With all of the recent hoopla surrounding Intel and NVIDIA as of late, today's announcement may come as a bit of a surprise to many of you. To say that there has been some tension between Intel and NVIDIA lately would certainly be putting it mildly. But the fact of the matter is, for now at least, the two companies share a somewhat symbiotic relationship. The products we'll be showing you here are proof of this fact.

What we have on tap for you today is a new NVIDIA chipset with an mGPU (or IGP) designed for the Intel Socket 775 platform, the GeForce 9300. We should note, however, that NVIDIA is also announcing the GeForce 9400 today as well, the same chipset that powers Apple's new MacBook line of products, but the two are nearly identical save for a few minor spec variations. Whether you're a small form-factor type, Home Theater PC buff or one of the newly enamored by Apple's new line of sleek notebooks, read on to find out about these enabling technologies that NVIDIA has launched today in platform chipsets with integrated graphics engines.

GeForce 9300 Block Diagram

The high-level block diagram above does a good job of clearly illustrating the GeForce 9300's (and 9400's) main features. As you can see, the GeForce 9300 supports virtually all flavors of Core 2, Pentium, and Celeron processors with front side bus speeds up to 1333MHz. The chipset is comprised of a single chip and the mGPU features 16 shader cores, with support for DX10 and many of NVIDIA's proprietary technologies, like CUDA, Hybrid SLI, PhysX and PureVideo HD. The GeForce 9300 mGPU has output support for virtually all of the major port types (DVI, DisplayPort, HDMI, etc.) with support for HD audio as well.

The chipset can support DDR2 or DDR3 memory technologies, although the two motherboards we'll be showing you in this article required DDR2. It also sports 12 USB 2.0 ports, gigabit Ethernet, six SATA 3.0Gb/s ports (with RAID), 20 PCI express 2.0 lanes, and support for 5 PCI slots.
 

CPU	Core 2 Family Pentium D Pentium 4 Celeron D Celeron	Core 2 Family Pentium D Pentium 4 Celeron D Celeron
FSB	1333 Mhz	1333 Mhz
Memory Interface	Dual-channel DDR2-800 / DDR3-1333	Dual-channel DDR2-800 / DDR3-1333
DirectX 10 Support	Yes	Yes
Graphics Cores	16	16
Core/Shader Clocks	580/1400 MHz	450/1200 MHz
Texture Fill Rate	3.6 Billion/second	3.6 Billion/second
Max. Anti-Aliasing (AA) Sample Rate	16×	16×
RAMDACs	400 MHz	400 MHz
Max. HDR Precision	128-bit	128-bit
Max. Analog Resolution	2048 × 1536	2048 × 1536
Max Digital Resolution	2560 × 1600	2560 × 1600
GeForce Boost Technology	Yes	Yes
HybridPower Technology	Only available in select designs	Only available in select designs
NVIDIA PureVideo HD	Yes With full HD decode (1080i/p)	Yes With full HD decode (1080i/p)
Display options	RGB, dual-link DVI, HDMI, DP	RGB, dual-link DVI, HDMI, DP
PCI-Express 2.0	20 lanes 1 × 16 4 × 1	20 lanes 1 × 16 4 × 1
SATA drives	6	6
SATA speed	3 Gbps	3 Gbps
RAID	0, 1, 0+1, 5	0, 1, 0+1, 5
NVIDIA MediaShield Storage technology	Yes	Yes
Networking	10/100/1000	10/100/1000
USB ports	12 / 2C	12 / 2C
PCI Slots	5	5
Audio	HDA (Azalia)	HDA (Azalia)
HD Audio Support	7.1 LPCM	7.1 LPCM

In terms of their specifications, the new GeForce 9300 and GeForce 9400 are nearly identical. As you can see in the spec list above, the only differences between the two are that the GeForce 9400 has a higher clocked GPU and shader cores. These attribute would obviously give the GeForce 9400 an edge in the 3D performance department, but as you'll see a little later, even the GeForce 9300 is far more powerful than any of Intel's current IGPs.

For the purpose of this article we have obtained a couple of GeForce 9300-based motherboards from Asus and Zotac. Both of the boards, in full retail trim, are pictured below...

