My new i7 machine..... any recommendations

if you are then deciding between the p6t and the rog 2, I would say the price diff between the gene2 and the p6t is not worth it then going for the gene 2.

Stick with the p6t then.

Ram, I believe that 6gb is more than adequate for your needs.

I would pass by the 4870's , and only consider the gtx 275 , 4890 and the gtx 260(core 216) maybe.

Only consider the 4870 if it's 1gb and like really cheap.

Other than that , I would say go for the more current cards.

Some of the okes on the Overclocking threads over at prophecy have some good things to say about the 4890's so it's worth a look as well.
 
The 4870 1gig actually outperforms the 216, but anyway. I say take the 4890. Directx 10.1 is actually useful these days.
 
What is dx 10.1 useful for again?

To bwecome future proof?

http://www.microsoft.com/games/en-US/aboutgfw/Pages/directx10.aspx

http://video.google.com/videosearch...&sa=X&oi=video_result_group&resnum=4&ct=title

http://en.wikipedia.org/wiki/DirectX#DirectX_10

DirectX 10
See also: Direct3D 10, Direct3D and Vista, and List of games with DirectX 10 support
A major update to DirectX API, DirectX 10 ships with and is only available with Windows Vista; previous versions of Windows such as Windows XP are not able to officially run DirectX 10-exclusive applications.[5] There are unofficial ports of DirectX 10 to XP[6], however these do not always work reliably[citation needed]. Changes for DirectX 10 were extensive.

Many former parts of DirectX API were deprecated in the latest DirectX SDK and will be preserved for compatibility only: DirectInput was deprecated in favor of XInput, DirectSound was deprecated in favor of XACT and lost support for hardware accelerated audio, since Vista audio stack renders sound in software on the CPU. The DirectPlay DPLAY.DLL was also removed and was replaced with dplayx.dll; games that rely on this DLL must duplicate it and rename it to dplay.dll.

In order to achieve backwards compatibility, DirectX in Windows Vista contains several versions of Direct3D:[7]

Direct3D 9: emulates Direct3D 9 behavior as it was on Windows XP. Details and advantages of Vista's Windows Display Driver Model are hidden from the application if WDDM drivers are installed. This is the only API available if there are only XP graphic drivers (XPDM) installed, after an upgrade to Vista for example.
Direct3D 9Ex (known internally during Windows Vista development as 9.0L or 9.L, the L standing for Vista's codename: Longhorn): allows full access to the new capabilities of WDDM (if WDDM drivers are installed) while maintaining compatibility for existing Direct3D applications. The Windows Aero user interface relies on D3D 9Ex.
Direct3D 10: Designed around the new driver model in Windows Vista and featuring a number of improvements to rendering capabilities and flexibility, including Shader Model 4.
Direct3D 10.1 is an incremental update of Direct3D 10.0 which is shipped with, and requires, Windows Vista Service Pack 1.[8] This release mainly sets a few more image quality standards for graphics vendors, while giving developers more control over image quality.[9] It also adds support for parallel cube mapping and requires that the video card supports Shader Model 4.1 or higher and 32-bit floating-point operations. Direct3D 10.1 still fully supports Direct3D 10 hardware, but in order to utilize all of the new features, updated hardware is required.[10] As of June 16, 2009, only ATI's Radeon HD 4000 and HD 3000 series, NVIDIA's GeForce 200M series and S3's Chrome 4xx GTX series of GPUs are fully compliant, NVIDIA has yet to release a DirectX 10.1 compliant desktop card.


[edit] Direct3D 11
Main article: Direct3D 11
Microsoft unveiled Direct3D 11 at the Gamefest 08 event in Seattle, with the major scheduled features including GPGPU support, tessellation[11][12] support, and improved multi-threading support to assist video game developers in developing games that better utilize multi-core processors. Direct3D 11 will run on Windows Vista, Windows 7, and all future Windows operating systems. Parts of the new API such as multi-threaded resource handling can be supported on Direct3D 9/10/10.1-class hardware. Hardware tessellation and Shader Model 5.0 will require Direct3D 11 supporting hardware. [13] Microsoft has since released the Direct3D 11 Technical Preview. [14]. Direct3D 11 is a strict superset of Direct3D 10.1 - all hardware and API features of version 10.1 are retained, and new features are added only when necessary for exposing new functionality. Direct3D 11 is scheduled to be released July 2009
 
PCINT have a radeon 4870 2GB for R2506, that sounds nice!
similarly ikonicit have the GTX 275 for R3k, that sounds nice too
Regarding the RAM I think i'm gong to need the 12GB's. Revit eats RAM alot.
Finally will a 600W power supply be enogh or should I go for a 700W
Thanks for the help guys, I really appreciate it.
 
