Summary: After years of providing chipsets for AMD processors only, today NVIDIA officially unveils their first chipset for the Intel Pentium platform: the nForce4 SLI Intel Edition. As its name implies, this new chipset is based on the same fundamental technologies found in today's nForce4 SLI chipset for the Athlon 64, including GigE with firewall, Serial ATA II support, 7.1-channel audio, 10 USB ports, and of course, 20 lanes of PCI Express for SLI! But how does this new chipset perform? Find out in today's article!
![]() Every time the question was asked, NVIDIA executives always gave the politically correct answer of what a wonderful partner AMD was and how they were happy to be in the AMD ecosystem. Early on, NVIDIA was also quick to admit their inexperience at designing, and most importantly, producing chipsets in massive quantities; this was a problem that nagged them when the original nForce chipset, and its follow-up part, nForce2 when they were first introduced. As far as NVIDIA was concerned, the AMD market was keeping them busy enough already, or at least that’s what they told us officially. Today, after selling millions of nForce chipsets to the AMD market, it’s hard to argue against NVIDIA’s strategy. [image]
Of course, another key reason why NVIDIA hasn’t produced an Intel chipset to date has nothing to do with chipset production or NVIDIA’s cozy relationship with AMD; the key hang-up between the two companies has been Intel’s asking price for a Pentium 4 bus license. Without a license for Intel’s front-side bus, NVIDIA couldn’t produce a Pentium 4 chipset without getting into legal trouble with Intel. In contrast to AMD, who licenses their bus for free (including HyperTransport) Intel enjoys royalties from the sale of any chipset that uses their bus to communicate to the P4 processor. And if you don’t own a license, Intel will vigorously defend their IP, just ask VIA. (NVIDIA gets around this issue on the Xbox because Microsoft holds a P3 bus license.) Then, in November of last year, shortly after the first batch of nForce4 SLI systems and motherboards appeared, NVIDIA and Intel announced a broad cross-license agreement. According to the PR the companies signed a “multi-year patent cross-license agreement spanning multiple product lines and product generations. Additionally, the companies signed a multi-year chipset agreement for NVIDIA to license Intel’s front-side bus technology. This will enable NVIDIA to deliver the NVIDIA nForce platform technology on Intel-based systems.” In other words, as a result of the agreement, NVIDIA finally had the green light to produce an Intel-based chipset. The product we’re taking a look at today, NVIDIA’s nForce4 SLI Intel Edition is the first product out of the gates from NVIDIA for the Intel market. With SLI support, it’s clearly targeted at one market: high-end desktop PCs.
Traditional North Bridge/South Bridge architecture
When AMD integrated the memory controller on the Athlon 64 processor’s core itself, a huge amount of real estate was freed up within the chipset. NVIDIA used this space to integrate all the functions found in the nForce3/nForce4 chipsets into a single chip. By moving to a single chip design, latency is reduced. Single-chip designs also free up more space on the board for motherboard manufacturers.
nForce4 SLI Intel Edition SPP
With dedicated chips for the North Bridge and South Bridge, NVIDIA has dusted off an old acronym from the nForce/nForce2 days for the North Bridge of the chipset, the SPP, otherwise known as the system platform processor.
Third generation DASP
First introduced in NVIDIA’s original nForce chipset, the Dynamic Adaptive Speculative Pre-Processor (DASP) acts as a data prefetch unit for the nForce4 SPP itself. If you’re not familiar with data prefetching, the concept is simple: the DASP intelligently looks for regular access patterns in memory access, predicts which data will be necessary next, and fetches and places that data inside its cache before it's actually needed. Once the CPU requests the data, it is available for the processor immediately, reducing system latency dramatically.
QuickSync
Besides the memory controller, the other main feature of the nForce4 SLI SPP is its PCI Express subsystem. Like the nForce4 SLI chipset for the AMD platform, NVIDIA provides twenty lanes of PCI Express for the Intel Edition of the chipset. Sixteen of these lanes are dedicated to the graphics processor, with three lanes reserved for the x1 PCI-E slots. Just like the original nForce4 SLI, when running in SLI mode on the Intel Edition chipset, the two x16 graphics slots operate as two x8 slots for optimal performance. The uppermost card is designated as the “Master” card, while the graphics card beneath it is slaved to it. In testing, NVIDIA found this configuration yielded the best performance. Of course, if the second card is removed, or if you don’t have two SLI-ready NVIDIA graphics cards, the chipset will devote all sixteen PCI Express lanes to the primary PCI-E slot. Processor support/FSB
nForce4 SLI Intel Edition supports all of the latest Intel processors, including Intel’s dual-core “Smithfield” chips, such as the Pentium Extreme Edition 840, which was first introduced on Monday. Front-side bus speeds up to 1,066MHz are also fully supported by the chipset.
