OC'd i7 930 vs OC'd 1055T/1090T, need advice for workstation

[a_drone]

I'm trying to decide between a 930 and a 1090T (or 1055T) for my new workstation. Thing is that I'd overclock either set-up as high as possible (tho stability is paramount) so it's harder for me to use reviews and benchmarks to compare the two as most do not include the OC'd CPU's.



TASKS:

3D modeling. Photoshop. Simulations. 3D rendering. After Effects. 1080p editing.

I can fully use any core, thread and Mhz while rendering. However, smooth performance while working is also very important.

CORE SYSTEM:

Samsung 1000GB SATA-300 7200RPM 32MB HD103SJ Spinpoint F3
CoolerMaster HAF 922 (High Air Flow)
Corsair TX650 Watt PSU
Noctua NH-D14
EVGA GTX 460 1024MB 675/3600 EE (External Exhaust)
LG H22LS30 DVD


THE DILEMMA

I found a few benchmarks which pitted the OC'd CPU's against each other.

http://www.extremeoverclocking.com/reviews/processors/AMD_Phenom_II_X6_1090T_15.html

Both at 3.8 Ghz and it's obvious that the 1090T is faster at multi-threaded applications, but only by a small margin. This while the 930 is faster in single-threaded applications, but by a larger margin.

Am I interpreting this correctly?

http://www.techreaction.net/2010/05/06/amd-phenom-ii-six-core-1090t1095t-thuban-review/

In this review it seems as if the 920 is able to catch up with the 1090T when they're both OC'd.

http://www.overclock3d.net/reviews/cpu_mainboard/amd_1090t_six_core_review/5

Strangely enough in this review the 1090 consistently beats the 930. Even when not overclocked, which doesn't seem consistent with other reviews.
Tho the review doesn't test any applications like Photoshop or After Effects.



Reviews of non OC'd CPU's only increase the confusion. One time the 1090T seems superior and the other time the 930.

http://www.anandtech.com/show/3674/amds-sixcore-phenom-ii-x6-1090t-1055t-reviewed/7
http://www.tomshardware.co.uk/charts/desktop-cpu-charts-2010/compare,2420.html?prod[4477]=on&prod[4438]=on&prod[4437]=on
http://www.behardware.com/articles/789-5/amd-phenom-ii-x6-1090t-1055t.html
http://www.behardware.com/articles/789-11/amd-phenom-ii-x6-1090t-1055t.html

But maybe the non OC reviews aren't really valid.


So if anyone can please make sense of which CPU is faster, because to me it seems reviews are contradicting each other.


As for other advantages/disadvantages per CPU:

I can't afford 12GB of RAM but I can afford 8. So my 1090T build would have 8 and the X58 set-up only 6. The more the better.
Apparently when both OC'd to the same Ghz the 1090T and the 1055T are equally fast. This would be quite nice since the 1055T is €100 cheaper.

Downsides of the 890FX build are the lack of SLI (which could be a large disadvantage as I'm going to use GPU rendering) and the fact that I can only get a Asus Crosshair IV Formula which is reported to have north bridge overheating issues.




Anyways, thanks immensely for all your help.

 

[iam2thecrowe]

i would go with the AMD system with the extra RAM. You can always look forward to AMD releasing its next line of CPU's and upgrade If you feel the system is not fast enough, without changing the motherboard. If intel release another CPU it will most likely need a new motherboard or some crap making upgrade more expensive.

 

[a_drone]

i would go with the AMD system with the extra RAM. You can always look forward to AMD releasing its next line of CPU's and upgrade If you feel the system is not fast enough, without changing the motherboard. If intel release another CPU it will most likely need a new motherboard or some crap making upgrade more expensive.

AMD will release its bulldozer with a new socket, I'm afraid.

Otherwise the choice would be easy as I'd prefer a bit of a potential upgrade.

 

[13thmonkey]

I'm not sure that gpu rendering works in SLI, or rather with two cards not in SLI it may work, you migth want to research that a bit, I've found the live chat on the nvidia support site to be quite good for simple ish questions like that.

 

[a_drone]

I'm not sure that gpu rendering works in SLI, or rather with two cards not in SLI it may work, you migth want to research that a bit, I've found the live chat on the nvidia support site to be quite good for simple ish questions like that.

I've looked into it for the CUDA rendering software I plan to use. It seems to be the case for other software too.

It uses multiple cards, even when they're no in SLI. So there's no advantage to SLI.

Problem with that for now is that your scene can only be as large as the RAM of the graphics card with the lowest RAM. So the RAM doesn't stack.

It's still very useful tho.

I would prefer to have SLI because GPU rendering is still in its infancy and I'd like to be ready for whatever the future brings.

 

[mosox]

Both CPUs are very good and both are very fast. From some point on it doesn't even matter who's faster it's not like if you'll gain 3 milliseconds when opening winamp or 30 seconds while video editing matters.

You have some AM3 SLI mobos - MSI NF980-G65, ASUS M4N98TD EVO but CUDA doesn't work with SLI anyway. Also with the GPU rendering you get lower quality than with the CPU one. If you want high quality processing look up X264.