Bottleneck? or bad GPU

[Anarchyz11]

Seems odd that it isn't hitting 100% usage.  Is your CPU at 100% usage?  If so that's the issue, but for BF4 I wouldn't think it is.

Most likely a bad card if the 270x is outperforming it.

 

Should have stuck with the 970.

[Ertman]

Bottlenecking shouldn't be a factor here, as, if I am understanding it correctly, the 270x was performing better than the current 280x. The better gpu should only do as well or better, not worse.

[GrantSor+1]

Seems odd that it isn't hitting 100% usage.  Is your CPU at 100% usage?  If so that's the issue, but for BF4 I wouldn't think it is.

Most likely a bad card if the 270x is outperforming it.

 

Should have stuck with the 970.

 

yes it was running at 96% to 99% based on afterburner

Bottlenecking shouldn't be a factor here, as, if I am understanding it correctly, the 270x was performing better than the current 280x. The better gpu should only do as well or better, not worse.

yep. you had it right.

[Ertman]

yes it was running at 96% to 99% based on afterburner

yep. you had it right.

Just throwing out ideas...

Are you running at the same settings as the 270x?

Not sure If you have addressed this, but have you tried reinstalling drivers?

[Anarchyz11]

yes it was running at 96% to 99% based on afterburner

 

Wait, so the 280x is at 60-70% usage, and your CPU is at 99%?

It that's the case, it's a CPU bottleneck clearly.

[CalculuS]

TBH, i'm starting to think that your 270x might just be a very good card and the 280x is a very bad card.

 

OR

 

This is the unspoken rule that certain hardware just runs better with certain hardware, which just like the silicon lottery is based on luck.

[BCPaury]

What about other games, isn't R9 280x downclocked somehow?

[D3MON]

Yes and Yes

 

you're overlooking the problem,

problem is, FPS of my 270x is higher compared to my 280x

wasnt too clear in original post but didnt you run the 280x on ultra and the 270 on high? (though it wont make a ton of difference it can easily be 10fps)

also real multiplayer can be hugely variable where a few subtle difference such as whats running in the background with a differnt server with differnt number o player can change results allot trust me ive done enough benchmarking for a lifetime (sympathise with luke).

are you able to do the comparison again?

personally i would just set it to 1600x900 or 1080p and downscale it, itll look better and your hardware will actually be "earning its keep".

[D3MON]

TBH, i'm starting to think that your 270x might just be a very good card and the 280x is a very bad card.

 

OR

 

This is the unspoken rule that certain hardware just runs better with certain hardware, which just like the silicon lottery is based on luck.

unless his 270 is at 1600mhz it wont be equal to a 1ghz 280x (ignoring memory bandwidth)

[GrantSor+1]

Just throwing out ideas...

Are you running at the same settings as the 270x?

Not sure If you have addressed this, but have you tried reinstalling drivers?

Yes. and Yes.

 

Wait, so the 280x is at 60-70% usage, and your CPU is at 99%?

It that's the case, it's a CPU bottleneck clearly.

That's a HUGE bottleneck, Ive seen people build a 6300 and a 280x and they seem fine

 

TBH, i'm starting to think that your 270x might just be a very good card and the 280x is a very bad card.

 

OR

 

This is the unspoken rule that certain hardware just runs better with certain hardware, which just like the silicon lottery is based on luck.

my 270x OC memory oc doesnt above +31mhz and for the clock speed only + 60 with max Power limit(But I dont use these OC settings because there's no point). Havent tested 280x's oc capabilities tho

 

What about other games, isn't R9 280x downclocked somehow?

The Highest AAA game I could play atm is BF4, Titanfall is still updating. also dota 2 is the same maxing 120fps(default max cap for dota 2)

[GrantSor+1]

wasnt too clear in original post but didnt you run the 280x on ultra and the 270 on high? (though it wont make a ton of difference it can easily be 10fps)

also real multiplayer can be hugely variable where a few subtle difference such as whats running in the background with a differnt server with differnt number o player can change results allot trust me ive done enough benchmarking for a lifetime (sympathise with luke).

are you able to do the comparison again?

personally i would just set it to 1600x900 or 1080p and downscale it, itll look better and your hardware will actually be "earning its keep".

ohh, I get that a lot of times

Summary

270x @ high 4xMsaa = 50-70fps

280x @ high 4xMsaa = 44-55fps

Gtx 970 @ ultra preset = 60+fps

 

unless his 270 is at 1600mhz it wont be equal to a 1ghz 280x (ignoring memory bandwidth)

[Anarchyz11]

 

 

That's a HUGE bottleneck, Ive seen people build a 6300 and a 280x and they seem fine

 

It doesn't matter what benchmarks say, if your 6300 is pinned at 96-99%, it's bottlenecking and outputting all the performance it can.  If that seems unusual (which it seems it does), the CPU itself may have issues such as throttling and downclocking.

 

 

But one way or another, idc if it's a 5960k, if the CPU is at 100% usage and the GPU isnt, the CPU is the bottleneck.  That's the definition. 

[GrantSor+1]

It doesn't matter what benchmarks say, if your 6300 is pinned at 96-99%, it's bottlenecking and outputting all the performance it can.  If that seems unusual (which it seems it does), the CPU itself may have issues such as throttling and downclocking.

 

 

But one way or another, idc if it's a 5960k, if the CPU is at 100% usage and the GPU isnt, the CPU is the bottleneck.  That's the definition. 

Temps at idle are at 10c. Max ive seen is at 49c (deepcool maelstrom 240 w/ sp120s in push)

As for downclocking, when all cores are active I dont seem to get a max of 3.5Ghz, I only see 3.472Ghz on task manager

I also managed to clocked it at 4.1Ghz at stock voltage but I decided to turn it back to stock because im scared that my PSU or Board may fail

[Anarchyz11]

Temps at idle are at 10c. Max ive seen is at 49c (deepcool maelstrom 240 w/ sp120s in push)

As for downclocking, when all cores are active I dont seem to get a max of 3.5Ghz, I only see 3.472Ghz on task manager

I also managed to clocked it at 4.1Ghz at stock voltage but I decided to turn it back to stock because im scared that my PSU or Board may fail

I have no clue why, but again if your CPU is at 100% load, it's the bottleneck, it's the peice of hardware that can't push your framerate further. 

[D3MON]

Temps at idle are at 10c. Max ive seen is at 49c (deepcool maelstrom 240 w/ sp120s in push)

As for downclocking, when all cores are active I dont seem to get a max of 3.5Ghz, I only see 3.472Ghz on task manager

I also managed to clocked it at 4.1Ghz at stock voltage but I decided to turn it back to stock because im scared that my PSU or Board may fail

didnt read a part saying your cpu was at 100%, get the game running in the test base area in windowed mode and see if your cpu is at 100% then and sort in task manager by cpu usage and see whats causing it.

if its not at 100% then go into a typical server game and try to replicate the problem and if it occurs then go to task manager and see what it its.

its possible that even if the cpu is being maxed by battlefield its actually something else as i have tested it and this didnt happen but i know that sometimes a dodgy hard drive thats very old and slow can cause game hitching and stuttering which makes the cpu hit 100% as it loads sections of the game then drop to a typical 50% then back up to 100% very quickly, my old q6600 used to do it with this old WD green (before caviar) but once i got a new drive with 64mb of cache it was fine.

[totjup5]

A bottleneck wouldn't make your frame rate lower than a less powerful card, only limit the increase that you would see. It could be a bad GPU or a driver issue.

[miagisan]

At that resolution the game is CPU bound and you won't get full usage on a 280x.