![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Are modular and multi rail PSU's already dying out?
- Thread starter Dunan
- Start date
Mike Clements
Gone but not Forgotten
- Joined
- Oct 31, 2006
- Messages
- 2,049
There is some bad and incorrect info floating around about modular PSUs. There is nothing wrong with a quality modular PSU in spite of what some might say.
As for multi rails, they are not dying out from what I can see. However, I think the trend will lean towards a smaller number of larger rails as opposed to several (4+) smaller rails.
If you are seeing higher rated PSUs with fewer features, my guess is that it is a cost saving measure as the features add to the cost of expensive units.
As for multi rails, they are not dying out from what I can see. However, I think the trend will lean towards a smaller number of larger rails as opposed to several (4+) smaller rails.
If you are seeing higher rated PSUs with fewer features, my guess is that it is a cost saving measure as the features add to the cost of expensive units.
Mike Clements
Gone but not Forgotten
- Joined
- Oct 31, 2006
- Messages
- 2,049
Only thing is, single rail doesn't meet SLI spex does it?
just curious
i'm seeing more and more PSU's (higher wattage ones) staying with the non modular design and going back to 2 rails (my PC&C 750 silencer for example), is this on purpose?
you mean single rail right? I'd pick 4 rails over 2 rails PSUs, as far as your concerned modular PSUs are the same like other PSUs, you can have your normal PSUs converted to modular PSUs like PCP&C can have their normal PSUs converted to modular requested by customer, modular PSUs aren't dying since it always come later in time!
So manufacturers who do not build modular PSUs got this bad information?
Only thing is, single rail doesn't meet SLI spex does it?
Single rail PSUs are SLI/Crossfire Cerified for example: PCP&C 1KW SR, I'd pick large single rail PSUs over multi-rails PSUs for heavy overclocking on CPU and GPU
Mike Clements
Gone but not Forgotten
- Joined
- Oct 31, 2006
- Messages
- 2,049
I cannot say why certain manufacturers do or do not build modular PSUs.
watertown28
2[H]4U
- Joined
- Jan 30, 2007
- Messages
- 2,216
I agree with this... i would now prefer to have one or 2 SOLID rails over 6 all split up.
As for modular psus dieing out , I dont think that they are, Its just been awhile sense a really good one came out....
Can you provide an example of these rumors and explain why they are not valid?
Redbeard
Official Corsair Rep.
- Joined
- Nov 3, 2005
- Messages
- 1,859
There aren't really any rumors, there's just some comments made by a few manufacturers as to why modular is not as good as non-modular, or vice versa. The resistance difference is negligible, in our original tests, and while you lose a very small amount of efficiency, the trade-off for having modular cables makes more of a difference to most consumers.
As for the single-rail vs. multi-rail thing, it's kind of a moot point, since no PSU I know of has truly electrical independent +12V rails. They're all coming from the same source and just have different physical routes to spread out the resistance, with each route having it's own OCP point so that there's never any serious danger due to overheating cables and all that. That's how multiple rail PSUs seem to work now, or at least how ours works.
This means that although the official spec says 18A each on all 3 +12V rails, what it's supposed to be saying is that the OCP point on each rail is 18A. In reality, that's been raised considerably to allow high-powered GPUs to take advantage of the power they need. With a single 8800GTX taking ~14A at full load, if you put one of those cards on the same rail as a couple of hard drives you could overload it at 18A. So the OCP point was raised much higher.
The reason it's still spec'd that way is due to some complicated issues with ATX 2.2 compliance and our label and box orders back when we designed the product. The product itself never shipped with actual 18A OCP limits on the +12V rails, to my knowledge, since the ATX 2.4 spec allowed us to modify that.
![[Spectre]](https://cdn.hardforum.com/data/avatars/m/89/89728.jpg?1455675927){kind=avatar}
- Joined
- Aug 29, 2004
- Messages
- 22,781
There isn't a manufacturer that has said this to my knowledge there is a brand who has and continues ot beat this drum.....and it's PC Power and Cooling. The validity, or lack there of, has been gone over ad nausem here in the forum.
That's all I wanted. The original post I responded to was a blanket statement and blanket statements make baby Jesus cry. With no disrespect intended, of course.
According to PC Power and Cooling - one of the most respected psu makers in the world - state that multirail psus are obsolete and inefficient. I wouldnt buy one imo.
Here is the link:
http://www.pcpower.com/technology/myths/#8
8. ARE MULTIPLE 12-VOLT RAILS BETTER THAN A SINGLE 12-VOLT RAIL?
With all the hype about multiple 12-volt rails (ads claim that two rails is better than one, five is better than four, etc.), youd think it was a better design. Unfortunately, its not!
