NVIDIA (with recap)
A Recap From HotHardware's Article Section: NVIDIA is the sort of company that can and will capitalize on opportunities quickly, whether it be market or product related, or something as simple as logistics. As such, in traditional fashion, with the huge insurgance of press and media swooping down on the never-ending sensory overload that is the city of Las Vegas and the CES show, NVIDIA took the opportunity to arrange an editor's day just prior to the opening day of CES. With a venue like this and a captive audience to boot it's a proverbial no-brainer and so we convened at the New York-New York hotel and casino for an update from virtually every product group at NVIDIA.
If you've been spending some time here at HotHardware, you'll recall our coverage of AMD's Hybrid CrossFire technology for the mainstream desktop segment and the benefits of combining on-board IGP graphics with a discrete card. Of course NVIDIA also has an answer for this, as you'd expect and we'll step through the salient points and highlights of Hybrid SLI for you in the pages ahead.
In addition, the company is also announcing the new nForce 700a series of chipsets for the AMD platform. It too brings with it an offering of new features and functionality that NVIDIA is disclosing today and we learned a bit about some of NVIDIA's patented PCI Express Switching IP in the process as well.
Like AMD's Hybrid CrossFire, NVIDIA's Hybrid SLI technology gives users the abilty to pair an IGP, or mGPU (motherboard GPU), with a discreet graphics card, or cards, for two new mdoes of operation. NVIDIA calls these new modes Hybrid Power and GeForce Boost.
GeForce Boost does what its name suggests. By coupling the motherboard's integrated GPU with a discreet graphics card, the 3D rendering workload is shared between the two GPUs for a boost in performance. Currently, GeForce Boost is only supported by GeForce 8400 GS and 8500 GT discreet cards, as for the IGP, only the next-gen nForce 700a series of chipsets due to arrive this quarter will be supported initially. New IGPs for the Intel platform that support Hybrid SLI are slated to arrive in Q2.
Hybrid Power mode is also fairly self explanatory. With a monitor connected to the mGPU, the discreet graphics card can be completely shut down to reduce power consumption when no in use. A side effect of the reduced power consumption will also be reduced heat and potentially less noise output from the system. The discreet card is shutdown (or renabled) using the SMBUS connection inherent to the PCI Express graphics slots. We should also note, that Hybrid SLI is supporteb by both PCI Express Gen 1 and Gen2.
NVIDIA's Hybrid SLI technology is controlled by system status and the Window Vista Power Plan. Hybrid Power, for example, is enabled when the system reaches an idle state to reduce power consumption. We spoke with NVIDIA and asked if an auto-switching system based on applications or workload for GeForce Boost and Hybrid Power was in the works, and company representatives explained that they would like to enable this type of functionality, but that it wouldn't be ready upon initial release.
At first, a new control panel icon in the system tray will give user's the ability to configure Hybrid SLI for other modes of operation. Because a Hybrid SLI system will have multiple GPUs installed, it's also possible to run multiple displays. The new control panel icon and final set of tools will be available in an upcoming driver release.
As we've mentioned, a key component to Hybrid SLI technology is a core logic chipset with a compatible mGPU. To that end, NVIDIA has developed a line of nForce chipsets for the AMD platform that feature an integrated graphics processor.
The flagship model in the nForce 700a series of chipsets is the 780a SLI. The 780a SLI features two video outputs (one digital, one analog) and an mGPU with similar functionality to a GeForce 8400 series discreet graphics card. This mGPU is Windows Vista Premium certified and DX10 capable. It also features NVIDIA's PureVideo HD engine for full CPU offload of all HD video codecs. The 700a series of MCPs (norhtbridge), will be manufatured at 65nm, and are mostly single-chip designs.
As you can see in this high-level block diagram, however, the 780a SLI uses NVIDIA's NF200 chip for 32 lanes of PCI Express Gen 2 and 3-way SLI support. The MCP also features an additional three lanes of PCI Express Gen 1 connectivity, a single GigeE network conroller, 12 USB 2.0 ports, Azalia HD audio, up to five PCI slots, and Media Shield and ESA support.
As we noted earlier, with this product launch NVIDIA has also disclosed details regarding NVIDIA patented technology that is built into their NF200 PCI Express Switch chip. Essentially, what NVIDIA has done is build a couple of fast paths inside the switch device, dedicated to optimizing mulit-GPU SLI transaction performance both back to the root CPU complex and peer-to-peer between GPUs. This is critical because there is a single Gen2 X16 PCIe link between the NF200 chip and the 780a MCP.
Specifically, there are two functional blocks as you'll not in the above diagram, denoted as "Broadcast" and "PWShort". The Broadcast block specifically provides a broadcast send mode for the root complex down to all GPUs in the system. This allows efficient transfer of data in one group transaction. PWShort, (which stands for Posted Write Short), is a dedicated cut-through mode for peer-to-peer commuinications between the GPUs, without the need to tap on upstream bandwidth to the CPU complex.