"Clocking down" is not technically correct as it is the number of lanes that are reduced, not the clock speed. An excerpt from
Anandtech explains this further, but it's a bit cryptic:
"NVIDIAs first task in assuring that the load distributed to both GPUs would be balanced and symmetrical was to equip their nForce4 SLI chipset with identical width PCI Express graphics slots. By default, PCI Express graphics cards use a x16 slot, which features 16 PCI Express lanes offering 8GB/s of total bandwidth. Instead of outfitting their chipsets with 16 more PCI Express lanes, NVIDIA simply allows the number of lanes to be reconfigurable to either a single x16 slot or two x8 slots, with the use of a little card on the motherboard itself.
The physical slots themselves are both x16 slots, but electrically they can be configured to be two x8 slots. This wont cause any compatibility issues with x16 cards, as they will just use fewer lanes for data transfers, and the real world performance impact is negligible in games, which is what NVIDIA is counting on."
I think what this means is that PCIe graphics cards use 16 lanes by default, but when SLI is enabled, the SLI Chipset only uses 8 lanes per card. The real world impact of this is negligible, but the graphics cards themselves must still plug into full-size 16-lane (x16) slots.