Here are the facts: A large, single 12-volt rail (without a 240VA limit) can transfer 100% of the 12-volt output from the PSU to the computer, while a multi-rail 12-volt design has distribution losses of up to 30% of the power supplys rating. Those losses occur because power literally gets trapped on under-utilized rails. For example, if the 12-volt rail that powers the CPU is rated for 17 amps and the CPU only uses 7A, the remaining 10A is unusable, since it is isolated from the rest of the system.
Since the maximum current from any one 12-volt rail of a multiple-rail PSU is limited to 20 amps (240VA / 12 volts = 20 amps), PCs with high-performance components that draw over 20 amps from the same rail are subject to over-current shutdowns. With power requirements for multiple processors and graphics cards continuing to grow, the multiple-rail design, with its 240VA limit per rail, is basically obsolete.
PC Power and Cooling is once again leading the industry. All of our power supplies now feature a large, single 12-volt rail. The design is favored by major processor and graphics companies, complies with EPS12V specs (the 240VA limit is not a requirement) and is approved by all major safety agencies such as UL and TUV.
Here is the link:
http://www.pcpower.com/technology/myths/#8
8. ARE MULTIPLE 12-VOLT RAILS BETTER THAN A SINGLE 12-VOLT RAIL?
With all the hype about multiple 12-volt rails (ads claim that two rails is better than one, five is better than four, etc.), youd think it was a better design. Unfortunately, its not!
Here are the facts: A large, single 12-volt rail (without a 240VA limit) can transfer 100% of the 12-volt output from the PSU to the computer, while a multi-rail 12-volt design has distribution losses of up to 30% of the power supplys rating. Those losses occur because power literally gets trapped on under-utilized rails. For example, if the 12-volt rail that powers the CPU is rated for 17 amps and the CPU only uses 7A, the remaining 10A is unusable, since it is isolated from the rest of the system.
Since the maximum current from any one 12-volt rail of a multiple-rail PSU is limited to 20 amps (240VA / 12 volts = 20 amps), PCs with high-performance components that draw over 20 amps from the same rail are subject to over-current shutdowns. With power requirements for multiple processors and graphics cards continuing to grow, the multiple-rail design, with its 240VA limit per rail, is basically obsolete.
PC Power and Cooling is once again leading the industry. All of our power supplies now feature a large, single 12-volt rail. The design is favored by major processor and graphics companies, complies with EPS12V specs (the 240VA limit is not a requirement) and is approved by all major safety agencies such as UL and TUV.
- Joined
- Aug 29, 2004
- Messages
- 22,781
Yeah...we all know...you keep posting that marketing page and like last time it is still misleading.
One - I dont work for any computer company I work in the dull banking world.
Two - I wasted some money on new "hyped" cpus and couldnt overclock it at all...plus other stupid investments like SLI.
Three - just trying to help others avoid makin same costly mistakes Ive made in the past.
Four - Isnt that what this forum is all about to take into account all points of view?
Two - that's all well and good, but what does that have anything to do with this thread?
Three - I understand, but blindly posting a copy-paste from some manufacturer's propaganda page doesn't help anybody.
Four - Yes, and if it were your own point of view with some actual data, we'd really appreciate it. But we like to rely on facts, not hype from different manufacturers, regardless of who they are.
PC Power & Cooling makes fantastic power supplies, but they've been known to overdo it sometimes when it comes to justifying why they don't follow a particular trend. If you want a solid 1KW PSU, PC Power & Cooling makes a great one, but it's going to be louder and larger than some of its competitors, and it doesn't have modular cables, either.
Maybe some of the people in here with modular, multiple-rail PSUs that are seeing none of these tragic failures that PC Power & Cooling is warning us about could chime in...
The problem with this thread is that its all one-sided. Just trying to add to the options.
I have a hard time with your comments in that your going up against PC Power and Cooling. Where are your facts? PC Power and Cooling is always been the standard to judge other psus...also the most expensive. They have also been making multirail psus for quite some time. Why the sudden shift?
Here is the issue. Is PC Power and Cooling just trying to shift back into a less costly design to sell more of their psus or are they simply stating fact?
I have a hard time with your comments in that your going up against PC Power and Cooling. Where are your facts? PC Power and Cooling is always been the standard to judge other psus...also the most expensive. They have also been making multirail psus for quite some time. Why the sudden shift?
Here is the issue. Is PC Power and Cooling just trying to shift back into a less costly design to sell more of their psus or are they simply stating fact?
It's not one-sided, it's open to interpretation.
You can't blindly believe ANYBODY with something to sell you. That's a fact. PC Power & Cooling makes great PSUs, but they're still trying to sell you a product. A product that is stable, reliable, powerful, overpriced, noisy, and not exactly feature-rich.