Why i ask have nvidia not chosen to worry about dx 10.1 then? Rather go 700 and be safe than go too low plus you will won't need to upgrade the psu should you go sli or crossfire in the future.
 
Why i ask have nvidia not chosen to worry about dx 10.1 then? Rather go 700 and be safe than go too low plus you will won't need to upgrade the psu should you go sli or crossfire in the future.

Because ati did it first. If they had done it they would've been second. So instead they say it's not important. Games like h.a.w.x get a 15% fps increase in dx10.1 mode.

Anyway nvidias latest laptop gpus have dx10.1.

Another reason going dx10.1 is smart is because some of dx11 features will be backwards compatible.
 
Look at this, maybe it helps

Micropoint is Zotac importers to SA.



http://www.techpowerup.com/reviews/Zotac/GeForce_GTX_275_Amp_Edition

http://www.techpowerup.com/reviews/Zotac/GeForce_GTX_275_Amp_Edition/29.html

perfrel.gif
 
With regards to power supply it depends on the brand. A corsair 650watt will be more than enough.
 
These are way out of my league, each of these boards costs almost 6.5K.

I've managed to source the rampage gene II for almost 20% cheaper than the P6T, so it seems I will be heading for a uATX board.
The Cooler-master HAF 922 chassis is a bit expensive, is there anything as good for half the price? I'm looking at the cooler-master gladiator 600, but reviews suggest its a tad on the cramped side.
I'm still confused about the graphics, is the 4870X2 2GB the same as the one for sale at pcint?
 
I would not buy the 4870x2 now hey not with the 4890x2 coming.

Well are you going to be overclocking or run water cooling at any point?
 
Ever looked at this? Why Nvidia?


http://www.nvidia.com/object/io_1236190120769.html?rssid=cuda_nerachvid0303

http://www.nvidia.com/object/cuda_h...Open;aid=b02413b0-0e15-11de-8c30-0800200c9a66

http://www.nvidia.com/object/cuda_h...Open;aid=4f4d41b0-f961-11dd-87af-0800200c9a66

Soft-mod (modify in software) a NVIDIA GeForce desktop graphics card into a NVIDIA Quadro professional graphics card

http://www.techarp.com/showarticle.aspx?artno=539

PhysX in gaming
http://www.nvidia.com/object/physx_new.html



http://www.nvidia.com/object/builtforprofessionals.html

Be Faster: nVidia Intros All-CUDA Quadro Graphics

As part of my report on SolidWorks World, I had included a chart of graphics board brandnames from nVidia:
1. GeForce = consumer and entertainment.
2. Quadro = professional applications, like CAD.
3. Tesla = high performance computing.



nVida invited CAD journalists to learn about its new FX x800 line of six Quadro graphics boards. nVidia marketing named the online event "The Power of 10 Launch," referring to the ten years the company has been in existence. During the decade, they have boasted a number of firsts, including these:

1st Linux and 1st 64-bit drivers for professional graphics boards
1st real-time rendered movie, "Final Fantasy," and 1st interactive raytracer
1st graphics board with 4GB RAM


The company was pretty pleased to show off a graph with their sales of professional graphics boards swamping that of an unnamed competitor -- ATi. I asked about Matrox, and was told they are mainly into financial graphics.



Quadro

The point to the Webcast was to describe nVidia's newest line of Quadro graphics boards, which boast dual-PCI Express graphics, support for multiple operating systems (at the same time), multi-threaded drivers, and the ability to handle up 144GB RAM -- "support for massive GPU memory footprint" indeed.

