xfire msi 290 oc with 8350

Rasch · 9 Feb 2014 at 16:51

The main problem is the weak IPC you gotta have really bad min fps in most of the games that not using lots of cores,hell even in Crysis 3 i got some really bad dips.Check out Cinebench R15 bench thread.I get like 120pts in single thread.Some guys with oced 4770K gets nearly 200.That tells everything.

Rasch · 9 Feb 2014 at 16:53

This tells everything really.

LtMatt · 9 Feb 2014 at 16:54

Tonester0011 said:
The link is, but his scores he wrote are normal FS.
I agree about not using FS to tell whether you're bottlenecking. Ingame benchmarks will give a truer picture as to how much.

Martini1991 said:
Just it says Firestrike Extreme on it.

Maybe that's why I'm getting confused.

Either way, I really can't see there being much if any difference with Firestrike.

EDIT : Agh, Tone explained

Crap. What's happened there. I've lost the link. I'll see if i can dig it up, BRB.

Martini1991 · 9 Feb 2014 at 16:55

Rasch said:
This tells everything really.

And that's with a GTX680, an R9 290 Crossfire is well over twice of that performance

OvEr_KiLL · 9 Feb 2014 at 17:05

guys im not bothered about benching my cards i just want better performance in games but if it isnt going to be that much then i wont bother

Martini1991 · 9 Feb 2014 at 17:08

You'll see gains, there's no doubts about that, but you'll be getting far from the full performance out of them.

Whether or not you think that's worth it is your call, but it'll bottleneck, and that is fact.

But with you buying a new motherboard, I'd strongly consider pushing more for the Intel set up.

LtMatt · 9 Feb 2014 at 17:11

Martini1991 said:
You'll see gains, there's no doubts about that, but you'll be getting far from the full performance out of them.

Whether or not you think that's worth it is your call, but it'll bottleneck, and that is fact.

But with you buying a new motherboard, I'd strongly consider pushing more for the Intel set up.

+1.

Kaapstad · 9 Feb 2014 at 17:23

Here is mine with the 290Xs @stock, anyone with a 8350 and a couple of 290Xs want to do a run and compare the graphics scores.

Score 16724

AMD Radeon R9 290X(2x) @1000/1250

Intel Core i7-4930K @4.8

Graphics Score 20642

Physics Score 18129

Combined Score 6586

http://www.3dmark.com/fs/1689650

The latest Haswell 4770k CPUs are more efficient than mine and will allow the cards to score slightly more than my 4930k

Choimaloy75 · 9 Feb 2014 at 17:39

This is what I've being trying for people to understand, some people aren't bothered about all this benching to prove what is better the op wants a bit more performance for games and getting another gpu regardless of weather he has amd cpu or not if gets the performance then great, from my point of view I changed from a 7990 to 2 290x's and I've seen good increase in performance I don't think the op wants to go out and spend more on a cpu and motherboard.

Martini1991 · 9 Feb 2014 at 17:40

Choimaloy75 said:
This is what I've being trying for people to understand, some people aren't bothered about all this benching to prove what is better the op wants a bit more performance for games and getting another gpu regardless of weather he has amd cpu or not if gets the performance then great, from my point of view I changed from a 7990 to 2 290x's and I've seen good increase in performance I don't think the op wants to go out and spend more on a cpu and motherboard.

He's buying a new motherboard.
OP asked if it bottlenecks, it does.

If you want to be ignorant, that's fine, but don't try and rubbish other people and calling them silly names like Intel preachers.

Avenged7Fold · 9 Feb 2014 at 17:41

Very specific example here but i used Valley, which is as game like as benches get really. I think valley is threaded for four cores, which i suppose is a fair place to start thinking about CPU bottlenecks in games.

When i moved my GPU to my other rig, i benched it at different common clocks to compare.

[email protected] with a 290x OCed @1241/1500 Was scoring 25xx on Valley with the leaderboard's settings.

Same card with same clock was scoring 30xx in Valley when i put the card in my 4770k @ 4.4

Now in Valley my GPU usage is at around 99% when it isnt changing scenes, while it was no more than in the low 80's for my 8320.