[image]
Paired alongside the SPP is NVIDIA’s nForce4 SLI Intel Edition Media Communications Processor, otherwise known as the MCP. The MCP handles the traditional I/O and storage duties of a conventional South Bridge with aplomb, sporting such features as support for Serial ATA II (3Gb/sec), GigE with a built-in hardware-based Firewall, 7.1 audio, and more. Ironically enough, the pathway between the MCP and the SPP is none other than AMD’s own HyperTransport link. Storage
As we just mentioned, the nForce4 SLI Intel Edition features robust storage support. The MCP utilizes dual controllers providing support for up to four Serial ATA hard drives (or four conventional IDE drives) and also supports native command queuing for improved disk performance. Networking and audio
Like Intel’s chipsets, NVIDIA’s MCP features Gigabit Ethernet support. A dedicated 2Gbps bi-directional link is used, maximizing the full potential of the controller. NVIDIA’s familiar hardware-based firewall (ActiveArmor) is also present, protecting your PC from spyware and hackers.
According to NVIDIA, the first nForce4 SLI Intel Edition motherboards should hit retail sometime around the end of this month, with retail prices hovering in the $200 range, just as you saw with AMD-based SLI motherboards when they first launched. NVIDIA expects that as additional partners release their boards, prices will slowly drop to the $160-$190 price range most AMD-based SLI motherboards sell for today. Memory
Launching alongside the new chipset and motherboards is NVIDIA’s new validation program for DDR2 memory manufacturers who have been tested and validated on the platform for speeds of up to 667MHz. (Specifically, compliant modules must surpass commodity JEDEC memory requirements and comply with minimal targeted performance levels, including 667MHz clock speeds with memory timings of 4-4-4-12-2T.) Power
In terms of power required, we were told that NVIDIA’s general recommendation for the nForce4 SLI Intel Edition, as well as the AMD-based SLI platform, was a 550-watt power supply. In particular, the new nForce4 SLI Intel SPP and MCP consume about twice as much power as the CK8-04 chipset used in AMD platforms, 30-watts in the Intel chipset versus 15 in CK8-04. You’ll also need to take into account the higher power consumption of Intel processors. For instance, the 3.73GHz Pentium 4 Extreme Edition consumes about 119 watts. In comparison, the Athlon 64 FX-55 dissipates 105W. Setup
If you’re familiar with the procedure for setting up SLI on NVIDIA’s AMD platform, the Intel SLI chipset will be pretty familiar. Simply orient the SLI selector card into the motherboard for dual GPU configuration, plug in two SLI-ready NVIDIA graphics cards (GPUs supported include the GeForce 6600 GT, 6800, 6800 GT, and 6800 Ultra), and connect them with the SLI connector that shipped with your motherboard.
With each new release, NVIDIA’s nTune software continues to get better and better. Originally known as the System Utility software, nTune provides the traditional hardware monitoring capability you’d expect from a third-party application, monitoring such critical aspects as CPU and system temperatures, fan speeds, and voltages, but goes many steps beyond that by also providing built-in tools for overclocking the front side bus and memory bus, tweaking memory timings, and even overclocking your graphics card! [image]
That’s right, if you have a newer GeForce FX or better graphics card installed inside your system, nTune can be used to automatically overclock your graphics card without having to resort to using the Coolbits registry hack, or other third-party software applications. With nTune’s profiles, you can automatically tune your system specifically for the best memory performance, or if you deal with large databases, the system can be optimized for best disk performance. Gamers will of course select the “Best Graphics Performance” setting, while those of you who like to watch movies on your PC will want to select “Silent Tuning”. [image]
One new feature NVIDIA has added to nTune recently is dynamic application sensing. With this feature, nTune automatically recognizes what program you’re loading up, then loads the appropriate profile. Say for instance you load up your DVD playback software, nTune can detect PowerDVD loading up and automatically load the profile for silent settings. Then, once you launch a game, nTune can sense that, cranking up your system settings for maximum performance. [image]
For manual overclocking, nTune provides settings for adjusting your graphics card speed, the chipset itself, memory timings, CPU and memory voltages, and fan settings, saving the end user lots of time that would otherwise be spent fiddling in BIOS. Once it’s done optimizing your system, nTune can then benchmark your system based on graphics, disk, and memory performance, where you can then compare your performance both before and after the modifications, as well as compare your system to NVIDIA’s baseline configuration, which is based on a similar system configuration. [image]
The real beauty of nTune though is its size. For an application that performs so many functions, you’d expect it to take up a huge amount of system resources. Fortunately, this isn’t the case at all, as we’ve found that nTune only takes up a small memory footprint. Of course, the final nTune implementation depends on your particular motherboard manufacturer. If a manufacturer wishes, they can disable certain features to prevent end user’s from damaging their hardware (or some features just may not work properly, for instance fan speed control), or they may skip support for nTune altogether. Since this is a product targeted at enthusiasts though, hopefully all nForce4 SLI Intel Edition motherboard manufacturers will get onboard and provide full support for nTune in their retail nForce4 products.