That's fine. That's their market. You want facts? Where are yours? You can't put out a "FAQ" entry from PC Power & Cooling as fact, you have to find something objective.
Here's some responses to the PC Power & Cooling thing you posted.
So can a PSU with multiple rails.
This is only true if that rail ONLY provides power to that specific area and if the other rails are limited to providing their "rated" spec. However, I don't know of a single power supply where the 4/8-pin PSU connector is its own separate rail. Not a single one. So in this mythical, bizarro-world PSU that PC Power & Cooling is talking about, they could be right.
The ATX 2.2 power suppy spec available on formfactors.org says you can only have 20A on any single rail in a PSU. It does not differentiate between single and multi rail PSUs so I don't understand their comment here. They're assuming that everyone is building to meet the 2.2 spec, but if this is true, and you're building a single rail PSU, you'd only be able to have a total of 20A on the +12V rail. So they're complaining that multi rail PSUs have a low over-current shutdown point, but the single rail PSU would have only one rail, which must shut down at 20A anyway. So what the hell?
And this being said, if what Redbeard says is true then everyone seems to be building to a newer 2.4 spec which isn't publicly available yet. (which would make sense since the 2.2 spec only goes up to 450W)
Intel and nVidia were largely behind the multiple-rail PSU thing so I don't know how PC Power & Cooling claims that a large, single 12-volt rail is supported by major processor and graphics companies.
As for them stating that the EPS12V spec not having the 240VA limit not being a requirement, Here's a PDF that has the latest EPS12V spec, 2.92, and according to page 35, the 240VA limit IS a requirement, in fact, it goes so far to state that no one +12V rail can carry more than 20A, and if the PSU needs to supply more than 20A on a +12V rail, it must be split into multiple (but no more than 4) +12V rails with 20A or less on each.
I don't mean to be a hardass or anything but you can't just say PC Power & Cooling makes awesome PSUs so they must be right.
What they're trying to do is to get you to buy more PC Power & Cooling power supplies. Which is a noble goal, and the goal every company has. But I'd like it more if they actually pointed out some factual data instead of just spouting off crap that we can easily disprove by looking up the specs ourselves.
Cliff Couser
[H]ard|Gawd
- Joined
- Dec 22, 2003
- Messages
- 1,860
just curious
i'm seeing more and more PSU's (higher wattage ones) staying with the non modular design and going back to 2 rails (my PC&C 750 silencer for example), is this on purpose?
makes no sense to me.
let me ask this question.....how many modular PSU's where there 3 years ago, 2 years ago....1 year ago?
manufacturers don't build on a dying technology/platform (i.e. something that won't sell)
Just to set the record straight...
PC Power & Cooling will not make one of there PSU`s mosular that is not an option directly from them!
Only one company is authorized by PC Power & Cooling to mod there PSU and that company is Performance PCS!!
jonnyGURU
Technical Marketing Manager at Corsair Memory
- Joined
- Apr 9, 2005
- Messages
- 4,844
Ok. So you bought a PSU that wasn't enough for your system, got pissed off and blew a wad on a PCP&C and now you're happy.
Nobody's trying to take that away from you.
PCP&C sell a very good product. Their big problem is the marketing hype they try to create with the exaggerated "facts" in their FAQ's and they count on n00bs to rehash it over and over again. Marketing works. Problem is, people that know facts get pissed off when exaggerated claims are regurgatated over and over. Does it make the PCP&C product any less of a product? Of course not, but spreading FUD that is obviously FUD is no way to gain friends.
PCP&C makes a claim that there is as much resistance in a modular connector as two feet of wire, while REAL TESTS have shown that there is actually FOUR TIMES the resistance in the 18" of wire going from the PSU to the connector even when IT'S NOT MODULAR than there is in one modular connector. There is no doubt that there IS resistance in a modular connector, but to take it and blow it out of proportion so you can better market your product is foul. That's all.
As for the single vs. multi-rail argument: There are good and bad things about both single +12V and multi-+12V designs. But because PCP&C only sell single +12V rail power supplies, they're only going to tell you about the positives of the single +12V rail and the negatives of multi-+12V rails. That's not to say that they're "lying," they're just not telling you the whole truth. Again, does it make the PCP&C and less of a product? No. But for those people that continue to copy and paste paragraphs from their "myths exposed" article, we're seeing the word "SUCKER" slapped across their foreheads and it's a damn shame to see brainwashing in any form happen. It's not too late to step away from the Kool Aid.