    

    

   
Zotac GeForce 9300

The Zotac GeForce 9300 is a micro-ATX motherboard, built upon a black PCB with color coded connections and headers. As you'd expect from a motherboard of this class, the Zotac GeForce 9300 ships with a basic accessory bundle, which includes a couple of SATA, IDA, and floppy cables, a case bracket with additional Firewire ports, an I/O shield, a user's manual and driver CD.

Right smack in the middle of the motherboard, you can see a small, active cooler mounted atop of the GeForce 9300 chipset. The Zotac GeForce 9300 motherboard features a single PCI Express x1 slot, a single PCI Express x16 slot, and two standard PCI slots.  In addition to all of the features inherent to the chipset (mentioned on the previous page), the Zotac board also features a handy POST code error reporter, Firewire support through the use of VIA controller, and IDE and floppy support through the use of a JMicron controller.  Generally, the layout of the motherboard is good, considering the tight quarters associated with the micro-ATX form factor, and users shouldn't have any problem fitting aftermarket coolers on the board.

The I/O port cluster on the Zotac GeForce 9300 consists of PS/2 mouse and keyboard ports, composite, HDMI, DVI, and VGA outputs, four USB ports, a, RJ45 LAN jack, and six assorted audio jacks.

    

    

    
Asus P5N7A-VM

The Asus P5N7A-VM has much in common with the Zotac GeForce 9300 board above.  They are both built upon dark colored PCBs, they both conform to the micro-ATX form factor, and they both have the same compliment of expansion slots.  Bundled with the Asus P5N7A-VM, we found the expected assortment of accessories, including SATA and IDE cables, a pack of Asus' Q-Connectors, an I/O shield, and of course a user's manual and driver CD.

The layout of the P5N7A-VM is very similar to the Zotac board, except for the placement of its SATA, IDE and ATX connectors.  The positioning of the boards' DIMM slots, expansion headers, supplemental power connection, and various other components are all similar.  The I/O port cluster on the P5N7A-VM though has much more going in.  In the P5N7A-VM's port cluster you'll find a single PS/2 port, six USB ports, analog and digital audio outputs, DisplayPort, HDMI, DVI, and VGA display outputs, and Ethernet port, and an eSATA connector.

Much of NVIDIA's recent marketing has centered around the notion that GPUs and mGPUs aren't just for gaming any longer, which you'd expect considering they've got GPU architectures in their arsenal capable of much more than graphics.  Of course, we've known this to be true for quite some time, but there definitely seems to be more happening in the GPGPU arena as of late. Over the past few weeks and months, NVIDIA has shown off a number of applications that all benefited from the power of a GPU, and none of them are games.

The application you see pictured above is a beta, pre-release version of Elemental Technologies’ BadaBOOM video encoder.  BadaBOOM takes advantage of ETI’s GPU-powered RapiHD Video Platform to offload some video encoding duties from the CPU, onto the GPU, to accelerate the process of converting standard-definition video from any format to H.264 for portable media devices, like an iPod, Zune, or iPhone.

NVIDIA, along with representatives from Stanford University, have also developed a brand new version of the Folding @ Home client which uses the GPU for its calculation. While using the GPU, the F @ H client shows massive speed increase compared to existing CPU architectures.  A typical CPU can do about 4ns / day, a PS3 about 100, a Radeon HD 3870 approximately 170, and GeForce GTX 280 about 500ns / day.

Which of course brings us to PhysX.  If you remember, NVIDIA acquired AGEIA not too long ago and incorporated PhysX support into all CUDA capable (GeForce 8, 9, and GTX 200 series) GPUs and mGPUs.  Since NVIDIA took over AGEIA, there has been much more buzz surrounding PhysX, and word is that many more developers have jumped, or will jump, on board.  Having NVIDIA's marketing muscle behind the PhysX technology has resulted in a number of new developers signing on to use the technology.

How We Configured Our Test Systems: When configuring our test systems for this article, we first entered their respective system BIOSes and set each board to its "Optimized" or "High performance Defaults". We then saved the settings, re-entered the BIOS and set memory timings for either DDR2-1066 (AMD) with 5,5,5,15 timings or DDR3-1333 with 7,7,7,20 timings (Intel). The hard drives were then formatted, and Windows Vista Ultimate was installed. When the Windows installation was complete, we updated the OS, and installed the drivers necessary for our components. Auto-Updating and Windows Defender were then disabled and we installed all of our benchmarking software, defragged the hard drives, and ran all of the tests.

