ways of removing it as a variable.

but that's not the point of this test. the point of this test was to see if 1 cpu held back the gpu more than the other, which in this case it doesn't. if you want to find the difference between the 2 with very little GPU bottleneck, watch the newest LTT video.

[othertomperson]

you didnt get anything from this!?

how about learning which games benefit from faster cores, how about which games are legitimately demanding and require more processing power and which ones dont?..

 

The CPUs are of the same generation, the speed of the cores is pretty much the same. It could have potentially been an interesting test of how much of a difference real cores vs logical cores made and to what extent hyperthreading is an effective tradeoff (would have been especially interesting if a Pentium were thrown in as well), but the heavy GPU bottleneck meant that actually you couldn't find that much out. You might be tempted to say that logical cores and real cores don't make much of a difference because the i3 kept up mostly with the i5, but that's because the testing methodology resulted in a strong bias against the i5 due to the aforementioned bottleneck.

 

 

but that's not the point of this test. the point of this test was to see if 1 cpu held back the gpu more than the other, which in this case it doesn't. if you want to find the difference between the 2 with very little GPU bottleneck, watch the newest LTT video.

 

But it does, and it's only the flawed testing methodology that leads to your mistaken conclusion.

 

Removing the GPU bottleneck would give you a raw FPS that each CPU was capable of, and you could have used that information to decide which GPU to pair with each -- you'd want the limit of the GPU and CPU to be similar. That would be a useful application of some results, but you don't get that information here.

[zappian]

if you wanted cpu benchmarks use cinebench, this is a gaming test.

-edit, thanks zappian for the testing, i know how much of a pain and effort goes into it.

then idiots come along and tell you that you know nothing and ignore your work.

 

Thanks for the kind words.

I mean I did the test but I already sold the i3 cpu .

So its as you get what you get type of situation.

I did work to make a decent thread for the community , if people already knew the outcome thats fine.

I was actually surprised.

Now calling it "useless" is pretty mean lol.

I know people that enjoyed the thread and liked the results and effort put into it .

If some people dont like it I just hope they are polite about it.

[zappian]

Just a question in BF4 multi-player how bigger fps difference would there be between an I3 and an I5?

 

Thanks.

 

The difference isnt that significant if you use mantle in both cpus.

The problem with bf4 is with the i3 it cripples your system meaning that if you have bf4 and another application open the computer will start to slow down because the i3 just cant handle it.

The game where I got the best results was far cry 4 where i got about a 10 fps increase , I didn't  do benchmarks because far cry 4 doesn't have a benchmark utility but the difference was significant.

[Mephy04]

I like this type of testing! Thank you, my wife and I both PC game so building two mid range machines is the name of the game most of the time.

[ADZ_123_!"£]

The difference isnt that significant if you use mantle in both cpus.

The problem with bf4 is with the i3 it cripples your system meaning that if you have bf4 and another application open the computer will start to slow down because the i3 just cant handle it.

The game where I got the best results was far cry 4 where i got about a 10 fps increase , I didn't  do benchmarks because far cry 4 doesn't have a benchmark utility but the difference was significant.

So would even having just BF4 and skype open would ruin in game performance?