System Setup
Benchmarks
Lock On: Modern Air Combat (Mig-29 custom demo)
Lock On: Modern Air Combat – Direct3D
IL-2 Sturmovik: FB - OpenGL
Pacific Fighters - OpenGL
Far Cry – Direct3D
Far Cry – Direct3D
DOOM 3 – OpenGL
DOOM 3 – OpenGL
Half-Life 2 – Direct3D
Chronicles of Riddick
Splinter Cell: Chaos Theory – Direct3D
![]() ![]() ![]() ![]() ![]()
![]() ![]()
The Intel Edition of the nForce4 SLI chipset incorporates all the major features found on its AMD counterpart, including GigE with native Firewall, 10 USB ports, NVIDIA’s impressive storage subsystem, and even AMD HyperTransport, which links the North Bridge and the South Bridge together. The only difference is that NVIDIA has adapted the chipset for Intel’s Pentium processors, this includes adding a new memory controller with support of the latest DDR2-667 memory, and of course Intel’s 1066MHz FSB. From a features perspective, NVIDIA’s nForce4 SLI Intel Edition chipset is in a unique class of its own, even excluding SLI support. But of course, no gamer willing to fork over $200 for a motherboard and another $400 or more for two graphics cards will forget about SLI. We witnessed performance gains that were comparable to the improvements seen on NVIDIA’s SLI platform for AMD users, sometimes in the order of just over 1.7X at 1600x1200 with 4xAA and 16xAF, but there were even multiple cases where we were pushing a 2X performance improvement under the same settings! Based on these kinds of results, clearly NVIDIA’s driver team has implemented quite a few performance enhancements inside their latest ForceWare release for SLI. At the same time however, there’s still a lot of work to be done. NVIDIA currently boasts SLI support for over games, but this is a small selection of the overall gaming market. In addition, some of the titles on the existing list aren’t quite up to snuff. Chronicles of Riddick is a perfect example of this, with SLI enabled, stability is severely compromised, and when you do get numbers, they’re below a single card configuration. We also ran into performance problems with the GeForce 6600 GT running in SLI mode in Splinter Cell: Chaos Theory. Performance would consistently begin to hitch in the same area of the timedemo, so we’re pretty sure the problem wasn’t overheating. Speaking of overheating, this is one aspect you’ll definitely have to take into account when building an nForce4 SLI Intel Edition system, especially if you plan on outfitting your system with a fast processor. Under load with our Pentium 4 3.73GHz Extreme Edition CPU, the GeForce 6800 GT cards running in SLI mode would begin to overheat when running looped demos in Far Cry for more than 10 minutes, causing the system to crash and a full reboot. We ultimately rectified the problem by removing the case cover our system shipped in (a CoolerMaster WaveMaster chassis with an NVIDIA case window on one side) so we could install an additional case fan to act as a blowhole, blowing cool air directly onto the graphics cards. The problem is caused by inadequate airflow. NVIDIA’s cards feature ducted cooling designs. These coolers work great when they have a steady supply of fresh air, but with the secondary “slave” graphics card in the way, airflow to the primary “master” graphics card is constrained – it’s literally sucking up the hot air off the card below it! As a result, the primary card typically operates 5-10 degrees Celsius higher than the secondary card. Teething problems aside though, NVIDIA’s SLI platform processors have brought quite a bit of excitement to the normally mundane chipset world. Obviously Intel recognizes that as well, why else would they sign NVIDIA so quickly after years of dismissing them previously? If you’re a hardcore gamer looking to get the most performance out of the Intel platform as possible, a motherboard based on NVIDIA’s nForce4 SLI Intel Edition chipset should be at the top of your list of components to purchase. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| © Copyright 2003 FS Media, Inc. |