HotHardware's Test Systems  Intel and AMD - Head To Head

System 1: Core 2 Quad Q9650 (3.0GHz - Dual-Core)  Zotac GF 9300 / Asus P5N7A-VM (NVIDIA GeForce 9300 Chipset) 2x1GB Corsair PC2-8500 CL 5-5-5-15 - DDR2-1066 GeForce 9300 IGP On-Board Ethernet On-board Audio WD740 "Raptor" HD 10,000 RPM SATA Windows Vista Ultimate DirectX Redist (June 2008)

System 2: Core 2 Quad Q9650 (3.0GHz - Dual-Core)  Intel G45PID (G45 Express) 2x1GB Corsair DDR3-1800 CL 7-7-7-20 - DDR3-1333 Intel IGP On-Board Ethernet On-board Audio WD740 "Raptor" HD 10,000 RPM SATA Windows Vista Ultimate DirectX Redist (June 2008)

System 3: AMD Phenom X4 9850 (2.5GHz) Asus M3N-HT Deluxe / Gigabyte MA790GP (nForce 780a SLI /  790GX) 2x1GB Corsair PC2-8500 CL 5-5-5-15 - DDR2-1066 GeForce 780a IGP On-Board Ethernet On-board Audio WD740 "Raptor" HD 10,000 RPM SATA Windows Vista Ultimate DirectX Redist (June 2008)

Preliminary Testing with SiSoft SANDRA XII  Synthetic Benchmarks

We began our testing with SiSoftware's SANDRA XII, the System ANalyzer, Diagnostic and Reporting Assistant. We ran three of the built-in subsystem tests that partially comprise the SANDRA XII suite with a Core 2 Quad Q9650 processor installed in the Asus P5N7A-VM motherboard (CPU Arithmetic, Memory Latency, and Memory Bandwidth).  All of the scores reported below were taken with the processors running a clock speed of 3GHz, with 2GB of DDR2 RAM installed.

Core 2 Quad Q9650 @ 3GHz CPU Arithmetic GeForce 9300	Core 2 Quad Q9650 @ 3GHz Memory Bandwidth GeForce 9300	Core 2 Quad Q9650 @ 3GHz Memory Latency GeForce 9300

The GeForce 9300's performance in the three SiSoft SANDRA benchmarks we ran fell right in line with expectations.  The CPU Arithmetic benchmark showed our GeForce 9300-based test system, right on par with the QX9650 based reference system.  The memory bandwidth benchmark showed significantly lower peak bandwidth available on the GF9300, but that is to be expected on a system equipped with an IGP that shares main memory.  And the latency benchmark reported a speed factor of 119.2, which was actually fairly strong versus the reference systems.

We also ran the GeForce 9300-based Asus and Zotac motherboards through Futuremark’s latest system performance metric built especially for Windows Vista, PCMark Vantage. PCMark Vantage runs through a host of different usage scenarios to simulate different types of workloads including High Definition TV and movie playback and manipulation, gaming, image editing and manipulation, music compression, communications, and productivity.  Most of the tests are multi-threaded, so they can exploit the additional resources offered by a quad-core CPU.

Futuremark PCMark Vantage Simulated Application Performance

The Asus and Zotac GeForce 9300-based motherboards we tested performed on roughly the same level--the Asus board finished ahead in a few tests, the Zotac board finished ahead in the others.  In comparison to the Intel G45 motherboard, the GeForce 9300 generally trailed slightly in the application and productivity tests, but in the gaming tests the extra GPU horsepower afforded by the 9300's mGPU allowed NVIDIA's platform to finish way out in front.

In our custom LAME MT MP3 encoding test, we convert a large WAV file to the MP3 format, which is a popular scenario that many end users work with on a day-to-day basis to provide portability and storage of their digital audio content.  LAME is an open-source mid to high bit-rate and VBR (variable bit rate) MP3 audio encoder that is used widely around the world in a multitude of third party applications.

