• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

**The OcUK Benchmarks thread!**

Rusty0611 said:
Tonester has triple screens and crossfire 7970's.
Click to expand...
gregster said:
I will nag him :)
Click to expand...

I'll set up my 3rd screen again, i decluttered my desk and took down my 3rd monitor as i never really used eyefinity and only really used that screen for football results and the odd benchmark.


I'll bench a couple of games at the weekend, BF3, MoH etc.
Crossfire scaling on 1 screen with 12.11s isn't anything to shout about though, hopefully they scale better over a larger resolution.
 
The case I'm going for is in terms of volume probably around the same/similar to the ST10. Reeeeeaaalllly can't wait to get my hands on an XSPC H1+ Cube :D

My mrs is gonna freak out when she see's the size of it (which would be a first :().
 
A few HD7970 Eyefinity Cat 12.11 Beta 4 "benchmarks" from me. Main purpose is to see if the performance advantages seen at 1080p also hold true at a much higher resolution.

System setup


- i5 2500K @ 4.2Ghz
- MSI P67A-GD55
- 4GB DDR3 1600Mt/s 8-8-8-24
- 64GB SSD (Boot/BF3), 120GB SSD (Steam)
- VTX HD7970 X Edition "Ghz"

- 3 x 23" LG IPS231P in Portrait Eyefinity, effective resolution 3620x1920 (Nvidia) / 3560x1920 (AMD) - difference caused by bezel management.

Comparison cards:

MSI GTX680 - 1084/6000 stock / 1200/6800 OC
MSI GTX670 PE OC - 1169/6000 "stock"

Note - all results are from cards I have owned personally and have been run through exactly the same tests.

Drivers used:

HD7970 - Cat 12.8 & 12.11 Beta 4
GTX680/670 - 301.40 (newer drivers tested but found to have no material difference)


Benchmarks / Results


BF3

Test Scenario: 5 minute play through of the campaign level, Operation Swordbreaker. Fraps run begins the moment you leave the APC and ends after the large battle in the car park.

Ultra Settings - No MSAA

(Note - MSAA not used as it is completely unplayable on any of the cards. Technically the HD7970 manages 30fps average but it certainly didn't feel like 30fps when "playing".)

BF3Ultra.png


12.8 - 45.06 Avg / 28.00 Min
12.11 - 50.58 Avg / 32.00 Min

Improvement - 12.3%

BF3Ultra128vs1211.png


Looking closer at the actual performance, it is clear that the 12.11 drivers are consistently faster than the 12.8 drivers in this test.

Custom Settings - "high"

BF3Custom.png


12.8 - 59.00 Avg / 45.00 Min
12.11 - 67.74 Avg / 52.00 Min

Improvement - 14.8%

Again it is clear that the 12.11 drivers are offering a noticeable gain in performance. Note also the performance improvement over the HD6970, 90% average and 100% minimum!


BF3CustomHD7970vs670OC.png


This graph is comparing the HD7970 Ghz on the 12.11 drivers with a MSI GTX 670 PE OC card. This particular 670 benches very closely to a reference 680 (2FPS down on Average but identical minimums).

The HD7970 Ghz ends up 27.9% faster overall with minimums 33.3% higher.

Overall this is the most interesting BF3 result as it the first time I have seen the HD7970 to be noticeably faster than the 670 OC/680 in game. Even over clocked, albeit relatively mildly at 1200/6800, the 680 can't catch up managing 61.47 average and 45 minimum. I would imagine that ah highly clocked GTX680 should be able to catch the HD7970, but the HD7970 does appear to now have the legs a bit at this resolution.


Just Cause 2

Test Scenario: Dark Tower benchmark - Max in game settings (No Nvidia specific features enabled).

JC2.png


12.8 - 44.86 Avg
12.11 - 46.87 Avg

Improvement - 4.5%

A small improvement overall is shown in JC2 but it certainly isn't anything to get excited about. The HD7970 was already noticeably faster than the GTX680 here anyway. In game performance is closer (but very hard to benchmark consistently). Unfortunately I no longer have the 670/680 to run any comparison tests.


TDU2

Test Scenario: 5 minute drive from your beginning Garage across Ibiza, Car used: Nissan GTR. Max in game settings with 4xMSAA.

TDU2.png



12.8 - 54.06 Avg / 45 Min
12.11 - 50.72 Avg / 41 Min

Improvement - (6.18%)

Where there are winners, there are also losers. It would appear that TDU2 performance has dropped with these latest drivers. Annoyingly I thought things has improved to begin with as I was seeing framerates into the 70's but alas the game had decided to change the resolution & AA settings all by itself.

I have tried re-testing to see if the performance drop is consistant but now the game won't allow me to select 3560x1920 from the settings menu (despite it being an option). I will try again another day.


Overall I am impressed with the 12.11 drivers. BF3 gains are very impressive especially once the raw performance is compared to the comparatively priced GTX670. I haven't noticed anything untoward yet with these drivers but no doubt there are still a few bugs that need ironing out. Overall this is a job well done by AMD.
 
I will spend a bit of time today getting the results in one image for BF3. This is the most popular game amongst us (it seems) and nice to see a few of us taking time to do these. Well done all.
 
Was thinking of doing the same :p

So thanks for taking a little work load off my hands :)

Same applies to all games at set resolution, soon as we have a fair few cards, I'll compile them all into a single chart for easier viewing.

Still need some 660Ti/670/580 scores too guys ;)
 
I only added the operation sword breaker results for BF3.

BF3MULTIGPURUNS-2.jpg


I just had a look back through this thread and I am very dissapointed to see the one upmanship kicking in. Joking or not, this wasn't the thread for that :(
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom