Contradicting reviews

[mypettycash]

Can anyone teach me how to interpret these 2 reviews using a same game for testing? Both shows totally 2 different results. Both are using different setup though but not to that extent isnt it? I am not comparing between 4670k and 3570k. I'm refering to, not all but, some CPU.

Intel Core i5-4670K Gaming Performance | bit-tech.net

Intel Core i5-3570K Gaming Performance | bit-tech.net

 

[GrimaH]

Optimization updates, maybe? There's more than a year between the two reviews, and Creative Assembly is pretty active with their game updates.

 

[eminus]

game version and although shogun 2 uses cpu test it could also attribute to GPU used in that testing and especailly is drivers.

so any of the two will do, just get what you really need..that will fall mostly to the motherboard and its features.

 

[betadyne]

They are not using identical hardware and drivers.

 

[mypettycash]

so for example if i am comparing i7-2600k and i5-3570k, which review should i choose? the latest review?

it's like 1 year ago i believe that i5-3570k is the best. So 1 year later, which is 2013, i7-2600k which is a dated processer is better than i5-3570k which is a more up-to-date processor? So you mean that the drivers, version or all those updates for any hardware or drivers are improving the performance for a dated processor and decreasing the performance for a later processor? It doesn't make sense.

1 year ago - i7-2600k fps is 33fps while i5-3570k is 43fps. (Average fps)
1 year later - i7-2600k fps is 40fps while i5-3570k is 37fps. (Average fps)

 

[eminus]

if we are talking about CPU test then it could be that developers have managed to optimized their games to use hyperthreading.

but Sandy/Ivy/Haswell if going for stock or minor OC they are just the same in terms of gaming it is the GPU that takes care most of the work..

 

[Cryogenist]

I will think that those reviews that focus on synthetic benchmarks instead of games for CPU performance will be a better reference.

Most games have too many factors to see the difference in performance for just CPU alone.

 

[mypettycash]

Is it right to say that reviews like this are useless in regards to gaming? OCing CPU to a higher or max speed doesn't really help much on gaming isn't it?

 

[Cryogenist]

Is it right to say that reviews like this are useless in regards to gaming? OCing CPU to a higher or max speed doesn't really help much on gaming isn't it?

Somewhat useless with respect to gaming..
Only a few games are very CPU-intensive.

OC-ing normally helps more in CPU-intensive workload.

EDIT:
For me, the reason I OC is for my workload, video encoding, Lightroom editing.
Gaming generally get boosted more by the GPU.

 

[mypettycash]

so if that is the case, i just conclude that just buy the highest clock speed possible within budget. So dun need to OC. Next time if feel slow then OC.

 

[Cryogenist]

so if that is the case, i just conclude that just buy the highest clock speed possible within budget. So dun need to OC. Next time if feel slow then OC.

You talking about which generation?

Haswell cannot OC without K.
IVB have limited unlock for non-K

 

[GrimaH]

so for example if i am comparing i7-2600k and i5-3570k, which review should i choose? the latest review?

it's like 1 year ago i believe that i5-3570k is the best. So 1 year later, which is 2013, i7-2600k which is a dated processer is better than i5-3570k which is a more up-to-date processor? So you mean that the drivers, version or all those updates for any hardware or drivers are improving the performance for a dated processor and decreasing the performance for a later processor? It doesn't make sense.

1 year ago - i7-2600k fps is 33fps while i5-3570k is 43fps. (Average fps)
1 year later - i7-2600k fps is 40fps while i5-3570k is 37fps. (Average fps)

They could have updated threading optimization for Shogun 2. i7 2600K has four cores and eight threads compared to the i5 3570K's 4/4.

If anything it's telling you that the number of cores/threads is going to become increasingly relevant in future gaming.

 

[sHrike]

jus for sharing...as I jus bought a GTX760 but still using a E8400 Core 2 Duo...I tried playing Crysis 3 all on high settings on a 1680X1050 monitor I was surprised I can get 20-30fps . All stock clock ..no over clocking...now I am really doubting should I upgrade my cpu or not.

 

[Maeda_Toshiie]

One problem with the games is how multithreaded the games are.

TW games, up to Shogun 2 (and upcoming Rome 2), are not heavily threaded. The core carrying the main thread is the serious bottleneck. Worse is that the CPU load is directly affected by unit size and unit detail settings. And everyone screams for ultra... So when you zoom into the fights, you get very large fps drops.

So the CPU choice depends very much on the kind of games being played. SC2 is another well known game for the main thread bottlenecking performance.

 

[Dr.Vijay]

Upgrade when you feel it's slow

You'll still get a decent boost, but it won't be day and night difference.

But the C2D platform is finally starting to show signs of age, though it's generally still quite usable.

jus for sharing...as I jus bought a GTX760 but still using a E8400 Core 2 Duo...I tried playing Crysis 3 all on high settings on a 1680X1050 monitor I was surprised I can get 20-30fps . All stock clock ..no over clocking...now I am really doubting should I upgrade my cpu or not.

 

[Dr.Vijay]

More like when you feel it's slow, just re-install Windows and your s/w from scratch - you'll get a more responsive system then an OC -_-

Otherwise, just dump the old system and get a new one when it's time or just upgrade the GFX.

so if that is the case, i just conclude that just buy the highest clock speed possible within budget. So dun need to OC. Next time if feel slow then OC.

 

[m3 th3 last]

Hi, just to share some of my personal experience i had over the weekend.

Upgrade my CPU from Q6600 to i5-4670k over the weekend. Changed the Ram also.

Game performance in World of Tank increased from avg of 30~40 fps to current 60~70 fps.

Never changed graphic card, still using MSI 285GTX.

What i found out is that since the game is single threaded, the CPU is bottle-necking the GPU on Q6600.

So that maybe the answer you are looking for.

 

[sHrike]

Upgrade when you feel it's slow

You'll still get a decent boost, but it won't be day and night difference.

But the C2D platform is finally starting to show signs of age, though it's generally still quite usable.

Yah i agree...an excellent workhorse...just that bought in 2008...almost 5 years liao...so backside itchy wanna change cpu....hahah

 

[mypettycash]

Upgrade my CPU from Q6600 to i5-4670k over the weekend. Changed the Ram also.

Game performance in World of Tank increased from avg of 30~40 fps to current 60~70 fps.

C2D to 4th gen processor of coz got a big leap. Not surprising the performance has increased.

What worried about is maybe someone like me will look at the one of the reviews, not both, at my first post and say "wow this CPU is better than that". And will buy the best CPU. But is there really a big difference for gaming if you buy i5-**** or i7-****?

For those non-hardcore gamer who feels the itch to buy new CPU becoz is faster, dun be silly like me and look at those useless reviews that will tell you nothing. 2nd gen cpu can be as good as 4th gen cpu in terms of performance. It's just that more up-to-date cpu has more features. Just like what is found out between H series and Z series Mobo. Just wait until you feel slow on your computer and then OC and the computer will tell you "please buy a new CPU, i cannot handle the stress if not i will explode".

 

[jarred1223]

if you are using the PC for games only, focus on your gfx card instead of your processor.