***Powercolor PCS+ R9 290 Mini Review***

[LtMatt]

Good advice from setter and i agree, but ill also say sandy still does a worthy job. I think in games the difference between them is less than that. In specific cpu benches and other stuff like that, the difference is bigger. In a game like bf4 i expect the difference will be no more than 1-3 fps on the minimums and averages.

 

[Diagro]

Hmmm what glue is that on the coils?
May take my cards out and glue them up to see if it helps with my coil whine.

I'm going to strip my pc down today or tomoz and getting a hot glue gun is easy

 

[Tute]

Damn you, I read the title and thought they were releasing a short version of the PCS+ card.

 

[LtMatt]

Both of my VTX reference cards have a bit of glue on some (but not all) of the coils as well. Probably explains their complete silence. Does make me wonder why its only on some and not all though.

 

[shankly1985]

My two also have no coil wine not sure if they have glue.

 

[Diagro]

I got no idea if mine do or not... Dont fancy pulling them apart to look and find they are sealed chokes tho.

 

[pgi947]

Or take a look at x79, its becoming eol soon and the sbe gear is getting dirt cheap, a 3930k, p9x79 and 16gb of ram recently sold for £400 on the mm!

Good review, but I have to ask the same as gregster, what's going on with the DX frame times? They're almost double what I see on nvidia even with 337 and its known frame time issues.

Glue on the coils is a good idea but from those pics it looks like an after thought and they've sent a guy round with a hot glue gun after manufacturing, very messy work.

 

[LtMatt]

Good review, but I have to ask the same as gregster, what's going on with the DX frame times? They're almost double what I see on nvidia even with 337 and its known frame time issues

Neither you or Gregster have a Phenom II cpu with slow memory though. I expect its quite a bottleneck on Siege Of Shanghai 64 player server at 1080P.

What's interesting to see ( i think) is his latency via Mantle with that slow cpu/memory is a fair bit lower than what you posted with a Haswell cpu at 4.9ghz, faster 1866mhz memory and two clocked 780's in SLI.

 

[wazza300]

you only need a z68/77 board and 2600k/3770k whatever you can get your mitts on cheapest

there's no diff once overclocked

 

[humbug]

Neither you or Gregster have a Phenom II cpu with slow memory though. I expect its quite a bottleneck on Siege Of Shanghai 64 player server at 1080P.

What's interesting to see ( i think) is his latency via Mantle with that slow cpu/memory is a fair bit lower than what you posted with a Haswell cpu at 4.9ghz, faster 1866mhz memory and two clocked 780's in SLI.

Didn't see his Latency.

In DirectX i average 13.24ms, in Mantle i average 7.29ms, its about half of what DirectX is, i think just over 7ms is excellent, if its better than a very overclocked Haswell, what's the problem?

Gregster?

 

[pgi947]

It was more the Star Swam latencies, I was under the impression bottlenecks on that all came from an API bottleneck rather then being CPU bound. If this is true (and to be fair what we've been getting told since it came out!) then SS AMD DX is giving you 75ms, while on nvidia its giving ~30ms on the same API.

BF4 single card latencies are around the same as an nvidia card in DX both hitting around 12-13ms.

I'm not commenting or taking anything away from the mantle results as it shows its doing what it says on the tin, more a concern that DX may still be fudged somewhat on the 14.x drivers.

 

[humbug]

It was more the Star Swam latencies, I was under the impression bottlenecks on that all came from an API bottleneck rather then being CPU bound. If this is true (and to be fair what we've been getting told since it came out!) then SS AMD DX is giving you 75ms, while on nvidia its giving ~30ms on the same API.

BF4 single card latencies are around the same as an nvidia card in DX both hitting around 12-13ms.

I'm not commenting or taking anything away from the mantle results as it shows its doing what it says on the tin, more a concern that DX may still be fudged somewhat on the 14.x drivers.

The Star Swarm results quite clearly states; DX 13.24ms and Mantle 7.29ms, 75ms does not exist, no idea where you got that from?

"BF4 single card latencies are around the same as an nvidia card in DX both hitting around 12-13ms."

Is that on a Phenom II? no, it isn't.

 

[Frenzy]

If you are going to get a sandy/ivy setup then don't touch the z68 chipset, get a z77 or something else.

My gigabyte z68 was blue screening all over the place with a 290x and a 290 on all drivers. Swapped for a z77 and it was 100% fine.

 

[humbug]

If you are going to get a sandy/ivy setup then don't touch the z68 chipset, get a z77 or something else.

My gigabyte z68 was blue screening all over the place with a 290x and a 290 on all drivers. Swapped for a z77 and it was 100% fine.

Thanks for that, i didn't know that about the z68

 

[LtMatt]

I'm Z77, as is Lamb Shanks and no problems here so another +1 for Z77.

Here are my starswarm results.


