Related articles

• TH: As I understand it, Adobe is OpenGL-accelerated. I’ve heard nothing from them about Stream. Yet your initial launch info shows Acrobat Reader, Photoshop CS4, After Effects CS4, and Flash 10 all being “Stream-enabled titles.” Can you elaborate on this? AMD: As I mentioned previously, ATI Stream refers to a framework (or environment) for both hardware and software that provides acceleration of tasks outside the usual game rendering or regular video playback acceleration. In the case of the Adobe apps you mentioned, they are enabling the hardware capabilities to accelerate their processing workflow. TH: Hardware acceleration, yes. But that’s not the same thing as Stream acceleration. AMD: Anything that uses GPU acceleration is ATI Stream, be it DirectX, OpenGL, or ATI Stream component. TH: Anything? Isn’t that getting a little too broad and vague? AMD: It is critical for a parallel compute language to be closely integrated with a graphics API. OpenCL will integrate tightly with OpenGL. Another example is DirectX 11, where the compute shaders are closely coupled with the rendering pipeline. In both cases, the rendering API is complimentary to the compute language in either OpenCL or DirectX 11. As parallel compute gets more standardized (with the ratification and release of OpenCL and DirectX Compute Shaders) “ATI Stream” is not defined by what underlying mechanisms are specifically tied to it, but more by what types of activities are being accelerated by the GPU. We tend to look at Stream as the GPU accelerating workloads that are beyond traditional 3D rendering that has previously been associated with GPU processing. Some productivity applications already use more traditional 3D API’s acceleration of certain tasks, hence we view this as an example of Stream processing.           

• It’d be easy to set this platform up on an open-air test bench, marvel at its size, rattle off its specs, run a couple of benchmarks that show it to be better than Intel’s own Atom-based boards, and give it an award. There’s nothing wrong with open-air testing. There’s nothing wrong with specs, benchmarks, or awards, for that matter. But you’re not going to get the feel of this diminutive little box unless you use it. So, with a HTPC usage model in mind and the issues first encountered in our review addressed, I built up a mini-ITX system around Zotac’s Ion offering, using Windows 7, 2 GB of DDR2 memory, an OCZ SSD, and a slim Blu-ray player that Nvidia was kind enough to send along. Let’s start with DVD playback. This is one area where the Ion is destined to shine given its GPU-based hardware capabilities. I played through several Blu-ray titles with Dolby TrueHD and DTS-HD Master Audio, and they were delivered smoothly. Though my modest theater room isn’t large enough for 7.1-channel sound to make sense, PowerDVD 9 had no trouble decoding the latest lossless codecs and outputting a 5.1-channel LPCM stream to my Onkyo SR507. All in all, as a platform for playing movies, Ion handles business—though once you’re done buying storage, memory, and a chassis, I almost wonder if it wouldn’t be easier to just grab that Playstation instead.  Next up is TV programming. I don’t watch any cable television, but I do hit Hulu on occasion when I eat dinner in front of the PC. The site employs Adobe Flash 9, which is going to need to be processed on the dual-core Atom processor. And indeed, this becomes a problem. In watching the pilot for Glee, it simply wasn’t possible to go full-screen at either low-res (320p) or high-res (480p) on my 50” Samsung running at 1280x720 (in the words of my wife, “It’s too distracting to try watching like this Let me know when you’re done playing around”). The standard browser window mode worked fine, but from 10 feet away, you’re really defeating the purpose of pulling this platform into a home theater environment. Of course, I didn’t expect .mp3 playback to be an issue, so I tried to give the Ion setup a slightly more challenging task constrained by I/O performance instead: streaming audio wirelessly from a NAS attached to the network via Gigabit Ethernet, while indexing a folder in the background. This is something that happens almost-transparently on most desktops; however, it brought this HTPC build to a stuttering stand-still (even with an SSD). Purpose-Built HTPC, Perhaps Fortunately for me, movies are all I really do in the theater room. For that purpose, Ion works. And it’s hard to argue against the board’s form factor (though, to that end, I’m still a fan of the LGA-775 mini-ITX alternative). Keep your expectations in line on this one, though. Any time you’re not taking advantage of the GeForce 9300’s GPU acceleration, you’re at the mercy of Intel’s Atom CPU. On the desktop, that’s really a recipe for frustration.

• There’s not much to say about the Intel Core 2 Duo other than the fact that the switch to Intel was the best move Apple ever made. The combination of Intel’s Haifa design team and international manufacturing prowess has helped produce some of the best CPUs in the history of computing. The Penryn core found in the Core 2 Duo is made possible with the use of 45nm manufacturing technology and the use of hafnium rather than silicon. While Dell allows entry-level laptops to ship with Pentium Dual Core CPUs or older generation Merom-based Core 2 Duo processors, Apple sticks with the Penryn architecture for their entire product line, relying on the 3MB L2 cache models for the MacBook and giving the option of 6MB L2 for the MacBook Pro. Our entry-level MacBook features an Intel Core 2 Duo P7350 running at 2GHz. GeForce 9400M The motherboard is considerably more exciting as it is the first implementation of Nvidia’s GeForce 9400M G, which integrates the GeForce 9400M and MCP71 chipset into a single package. Like any modern chipset, the GeForce 9400M has a full complement of support for SATA 3 Gb/s, Gigabit Ethernet, and dual-channel DDR-3 support. Power consumption of the GeForce 9400M line is on par with Intel’s Centrino 2 architecture at idle, an impressive feat in itself given Intel’s renowned expertise in chipset design. Under heavy load, the GeForce 9400M actually consumes less power than the Intel G45 chipset and from a pure chipset perspective, the GeForce 9400M is a solid choice over the Intel design. The real magic happens with the graphics. While gaming enthusiasts will still want to invest in the faster GPU than what the MacBook Pro includes, the GeForce 9400M is a veritable GPU, adding considerable benefits to the system. From a pure gaming standpoint, the GPU is about half the performance of the GeForce 9600 GT. In practice, traditional GPU-intensive games such as Call of Duty 4 will run at ~30 frames per second at resolutions of 1024x768 at high quality settings. With the exception of games such as Crysis, it appears that the 9400M is capable of running most modern games at medium or high image quality settings at 30 fps or greater, making it perfectly suitable for casual gaming. For more information on the performance of Nvidia's desktop GeForce 9300 chipset, check out that review.  Importantly, the GeForce 9400M has the same quality capabilities as its faster discrete counterparts. It’s just slower. That means that the GeForce 9400M is perfectly suited for the GPU accelerated capabilities of Adobe Photoshop CS4, including anti-aliased text, smooth zoom, rotation, and brush resizing. The integrated Intel X4500, in contrast, lacks all of the OpenGL features required for the full complement of GPU accelerated features in Adobe Photoshop CS4, such as rotation. In the Windows environment, GPU-accelerated H.264 encoding for the iPhone target resolution is about 50% faster than a quad-core CPU. The GeForce 9400M also features Nvidia’s third-generation PureVideo technology. This means full HD Codec offload support for H.264, VC-1, and MPEG-2 including support for Blu-ray’s Picture-in-Picture capabilities as well as Nvidia's excellent image quality and advanced cadence detection and superb region-based deinterlacing. When Mac OS X 10.6 Snow Leopard is released, we anticipate seeing more Mac applications taking advantage of the computing-on-GPU capabilities.