Antec TrueBlue 2.0 480W Power Supply

Split 12V Rails & SLI:

Here's some interesting points to ponder (this is all just my thoughts on the matter). For most PCs split 12V rails is good (for the reasons mentioned on the previous page), however with the advent of SLI (dual video cards) 12V power consumption made a noticeable jump. If you consider the peak power consumption from a pair of top-of-the-line video cards can reach 12-16A, that doesn't leave much power (if any) for the rest of the devices on that rail. UL/EN spec says a maximum of 240VA (translates to 20A on each 12V rail, or even more stringent is 18A per ATX12V spec).

So what is a power supply company supposed to do (for both short-term and long-term)? The reality is unfortunately the UL/EN safety spec will be in direct conflict with what can be done with respect to a power limit of 240VA per rail. More often than not, a PSU manufacturer will simply join the two rails together on the supply-end to maximize available 12V power.

The SLI market is relatively small compared to the entire PC market, take a subset of that number for users that have the top-end video cards, and then divide up that small percentage among all the various PSU makers. The end number that would be roughly each PSU manufacturer's share of the market is extremely small when compared to the mass-market of more "traditional" requirements. So, creating a special power supply just to cater to this nice market really isn't cost-feasible (at this time).

Until some new official guidelines are published, and there is a clear direction that both video card & motherboard manufacturers are going to take, I don't think we are going to see any huge change in power supply design. The current route of having a combine 12V rail seems to be the more cost-effective solution than something like a 3rd 12V rail dedicated for video.

How We Test Power Supplies:

In order to thoroughly test a power supply we have built a special rig that is capable of dynamic load levels on each rail. A short article covering our power supply testing methodology in detail can be found here. Please read over the guide if you haven't already, you might just learn a thing or two. Simply plugging a power supply into a regular PC for "testing" is not accurate by any means, nor can any useful numbers be extracted by testing that way.

Also worth reading is our power supply primer article titled, Powering Your PC: Truths and Misconceptions - A Guide To Power Supplies, which covers how power supplies work in great detail.

If you have read the power supply testing article, you know that we publish four (4) groups of numbers for each power supply to give a simplified overall representation that can be compared to other power supplies. Many other load ranges are tested (but not published unless there is a serious problem) to ensure voltage levels remain within ATX12V specifications. I like to refer to it as "working the switchboard" since it involves flipping lots of switches tied to resistors on each rail to achieve various loads. Unfortunately we can't always dial in exact maximum numbers because of the fixed levels of resistance, but load levels we test are often far in excess of any load a regular PC will create.