Intel Core i5 4690K "Devil's Canyon" - Review - Overclocking & CPU Benchmarks

[Faa]

I'll see what I can manage, downloading now will be done in 40 mins :rolleyes:

 

Not sure how much faster a 780 is than a 7950. At the overclock I run I get 2000 in valley benchmark (max settings) if that's of any help.

Well can't manually downclock it any further than 105MHz, so have to notch the temp target down to make it downclock any further but thats complety messy. I got around 2170 in valley

[harrynowl]

Well can't manually downclock it any further than 105MHz, so have to notch the temp target down to make it downclock any further but thats complety messy. I got around 2170 in valley

It'll do, just knock 5% from your star swarm score I guess as I got 2062 from memory in valley.

[Faa]

It'll do, just knock 5% from your star swarm score I guess as I got 2062 from memory in valley.

Here you go: http://i.imgur.com/HaASjkB.png

Extreme preset with cpu at stock

[harrynowl]

Here you go: http://i.imgur.com/HaASjkB.png

Extreme preset with cpu at stock

I'll add my results in the OP, CPU is at 4.7.

[Faa]

 

My result. I think GK110 and Tahiti aren't really worth comparing if the main goal of the comparison is to benchmark SB-E against Haswell

 

I'll try with Mantle and see if it improves the score by a notable amount.

 

Also extreme with "follow". 

 

CPU @ 4.7

Yeah agreed The scores didnt make any sense lol.

Nvidia released lately that 337.50 driver, reduced cpu bottlenecking by a huge difference. R9 290x used to be dunno 30-40% faster than 780ti's with mantle, now its the 780ti but not sure by how much. 

[harrynowl]

Yeah agreed The scores didnt make any sense lol.

Nvidia released lately that 337.50 driver, reduced cpu bottlenecking by a huge difference. R9 290x used to be dunno 30-40% faster than 780ti's with mantle, now its the 780ti but not sure by how much. 

I don't think I'll be getting my hands on a 780 anytime soon unfortunately

 

AMDs Catalyst drivers have been doing quite well though in my opinion. Going from 12.something a while ago to 14.6 has given the 7950 so much better performance for me.

[Faa]

I don't think I'll be getting my hands on a 780 anytime soon unfortunately

 

AMDs Catalyst drivers have been doing quite well though in my opinion. Going from 12.something a while ago to 14.6 has given the 7950 so much better performance for me.

Star Swarm is more of a cpu comparison test, although if you compare an AMD with nvidia there youre rather testing the driver overhead instead of how powerful the gpu's are if you do dx11 vs dx11. 

Here are a few: 

http://pclab.pl/art58123-3.html (Skyrim shows such a huge difference?)

http://cdn.wccftech.com/wp-content/uploads/2014/02/battlefield-4-mantle-benchmarks.gif

http://www.pcgameshardware.de/screenshots/original/2014/05/Watch_Dogs-CPU-Driver-Overhead-Benchmarks-pcgh.png

Wasnt trying to say that AMD drivers are bad honestly theyre even better than nvidia, just trying to explain why the difference was so huge because normally a 780 performs the same as a 290 when being cpu limited. This is probably why you see sometimes a 30% difference orsomething between a 780 ti & R9 290x (at 99% vs 99% its at best a 5% difference). This theory kinda falls in water, many 780ti/R9 290x owners are gpu bound in BF4 so the difference is mostly zero or just the 5% they all fight about. For mmo's? The game itself is CPU bound, drivers can't do much besides reducing the overhead but Mantle can do this.

[harrynowl]

Star Swarm is more of a cpu comparison test, although if you compare an AMD with nvidia there youre rather testing the driver overhead instead of how powerful the gpu's are if you do dx11 vs dx11. 