The new boards range in price from a dollar under $100 to an astounding $3,300. Don't worry: there will be no peer (or other) pressure on you to buy the $3.3K board, because its 240 parallel CUDA cores are meant for 4D seismic and medical imaging. By "4D," nVidia means 3D graphics + time: graphics boards that show, say, the development of a tumor.

Instead, we CAD users will be marketed with boards that range from $99 to $899:

Quadro FX 380 ($99) for "volume" CAD -- 50% faster than its predecessor.
Quadro FX 580 ($149) for "volume" CAD -- adds 30-bit color through DisplayPort.
Quadro FX 1800 ($599) for midrange CAD -- 60% faster than FX 380.
Quadro FX 3800 ($899) for high-end MCAD -- adds stereo imaging and multi-OS support.



All of the new models in this line of Quadro graphics boards contain the CUDA ISA, even the cheapest FX380 with its 16 CUDA processors.



About CUDA

CUDA [compute unified device architecture] is the programming interface that runs software on the graphics board. nVidia sees this as a good thing, because their GPUs are so much faster than Intel's CPUs. A chart showed Intel's i7 965 CPU maxing out at 104GFLOPS (giga floating point operations per second), with the nVidia FX3800 archiving 933GFLOPS -- about 9x faster. nVidia says developers are seeing an 18x-150x speed up in graphics display using CUDA.

You can program for CUDA using these programming languages:

C for CUDA (available now).
OpenCL, also being implemented by Apple in its OS (beta available in summer).
DirectX 11 Compute.
Other languages being added, such as Python, C++, and Fortran.


The drawback to CUDA: the software running on it has to do parallel processing. nVidia claims that 25,000 developers are busy porting code from Intel CPUs to CUDA GPUs. However, for the most part they are specialized medical and imaging programs. Still, nVidia is handing out cash -- $500 thousand to $5 million -- to help develop software to exploit their GPU through their new GPU Ventures Program < http://www.nvidia.com/object/gpu_ventures_program.html >.



What's In It for CAD Users?

Other than buying an FX 380 or 580 for their somewhat faster speeds, I wonder how else do CAD users benefit from CUDA-ization. It's hard to say. nVidia described five advanced graphics technologies that CAD users could benefit from:

1. NVSCALE allows apps to work across multiple GPUs (available spring 09).

2. NVIRT performs realtime raytracing with specular reflections, like metal and glass (fall 09). This is a result of nVidia's purchase of mental ray a year ago; AutoCAD uses mental ray for its rendering.

3. 3D Vision Pro provides a professional version of 3D glasses: wireless active shutter 3D glasses that include a compass, accelerometer, and API. The IR [infra red] transmitter lets several people view the same 3D drawing at the same time (second half 09).

4. SDI Input Card reads video from multiple digital cameras, and then stores the video on the GPU. Useful for comparing inspections of products (summer 09).

5. Multi-OS allows graphics from more than one operating system to be virtualized and accelerated on GPUs. Until now, you needed either two GPUs or one GPU that accelerated only one OS (available now). The practical effect is that one computer can run Windows and Mac and/or Linux-based CAD with the graphics board doing the hard work of handling rendering, etc. for all of them.



When I asked which CAD vendors were implementing them ("No need to name names," I added), the response was vague. There is apparently some initial interest from CAD vendors.



Paraphrased Q&A

Q: Your comparison chart showed "Competitor" as singular. Do you have just one competitor?

A: Yes, ATI seen as the only competitor in the workstation market.



Q: What about Matrox?

A: They no longer are; they specialize in multi-displays for financial companies.



Q: What is DisplayPort?

A: DisplayPort is the new interface port, which Apple and PC vendors are starting to implement. It allows 30-bit color.



Q: Why are your GPUs so much faster than CPUs?

A: GPUs are faster because they use a parallel architecture; CPU is serial.



Q: This isn't the best time to introduce cutting edge cool stuff?

A: This lets customers become more productive (less time to design), which provides costs savings.



Q: Even the cheapest $99 board has CUDA in it?

A: Yes. There is a perception that Quadra is really expensive; this is not the case.



Q: Is CAD taking advantage of CUDA?

A: CAD is not being ported to CUDA right now; science and education currently primarily doing it. Ray tracing within the CAD app could take advantage of Cuda.
 
Top
Sign up to the MyBroadband newsletter
X