In your opinion John is there there any appreciable difference in quality and performance between a $200 PC&C 500 watter and the Seasonic SS-500HT that I can get for 99 bucks? Not that I would spend 99 mind you because my needs are met by the "free after rebate" 500W Ultra V-series.
But I tend to agree with you there is a whole lot of marketing department BS in the power supply game, over trying to turn a power supply which is a commodity item, into a boutique object like it's a Rolex watch or something. I understand that higher quality components cost more and tighter tolerances cost more, and I understand thermal efficiency costs more up front but you recoup some later on the electric bill.
And if power supply A fluctuates by +/- 0.2 on the 12V and power supply B only is +/- .05 V for example does even that necessarily make a difference? Or does it even make a difference by the time to it gets to my motherboard that has 6 phase voltage regulator on the board? Because based on my experience with the Ultra V-Series I am overclocking to the bleeding edge of the chip where adding 1/100 of a volt causes it to reboot. Yet, as long as I do not add that final 1/100th of a volt system is stable and the results are repeatable. Power supply is not getting hot at least the exhaust air, more than about 10% of the time. The thing is quiet too the fan has only sped up a few times. So as long as the PS does not set my house on fire I am more than happy with it. You would think my Vcore would be jumping up and down all by itself but it is remarkably stable. And my GPU is not crashing either.
Anyone can see what a guy who spent $200 more than me on a vid card is gettting for his dollar, and obvious what the guy who spent $200 more on a faster CPU is getting in terms of payback. I am having trouble seeing what the guy who spends $200 more on a PS is getting though.
But I tend to agree with you there is a whole lot of marketing department BS in the power supply game, over trying to turn a power supply which is a commodity item, into a boutique object like it's a Rolex watch or something. I understand that higher quality components cost more and tighter tolerances cost more, and I understand thermal efficiency costs more up front but you recoup some later on the electric bill.
And if power supply A fluctuates by +/- 0.2 on the 12V and power supply B only is +/- .05 V for example does even that necessarily make a difference? Or does it even make a difference by the time to it gets to my motherboard that has 6 phase voltage regulator on the board? Because based on my experience with the Ultra V-Series I am overclocking to the bleeding edge of the chip where adding 1/100 of a volt causes it to reboot. Yet, as long as I do not add that final 1/100th of a volt system is stable and the results are repeatable. Power supply is not getting hot at least the exhaust air, more than about 10% of the time. The thing is quiet too the fan has only sped up a few times. So as long as the PS does not set my house on fire I am more than happy with it. You would think my Vcore would be jumping up and down all by itself but it is remarkably stable. And my GPU is not crashing either.
Anyone can see what a guy who spent $200 more than me on a vid card is gettting for his dollar, and obvious what the guy who spent $200 more on a faster CPU is getting in terms of payback. I am having trouble seeing what the guy who spends $200 more on a PS is getting though.
jonnyGURU
Technical Marketing Manager at Corsair Memory
- Joined
- Apr 9, 2005
- Messages
- 4,844
There is no PC Power & Cooling 500W.
And something like a V-Series 500W is probably more than suitable for 75% of the people out there w/o SLI or large drive arrays IMHO.
I think people need to put into perspective the quantifiable differences between power supplies. Some people rag on bling, but if bling is what you want keep in mind that bling costs money. LED's cost money. Modular interfaces cost money. And, yes, 80%+ efficiency costs money. And there's nothing wrong with buying a $500 PSU that's built like a tank and puts out twice the power you need, but it needs to be said that a $150 PSU will do the job too.
You need to understand the difference between load regulation and ripple/noise. A PSU that drops even as much as 5V from load to idle is not going to have much of an effect on your component's life or even overclockability. But poor regulation may be a symptom of a much greater problem, like heavy and hard loads kicking the voltage way below a tolerable threshold and then quickly bouncing back so fast that your DMM can't register it (that's why I actually perfer analog multi-meters because you can see the needle bounce.) But ripple and noise is what's going to tear up the regulators on your motherboard, not a .02V drop going from idle to load. And that you can't meansure with a DMM.
Me too.
You know Systemax has a build that comes from the factory with three hard drives, a QX6700 and a pair of X1950XTX cards and it's all powered by an X-Finity 600W? Not even the "2G" but the old version? What's that.... like a $100 PSU? For ages I ran a pair of 7800GTX's on a SevenTeam external 500W. Never ran into a hitch. Go figure. My quad-father build with a pair of 8800GTX's in it pulled a max of 450W from the wall. Hmm... I admit, I use a 1kW now. But I don't have it because it's a 1kW PSU. I have it because it's modular, 80% efficient, smaller than most 1kW PSU's at 6" deep and has a 135MM fan which is very quiet. Like I said earlier, it has quantifiable features that I can appreciate.