Here are a few: 

http://pclab.pl/art58123-3.html (Skyrim shows such a huge difference?)

http://cdn.wccftech.com/wp-content/uploads/2014/02/battlefield-4-mantle-benchmarks.gif

http://www.pcgameshardware.de/screenshots/original/2014/05/Watch_Dogs-CPU-Driver-Overhead-Benchmarks-pcgh.png

Wasnt trying to say that AMD drivers are bad honestly theyre even better than nvidia, just trying to explain why the difference was so huge because normally a 780 performs the same as a 290 when being cpu limited. This is probably why you see sometimes a 30% difference orsomething between a 780 ti & R9 290x (at 99% vs 99% its at best a 5% difference). This theory kinda falls in water, many 780ti/R9 290x owners are gpu bound in BF4 so the difference is mostly zero or just the 5% they all fight about. For mmo's? The game itself is CPU bound, drivers can't do much besides reducing the overhead but Mantle can do this.

Well I have a few CPU bound games I'd like to try.

 

I have WoW but can't really think of a place to do a synthetic loop (is there a quest or something that's particularly heavy?)

 

I'd like to do Skyrim but this will probably happen tomorrow

 

Total war is insanely CPU bound but it's very hard for me to test.

[Faa]

Well I have a few CPU bound games I'd like to try.

 

I have WoW but can't really think of a place to do a synthetic loop (is there a quest or something that's particularly heavy?)

 

I'd like to do Skyrim but this will probably happen tomorrow

 

Total war is insanely CPU bound but it's very hard for me to test.

Do you have BF3? No need to do a loop Standing still here: http://i.imgur.com/9vI7cYn.jpg

You could notice how hard I was cpu limited, at 4.8GHz I still cant get the gpu above 80% load there. Its even worse in a full server

[harrynowl]

Do you have BF3? No need to do a loop Standing still here: http://i.imgur.com/9vI7cYn.jpg

You could notice how hard I was cpu limited, at 4.8GHz I still cant get the gpu above 80% load there. Its even worse in a full server

I do have BF3 as I picked it up in that humble bundle.

 

I'll try both BF3 and 4 on various player size servers and see how it does

[Faa]

I do have BF3 as I picked it up in that humble bundle.

 

I'll try both BF3 and 4 on various player size servers and see how it does

Could you try to take the same position there as I did? Its strike at karkand, spawn at G gatehouse, stand like I was (90pov & ultra) and take screenshot. Would love to see the gpu load though and a 4670k with a 7950 outperforming a 3930K & 780 lol.

[harrynowl]

Could you try to take the same position there as I did? Its strike at karkand, spawn at G gatehouse, stand like I was (90pov & ultra) and take screenshot. Would love to see the gpu load though and a 4670k with a 7950 outperforming a 3930K & 780 lol.

Is that a DLC map?

 

I don't have DLC for anything really

[Faa]

Is that a DLC map?

 

I don't have DLC for anything really

Yes

[harrynowl]

Yes

I'll have to try another then, what is a demanding vanilla map do you think? (64p)

[Faa]

I'll have to try another then, what is a demanding vanilla map do you think? (64p)

Hmm it's only strike at karkand thats so extremely cpu bound, in most maps I'm much closer to hit the limit on the gpu though.

Anyways I've found one: http://i.imgur.com/nKHxkm0.jpg 

Load is actually 64%, had to record the game and open it with vlc to take a screenshot lol. It's noshar canals, just empty server 64 TDM.

[harrynowl]

Hmm it's only strike at karkand thats so extremely cpu bound, in most maps I'm much closer to hit the limit on the gpu though.

Anyways I've found one: http://i.imgur.com/nKHxkm0.jpg 

Load is actually 64%, had to record the game and open it with vlc to take a screenshot lol. It's noshar canals, just empty server 64 TDM.

I'll do that tomorrow then.

 

Hopefully Haswell master race continues

[AEA]

Probably be picking up a 4690k so this is very interesting to me

[harrynowl]

Probably be picking up a 4690k so this is very interesting to me

Hopefully you get a good chip like mine

[AEA]

Hopefully you get a good chip like mine

 

Hopefully