Originally posted by: FastEddie
"Futureproof?" Dual core processors is just a matter of a couple months away. That's not asking for much. A known is that they will work using a single rail psu with enough amperage on the 12v rail.
Let me rephrase what I meant about dual cores & the future. I'm aware they're coming out soon but you never specified how far into the future you meant. Do you think a 110W/9.2A (max) s939 dual core Athlon 64 will have the same power envelope that a dual core processor on socket M2 would? I can't speak for Intel's dual core but as you've said, it's their PSU spec ...assuming Intel dictated the ATX12V 2.0 spec for their own benefits to begin with.
Now it could work on a single rail PSU but as previously quoted, I wouldn't wanna risk the life of the PSU & possibly hardware connected to it.
Originally posted by: FastEddie
And you make it sound like the dual rail psu is the next best thing to come down the pipe since sliced bread.
If all it took was a quote from THG to make it sound like the next biggest thing, I don't think I'm doing multi-rail PSUs justice in all fairness. I could've went further into the merits of dual/multi-rails if that was my intention.
Originally posted by: FastEddie
My point is that the technology has been here awile, and was designed for a totally different animal than what we have with NF4/Sli.
That's your conclusion, that argument will be in complete reverse once SLI's power needs (assuming it flourishes) exceeds the power output capacity that single rail PSUs can deliver. Meaning I don't think either, single or multi rail, was designed for SLI or vice versa. I surmise it's similar to saying there's the Northwood & Prescott, & an application comes along which happens to favor Northwood at lower CPU frequencies & Prescott at higher CPU frequencies - minus the heat problems. Would you say that application was designed for either of them, no. It's just the circumstance at the time.
Originally posted by: FastEddie
The dual rail tech evolved around the Intel architecture, not AMD.
If dual rails have to start somewhere, why not Intel to start it up? They are afterall the market leader & anyone that wasn't sleeping underneath a rock would see the trend that a lot of things are going dual on the desktop e.g. CPUs, memory, GPUs, hard drives (RAID 0), etc.
Originally posted by: FastEddie
But when incorporated into AMD, it was tweaked for multiprocessor/multi-memory bank K7 architecture---not the current K8, which moved most of the current demand to the 12v rail.
I'm lost here, how was the PSU design "tweaked"? AFAIK some dual rail PSUs are made for the P4/K8 and disregard the K7 platform & then there are dual rail PSUs that have no problems powering the K7 platform with both rails. I'll explain further what I mean, the Antec Neopower has one rail dedicated to the P4 connector alone, now since the K7 platform doesn't receive +12V power via the P4 4-pin connector, that rail is pretty much unused (shame on Antec in this case). On the other hand the Enermax unit is designed to power the P4 connector & main 20/24-pin connector via the same rail, therefore both rails are utilized on a K7 based system.
Note that I don't see it being a big factor since dual rail PSUs for the desktop came out at or a little after the A64 and by that time both companies [Intel & AMD] utilized the +12V P4 connector on their latest platforms. Granted there are cases where someone who needs a new PSU cause their old PSU crapped out on their K7 based system may shop for a dual rail PSU, which is unfortunate if they get a Neopower. Which brings the question, does Antec recommend you not use the Neopower on a K7 based system on their packaging? If not supposedly they think 18A was strong enough for any K7 based system?
Originally posted by: FastEddie
That particular version of the Enermax hasn't changed the internal components or ratings to take K8 into consideration, let alone the lopsided demands of Sli. So to mearly add a couple of pci-e power connectors and claim that it's the holy grail for NF4/Sli is really stretching it's capabilities to fulfill that demand.
It powers K8 & SLI just fine until I see otherwise. Again, got proof the Enermax unit can't? Cause if wanted I can pull evidence that it does indeed work with K8 & SLI.
Originally posted by: FastEddie
The key here is Older ATX (1.0) power supplies..... The statement is true for that qualifier---older. Not all ATX 2.0 psu's have two 12v lines or seperate rails
I suppose none of us here in ATF are guilty of having said ATX 2.0 when we really meant ATX
12V 2.0? If you look into the white papers, if memory serves the ATX spec doesn't suggest multi-rails, however ATX12V 2.0 does & the only ATX12V 2.0 PSUs that shouldn't have more than 2+ rails are the ones that don't exceed 18A, as specified in the ATX12V 2.0 specs.
Originally posted by: FastEddie
high output/high amperage single rail ATX 2.0 psu's are designed to function properly using a constant power drain in a range of 60-80% of its peak output. No psu should have to run above 90% of its peak output on a continous basis (as in constant use). That is not an element of their design, nor is it an efficient use of the power resource.
Are you speaking as an electrical engineer? Are there sources that say you can't fully load one of the rails up on a multi-rail PSU without experiencing problems? Perhaps that applies only to single rail PSUs cause when the rail on a single rail PSU is fully utilized, the total output of the PSU is maxed out as well. With the dual rail in question, when you max out just one of it's rails, you don't exactly max out the PSU's entire +12V power output capability.