AMD Catalyst 11.12 8.921.5 29 December Benchmark
Yesterday, a new 11.12 AMD Catalyst that was built on 29 December surfaced, this driver is WHQL certified. Knowing that the best performing driver to date is a 11.12 one (11.12 Beta 18 Nov), I think it is a good opportunity to find the next best driver or even the best one yet.
Having seen that 8.95 branch drivers don’t bring big improvements, 11.12 8.921.5 29 December is up for testing. Right after the driver was available I did some “preliminary testing” in Battlefield 3 and was surprised to see how well it performed. For me, the best Battlefield 3 driver is 11.12 Beta 18 Nov but I think that that is about to change.
To test 11.12 8.921.5 29 Dec driver I will compare it to 11.12 WHQL and 11.12 Beta 18 Nov.
Prior to installing a driver, a sweep was done in safe-mode using Phyxion Driver Sweeper.
| Catalyst Version | 2D Driver | 3D Driver | OpenGL Driver | Driver Packaging Version | CCC Version | CCC |
| 11.12 WHQL | 8.01.01.1215 | 7.14.10.0879 | 6.14.10.11318 | 8.92-111109a-129966C-ATI | 2011.1109.2212.39826 | 11.12 |
| 11.12 Beta 18 Nov | 8.01.01.1221 | 7.14.10.0883 | 6.14.10.11384 | 8.93-111118a-128864E-ATI | 2011.1118.1246.22742 | N/A |
| 11.12 8.921.5 29 December | 8.01.01.1215 | 7.14.10.0885 | 6.14.10.11332 | 8.921.5-111229a-131232C-ATI | 2011.1229.1025.18582 | 11.12 |
All 3 drivers are tested in 3 synthetic benchmarks and 6 games.
- 1. 3DMark06 1.02 Professional Edition - 1680×1050 – 2xAA – Anti Aliasing Quality = 0 – 4xAF – Shader Model 2.0 = Return to Proxycon + Firefly Forest – HDR/Shader Model 3.0 = Canyon Flight + Deep Freeze – 2 Runs.
- 2. Unigine – Heaven Benchmark 2.5 – 1680×1050 – High Shaders – Normal Tessellation(DirectX 11) – 2xAA – 4xAF – 2 Runs.
- 3. Cinebench 11.5 – 3 Runs.
- 4. Crysis 2 1.9 using Adrenaline Benchmark Tool – Extreme Preset – 1680×1050 – 4xAA – Hi-Res Textures – Central Park – 3 Runs.
- 5. Metro 2033 using Update 1.2 Benchmark Tool – Very High Settings – 1680×1050 – 16xAF – No AA for DirectX 9 and 4xMSAA for DirectX 10 and DirectX 11 – 3 Runs.
- 6. S.T.A.L.K.E.R. – Call of Pripyat – 1680×1050 – Extreme Preset (DirectX 9) – Ultra Preset (DirectX 10 & 11)
- 7. DiRT 3 – Ultra Preset – 1680×1050 – 8xMSAA – 3 Runs
- 8. Battlefield 3 – Ultra Preset – 1680×1050 – 4xMSAA – 16xAF – Multiplayer @ Strike at Karkand (32p map, Large Conquest), main street: US Deployment -> B Flag – 3 Runs x 60s.
- 9. Elder Scrolls V: Skyrim v1.3.10 – Ultra Preset – 1680×1050 – 8xAA – 16xAF – Tested in the area East of Rorikstead – 3 Runs x 60s.
All synthetic benchmarks were ran 2 times except Cinebench (3 runs), all games were ran 3 times.
| Test Hardware | AMD Catalyst 8.94 22 December Driver Benchmark | |
|---|---|
| Processor | Intel Core i5-2500K (Sandy Bridge)
4.5 GHz, 6 MB L3 Cache, power-saving settings disabled, Turbo Boost disabled. |
| Motherboard | MSI P67-C43-B3, Intel P67 Chipset |
| Memory | 2 x 2 GB DDR3 1600MHZ |
| Hard Drives |
WD 500 GB SATA III (OS) Samsung 750 GB Sata II (Game) |
| Graphics Card | Sapphire HD6950 1 GB |
| Power Supply | Corsair TX 650 W |
| System Software And Drivers | |
| Operating System | Windows 7 SP1 x64 |
Table of Contents:
Page 1 – Introduction
Page 2 – 3DMark06, Unigine and Cinebench
Page 3 – Crysis 2, Metro 2033 and Call of Pripyat
Page 4 – Battlefield 3, DiRT 3 and Skyrim
Page 5 – Image Quality
Page 6 – Conclusion
Page 2 – 3DMark06, Unigine and Cinebench
Hey, I am usually trusting your measured averages, but I am wondering if the somewhat bigger differences in the minimum department are consistent and conclusive or just a random effect caused by e.g. VRAM/RAM emptying and cleaning or just spikes in voltage and adjustment, or something else comparable in randomness.
Btw, I just wrote a similar text and I forgot to answer the Security Question, a new page is loaded with the error message, while my text is gone forever
. Just wondering if something could be done to prevent that. E.g. a popup message or new window containing the error instead of reloading the page with the text box.
The minimum values are pretty consistent, if any of the 2-3 runs show big differences between Minimum and/or Average values I re-do the test. Usually all runs show about the same results. Tests aren’t 100% bulletproof but you’d be surprised to know how small are the differences between each run (with few exceptions).
As for the deleted comment, if you hit the back button the full comment should be “written in”.
good work !!
appears that the 11.12 29 dic. drivers are pretty stable, are the ones I’m using now.
Don’t forget to write the link where you get these drivers, the guys over at guru3d are doing a great job and we should be grateful to them also.
And thanks to you, please keep it up you are doing a great service here!
Great work. I am switching to this drivers.
Can you plz upload the drivers to mediafire or multiupload ,, it’s faster than filesonic which is very slow < 100 ko/s ! Thx
Great Job Man
PLEASE!
Use graph that start from ZERO!!!!
thanks, between here and guru3d i can always guarantee i am running the best drivers, cheers
Just tested this drivers with my 7970, the results are very bad compared to previous 8.921.2.
In Unigine the score has passed from 1025 to 961 (FullHD 8x AA extreme aniso 16x)
In Batterfield3 the minimum fps has passed from 47 to 35-40 fps.
Please can you make the benchmark in the next review with Tahiti GPU too (if you have one)?
With this chip the drivers works completly different.
I roll-back and reinstalled again the 8.921.2
Best regards, Pasquale from Spain.
Can you please test ”new” 8.95, 17 January drivers?
Thanks in advance!
It’s on the way…
I am getting micro stutters in bf3 with my Sapphire 6950 1gb using these drivers. It seems 11.12 or 12.1a are best for me.