I would look towards getting an i5, especially if you are going to dish out for a ud5 board. An i7 if you plan on Crossfiring eventually.

Martini1991 · 9 Feb 2014 at 17:43

Valley isn't good at all for testing bottlenecks with CPU's for games to be perfectly honest, even though it's completely agreeing with me with bottlenecks

As it can vary game from game, take two set ups and compare them in Frostbite 3, then compare them in Anvil Next, you'll get a massive difference.

Don't think I'd personally fancy an i5 4670K with an R9 290 Crossfire either like

Kaapstad · 9 Feb 2014 at 17:43

Choimaloy75 said:
This is what I've being trying for people to understand, some people aren't bothered about all this benching to prove what is better the op wants a bit more performance for games and getting another gpu regardless of weather he has amd cpu or not if gets the performance then great, from my point of view I changed from a 7990 to 2 290x's and I've seen good increase in performance I don't think the op wants to go out and spend more on a cpu and motherboard.

It is still worth knowing the answer, does a 8350 bottleneck the cards or not. I think someone with a 8350 and a couple of 290Xs should give it a go, I think they will be pleasantly surprised.

It is a whole lot better finding the answers than people getting in endless debates about it with limited facts.

Kaapstad · 9 Feb 2014 at 17:44

Martini1991 said:
He's buying a new motherboard.
OP asked if it bottlenecks, it does.

If you want to be ignorant, that's fine, but don't try and rubbish other people and calling them silly names like Intel preachers.

They can call me an intel preacher

My name is Robert Noyce lol.

Rasch · 9 Feb 2014 at 17:45

It does even bottleneck 1 290X.End of story

Avenged7Fold · 9 Feb 2014 at 17:46

Martini1991 said:
Valley isn't good at all for testing bottlenecks with CPU's for games to be perfectly honest, even though it's completely agreeing with me with bottlenecks

As it can vary game from game, take two set ups and compare them in Frostbite 3, then compare them in Anvil Next, you'll get a massive difference.

Like i said, very specific example but its all i have from personally benching it.

In COD ghosts with the AMD set up, my usage never went above 87% and averaged around mid 70's in game.

Now i am hitting 100% all the time

Rasch · 9 Feb 2014 at 17:48

Yeah i get bad gpu utilization in Ghosts too.

Choimaloy75 · 9 Feb 2014 at 17:50

So if I ran this bench mark my graphics score would less than someone with a 4930k

Kaapstad · 9 Feb 2014 at 17:50

Avenged7Fold said:
Very specific example here but i used Valley, which is as game like as benches get really. I think valley is threaded for four cores, which i suppose is a fair place to start thinking about CPU bottlenecks in games.

When i moved my GPU to my other rig, i benched it at different common clocks to compare.

[email protected] with a 290x OCed @1241/1500 Was scoring 25xx on Valley with the leaderboard's settings.

Same card with same clock was scoring 30xx in Valley when i put the card in my 4770k @ 4.4

Now in Valley my GPU usage is at around 99% when it isnt changing scenes, while it was no more than in the low 80's for my 8320.

I would look towards getting an i5, especially if you are going to dish out for a ud5 board. An i7 if you plan on Crossfiring eventually.

Valley is an awful bench to use as it is rubbish on any decent GPU setup regardless of whether it's on an intel or AMD CPU, this is why I did not use it. Intel will get an advantage on it though.

Tonester0011 · 9 Feb 2014 at 17:50

Choimaloy75 said:
This is what I've being trying for people to understand, some people aren't bothered about all this benching to prove what is better the op wants a bit more performance for games and getting another gpu regardless of weather he has amd cpu or not if gets the performance then great, from my point of view I changed from a 7990 to 2 290x's and I've seen good increase in performance I don't think the op wants to go out and spend more on a cpu and motherboard.

Run the benchmark Kaap provided and prove how little you bottleneck (even though i don't think FS is a fair test to prove such thing, as it's so heavily GPU bound)

The OP will be bottlenecked regardless of game and clock, don't see the merit in denying such things just because that's what you own?