DirectX 11.1

===========================================================
Oxide Games
Star Swarm Stress Test - ©2013
C:\Users\matt\Documents\Star Swarm\Output_14_04_14_1340.txt
Version 1.10
04/14/2014 13:40
===========================================================

== Hardware Configuration =================================
GPU: AMD Radeon R9 200 Series
CPU: GenuineIntel
Intel(R) Core(TM) i7-2700K CPU @ 3.50GHz
Physical Cores: 4
Logical Cores: 8
Physical Memory: 17118543872
Allocatable Memory: 140737488224256
===========================================================


== Configuration ==========================================
API: DirectX
Scenario: ScenarioFollow.csv
User Input: Disabled
Resolution: 1920x1080
Fullscreen: True
GameCore Update: 16.6 ms
Bloom Quality: High
PointLight Quality: High
ToneCurve Quality: High
Glare Overdraw: 16
Shading Samples: 64
Shade Quality: Mid
Deferred Contexts: Disabled
Temporal AA Duration: 16
Temporal AA Time Slice: 2
Detailed Frame Info: Off
===========================================================


== Results ================================================
Test Duration: 360 Seconds
Total Frames: 15324

Average FPS: 42.57
Average Unit Count: 4273
Maximum Unit Count: 5662
Average Batches/MS: 640.49
Maximum Batches/MS: 1689.98
Average Batch Count: 17323
Maximum Batch Count: 133495
===========================================================



Mantle

===========================================================
Oxide Games
Star Swarm Stress Test - ©2013
C:\Users\matt\Documents\Star Swarm\Output_14_04_14_1333.txt
Version 1.10
04/14/2014 13:33
===========================================================

== Hardware Configuration =================================
GPU: AMD Radeon R9 200 Series
CPU: GenuineIntel
Intel(R) Core(TM) i7-2700K CPU @ 3.50GHz
Physical Cores: 4
Logical Cores: 8
Physical Memory: 17118543872
Allocatable Memory: 140737488224256
===========================================================


== Configuration ==========================================
API: Mantle
Scenario: ScenarioFollow.csv
User Input: Disabled
Resolution: 1920x1080
Fullscreen: True
GameCore Update: 16.6 ms
Bloom Quality: High
PointLight Quality: High
ToneCurve Quality: High
Glare Overdraw: 16
Shading Samples: 64
Shade Quality: Mid
Deferred Contexts: Disabled
Temporal AA Duration: 16
Temporal AA Time Slice: 2
Detailed Frame Info: Off
===========================================================


== Results ================================================
Test Duration: 360 Seconds
Total Frames: 17400

Average FPS: 48.33
Average Unit Count: 4449
Maximum Unit Count: 5599
Average Batches/MS: 933.97
Maximum Batches/MS: 2983.30
Average Batch Count: 21678
Maximum Batch Count: 99092
===========================================================

 

[pgi947]

The Star Swarm results quite clearly states; DX 13.24ms and Mantle 7.29ms, 75ms does not exist, no idea where you got that from?

"BF4 single card latencies are around the same as an nvidia card in DX both hitting around 12-13ms."

Is that on a Phenom II? no, it isn't.

So from your OP...

Star Swarm Extreme Follow 1080

Mantle: Final Score Latency of 25.38ms

D3D: Final Score Latency of 76.20ms

Or am I missing something in the big black letters in the screens in your OP?

 

[LtMatt]

During my dx11 run of Star swarm fps were down in the single digits at times, so i would expect the latency to be fairly high at points. Mantle didn't drop below 29 fps afaik. Its just a silly benchmark though to be honest so who cares about latency. Theres ten minutes of my life i won't get back thats for sure.

 

[pgi947]

During my dx11 run of Star swarm fps were down in the single digits at times, so i would expect the latency to be fairly high at points. Mantle didn't drop below 29 fps afaik. Its just a silly benchmark though to be honest so who cares about latency. Theres ten minutes of my life i won't get back thats for sure.

I'm not the one getting bent out of shape and on the defensive about it though matt, my overlying question was is there still issues with DX on the 14.x drivers.

 

[humbug]

It was more the Star Swam latencies, I was under the impression bottlenecks on that all came from an API bottleneck rather then being CPU bound. If this is true (and to be fair what we've been getting told since it came out!) then SS AMD DX is giving you 75ms, while on nvidia its giving ~30ms on the same API.

BF4 single card latencies are around the same as an nvidia card in DX both hitting around 12-13ms.

I'm not commenting or taking anything away from the mantle results as it shows its doing what it says on the tin, more a concern that DX may still be fudged somewhat on the 14.x drivers.

"SS AMD DX is giving you 75ms, while on nvidia its giving ~30ms on the same API."

Completely useless information. if the Nvidia setup is using a £450 i7 then i'm doing very well in getting 25ms latency on a 4 year old AMD CPU, thats the whole point of Mantle, that proves its working, 25ms is better than NV ~30ms, whats the problem?

"BF4 single card latencies are around the same as an nvidia card in DX both hitting around 12-13ms."

As above, yet it doesn't anyway, a 780TI gets less FPS http://forums.overclockers.co.uk/showpost.php?p=26157610&postcount=58

you have no point to make here. just a lot of noise about nothing.

 

[LtMatt]

I'm not the one getting bent out of shape and on the defensive about it though matt, my overlying question was is there still issues with DX on the 14.x drivers.

As far as i know there never has been. I think the first Mantle drivers, 14.1 had some buggy dx performance in BF4 but that was rectified two weeks later with 14.2. Anything else just sounds like someone is making stuff up to me. The majority of benchmarks on the 295x2 (14.4 Beta) vs 780TI SLI (wonder drivers) had the 295x2 winning on most sites apart from TechReport who had three GameWorks titles in their testing suite. So i guess the dx performance isn't as bad as people would have you believe.