LAME MT Audio Encoding

In this test, we created our own 223MB WAV file (a hallucinogenically-induced Grateful Dead jam) and converted it to the MP3 format using the multi-thread capable LAME MT application in single and multi-thread modes. Processing times are recorded below, listed in seconds. Once again, shorter times equate to better performance.

Out custom LAME MT benchmark was essentially a wash between the GeForce 9300s and Intel G45, although the NVIDIA platform did finish slightly faster in the single-threaded version of the benchmark.

Kribibench v1.1 CPU-Bound 3D Rendering

For this next batch of tests, we ran Kribibench v1.1, a 3D rendering benchmark produced by the folks at Adept Development.  Kribibench is an SSE aware software renderer where a 3D model is rendered and animated by the host CPU and the average frame rate is reported.  We used two of the included models with this benchmark: a "Sponge Explode" model consisting of over 19.2 million polygons and the test suite's "Ultra" model that is comprised of over 16 billion polys.

In the two Kribibench rendering tests we ran, the system built around Intel's G45 finished just slightly faster than the Asus and Zotac GeForce 9300 motherboards.  The difference can be attributed to the G45's use of faster, and hence higher bandwidth, DDR3 RAM.

Cinebench R10 is an OpenGL 3D rendering performance test based on Cinema 4D. Cinema 4D from Maxon is a 3D rendering and animation tool suite used by 3D animation houses and producers like Sony Animation and many others.  It's very demanding of system processor resources and is an excellent gauge of pure computational throughput.

Cinebench R10 3D Rendering

This is a multi-threaded, multi-processor aware benchmark that renders a single 3D scene and tracks the length of the entire process. The time it took each test system to render the entire scene is represented in the graph below, listed in seconds.

Unlike the Kribibench rendering tests on the previous page, the Cinbench R10 benchmark had the GeForce 9300-based motherboards out in front of the Intel G45 by a few points in the multi-threaded test.  In the single-threaded test, however, the G45 was a bit ahead.

Futuremark 3DMark06 Synthetic DirectX Gaming

3DMark06's built-in CPU test is a multi-threaded DirectX gaming metric that's useful for comparing relative performance between similarly equipped systems.  This test consists of two different 3D scenes that are processed with a software renderer that is dependent on the host CPU's performance.  Calculations that are normally reserved for your 3D accelerator are instead sent to the CPU for processing and rendering.  The frame-rate generated in each test is used to determine the final score.

3DMark06's built-in CPU benchmark reported scores for Asus and Zotac GeForce 9300-based motherboards that were just slightly better than the Intel G45.

For our next set of tests, we moved on to some in-game benchmarking with Crysis and F.E.A.R. When testing motherboards or processors with Crysis or F.E.A.R., we drop the resolution to 800x600, and reduce all of the in-game graphical options to their minimum values to isolate CPU and memory performance as much as possible.  However, the in-game effects, which control the level of detail for the games' physics engines and particle systems, are left at their maximum values, since these actually do place some load on the CPU rather than GPU.

Low-Resolution Gaming: Crysis and F.E.A.R. Taking the GPU out of the Equation

Intel's G45 chipset, which is equipped with the GMA X4500HD IGP simply couldn't hold a candle to NVIDIA's offerings in our low-resolution game tests.  Both the Asus and Zotac GeForce 9300-based motherboards were significantly faster than Intel's offering here.

Next, we moved on to some Hybrid SLI testing with 3DMark06 and Crysis.  As we've mentioned, when coupled with a GeForce 8500 GT or 8400 GS, the GeForce 9300 can utilize NVIDIA's GeForce Boost technology, for increased 3D performance.  We tested the GeForce 9300 GT with 3DMark06 and Crysis using the built-in IGP, with a GeForce 8500 GT, and with the IGP and 8500 GT working in tandem in GeForce Boost mode.

Hybrid SLI Tests: 3DMark06 and Crysis Doubling Up On The GPUs

As you can see, when using Hybrid SLI with the GeForce 9300 / 8500 GT combination working in tandem in GeForce Boost mode, the platform offers significantly better 3D performance.  We saw increases of 69% and 87% in Crysis and 3DMark06, respectively, when running in GeForce Boost mode.

We also did some quick testing of the GeForce 9300's video processing engine as it is implemented on the Asus P5N7A-VM, in terms of both image quality and CPU utilization with some HQV and H.264 playback tests.

Video Playback Performance: SD and HD HQV and H.264

HQV is comprised of a sampling of SD video clips and test patterns that have been specifically designed to evaluate a variety of interlaced video signal processing tasks, including decoding, de-interlacing, motion correction, noise reduction, film cadence detection, and detail enhancement. As each clip is played, the viewer is required to "score" the image based on a predetermined set of criteria. The numbers listed below are the sum of the scores for each section. We played the HQV DVD using the latest version of Cyberlink's PowerDVD Ultra, with hardware acceleration for NVIDIA PureVideo, ATI UVD, and Intel video acceleration extensions enabled.

The GeForce 9300 performed much like NVIDIA's discreet graphics cards in the HQV benchmark, clearly surpassing both Intel's and ATI's offerings here. We should note, however, that in the Jaggies 1 and Jaggies 2 tests, the GeForce 9300 just barely squeaked by with its 5 and 3 scores, but due to HQV scoring system the scores are justified.  If we disregard the scores, in a side-by-side test versus a discreet card in the GeForce 8500 range or higher, the discreet card actually performs better in those tests.

GeForce 9300 H.264 HD Playback

Next we conducted a test using an H.264 encoded movie clip which is available for download from NASA's HD showcase website. The CPU utilization data gathered during this tests was taken from Windows Vista's built-in Performance Monitor while playing back on the GeForce 9300 using PowerDVD 8 Ultra. The graphs show the CPU utilization while playing back the 1080i QuickTime clip on the GeForce 9300. As you can see, utilization was quite low and HD video playback shouldn't be a problem for this chipset.

Before we bring this article to a close, we'd like to cover a few final data points. Throughout all of our benchmarking and testing, we monitored how much power our test systems were consuming using a power meter. Our goal was to give you all an idea as to how much power each configuration used while idling and under a heavy workload. Please keep in mind that we were testing total system power consumption at the outlet here, not just the power being drawn by the motherboards alone.

Total System Power Consumption Tested at the Outlet

The GeForce 9300 proved to be quite power friendly in our testing.  While idling, the Intel G45 and GeForce 9300-based motherboards all consumed similar amounts of power, with the G45 board falling in between the Asus and Zotac boards (the Zotac board's active chipset cooler accounts for the slight increase over the G45 at idle).  Under load though, the GeForce 9300s both consumed less power than the G45--although the difference was only a few watts.  It is worth noting, however, that the GeForce 9300 offers lower power consumption with significantly increased GPU performance, so in terms of efficiency there is really no match here; NVIDIA wins.

Performance Summary: There are number of performance comparisons we need to make while summarizing the benchmarks scores from this article.  If we compare the two GeForce 9300-based motherboards we tested here to each other, we find that the Asus P5N7A-VM and Zotac GeForce 9300 perform about on par with one another; they trade victories depending on the benchmark, but the deltas are quite small.  In comparison to Intel's G45 chipset, the GeForce 9300 generally performs on roughly the same level in productivity applications, with a slight edge going to the G45 where its increased memory bandwidth comes into play.  In any GPU intensive situation, however, the GeForce 9300 is simply on a different level than the G45.  In the game tests, for example, the GeForce 9300 was up to 4.4X faster than the G45.

NVIDIA has a solid offering on their hands with the new GeForce 9300 / 9400.  We can definitely see this chipset being popular with the HTPC crowd.  If you wanted an Intel-based HTPC built around a current Intel CPU, up until now, that usually meant settling for an Intel IGP or an older GeForce 7-series mGPU, which couldn't offer the same kind of 3D or video playback performance.  Although this is going to sound cliche', with the GeForce 9300 / 9400, HTPC afficionadoes no longer have to make any comprimises. The GeForce 9300 and 9400 represent full-featured, low-powered, core logic chipsets, with competitive application performance and the best integrated graphics processor available for the Intel platform.  Video playback performance is great, and the Intel platform finally has an IGP that can play many current games with acceptable framerates.  Even if you're not an HTPC afficionado, we suspect many of you can see the value in that type of product, which is probably why these chipsets landed in Apple's brand new Macbook offerings.