Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Asgorath

macrumors 68000
Mar 30, 2012
1,573
479
Will the nannies that keep repeating the same thing please exit this thread. You've crapped on this post well enough and everyone has seen your warning stickers. Thanks.

The very same LB/BP's you guys keep suggesting can't carry 150w already route 300w to the PCI's (single trace from start, then branches) and another 250w to the Xeon's. And those channels also have a fault tolerance. Further, 6-pins are rated for 150w each with the 3x 12v rails. I've been running it under heavy load and with multiple platforms and for a few weeks now. Last, I'm posting this message on one of those same platforms while I crunch down a particle system for a vfx shot.

We get it. You think we're all a bunch of naysaying *******s. Let's move on, since some of us are actually trying to get to the bottom of what you're saying in a somewhat productive manner, though you really are making that difficult.

Do you have a photo or other link to the layout of the Mac Pro motherboard? I haven't disassembled mine to the point at which I can see the traces on the motherboard. You keep saying you have docs from HP/Intel that prove what you're saying, do you have links to those? I'm genuinely interested in learning more about it. Where does the trace branch occur exactly?

Again, we've simply been responding with the currently-held understanding of the PCIe power situation in the Mac Pro. If our collective understanding is incorrect, that is extremely useful information.
 
Last edited by a moderator:

blackwoodfx

macrumors member
Original poster
Nov 15, 2010
76
3
Las Vegas, NV
WOW. If I had even 10% of that happen to me, I wouldn't have a single Apple product any more. You are either exceptionally forgiving, or your livelihood depends on Apple products! :)

Apple's got a large portion of my business wrapped around their little ProRes finger. The rest is a general enjoyment of the environment and the performance of their products. There's certainly a lot of forgiveness there and after-all, things are amazing today. iOS is just incredible and I really enjoy most of what Apple is doing. While these issues I listed (and many more like them) make the 'just works' bullet point pretty silly (and getting more ironic every product cycle), the fact remains, life with computers is more enjoyable than ever today and Apple is still the top of that heap.
 

theSeb

macrumors 604
Aug 10, 2010
7,466
1,893
none
WOW. If I had even 10% of that happen to me, I wouldn't have a single Apple product any more. You are either exceptionally forgiving, or your livelihood depends on Apple products! :)

I don't know if I feel like going through that rant point by point. The op seems angry and excitable and I am not sure if I dare incur its wrath.
 

SDAVE

macrumors 68040
Jun 16, 2007
3,573
599
Nowhere
At some point, I bought into that Apple mantra too. Then I quickly realized it's a load of bunk.

For starters; Apple TV never works correctly. There are constant software bugs and connectivity niggles and after snapping up three generations of hardware for multiple televisions across home and studio, Apple announces "it's a hobby". Nice. How about the fact that the iMacs have, for multiple generations now, not matched the hardware dimensions of the Apple LED Cinema Displays? You need to buy an after-market stand to get the monitors leveled for multi-monitor mini-seam transitioning.

Or what about the Airport Express modules that would limply hang off the wall with their prongs exposed, luring children (worst-case) or would simply fall off the wall with the slightest bump (common)? Or how about the fact that the current generation of Airport Extreme and Airport Express hockey pucks are not compatible when attempting to use the Express to expand a network (even though Apple claims they are)? What of the fact that every single 30 pin iToy connector I've ever seen frays at the plug? Or the fact that Apple had to replace all of the original 5v cubes? Or that the first and second generation Apple Garage Band apps had a bug that only allowed monaural output? Pretty significant when you're talking about a music program... Or that Apple bought Logic and drove the product into the ground (which they also did with other major professional tools)?

Remember the iPhone antenna-gate problems? How about the first generation iPhone that had a "lock screen" that didn't actually lock/secure anything? Did you ever use the iPhone 4.* release that caused visual voicemail to not display for up to 30 seconds each time?

I once had an iMac DVD jam in the side-screen slot load. That DVD and the HDD inside of the iMac contained an unreleased album from a major recording artist, plus unreleased/unapproved video footage from a music video rough. I called Apple Support and after attempts at software troubleshooting, they told me it was not possible for me to fix this by myself. I would need to take the iMac into an Apple store to fix it. Because Apple neglected to include the paperclip force eject hole that every other computer on Earth has -- presumably this was because Apple didn't like the aesthetics of a pin hole -- I had to bring everything into an Apple Store to fix this and they would need to bring the computer into the back room (where I'm not allowed) to perform the work. Nope. Couldn't risk it. Had to do the disassembly and dirty work myself. All.. to... eject... a DVD. The drive never jammed again and I made sure to only use external drives for vital work.

Have you ever downloaded an album from iTunes that was corrupted and then had to email Apple and ask for a re-download? I have. Did you ever need to contact Apple support because you had reached the limits of iTunes activations and needed to reset all of them so you could use iTunes again? Yep, been there.

How many times in a week do you need to provide your administrator password, sometimes even multiple times in a row. I distinctly remember Apple making fun of Microsoft for this (which no longer happens in Windows since 7, by the way).

Ever max out your RAM shopping with Apple only to find out later, the hardware can safely support much more but Apple inexplicably won't sell or advertise that? That's fun. And expensive.

Or how about all of us MacBook Air buyers that bought the first generation hardware and only found out later, after light-weight use (browsing internet or playing music) the thing would heat up on battery and the second processor core would shutdown. Nice. Great design there. Again, expensive.

I've also got an iTunes-bought Disney movie I bought for my daughter that inexplicably pops French credits and then de-synchronizes the vocal track in the middle of playback. I've emailed Apple about that and never heard a word back. But because the video file is 'in the cloud' via iTunes Match, it can't be fixed on our end.

Or that Apple Maps doesn't work as-well-as Google Maps and even though they had a license that ran until next year, they pulled the plug in the interest of competition only.

There are countless more. These just off the top of my head. "It just works" is a marketing bullet point.

But hey, these 7970's do 'just work' and no one here wants to believe that...

I'm not really arguing with you :)
I'm glad it works and am happy for you and for others who want to use it!

My point was, for me, I'd rather treat it as a temporary product which I do work on, and then sell it or keep it as a backup and buy a new one and continue working. That is all. I do not care about computers anymore, I barely have time to myself. I just want them to work "ok".

Coming from a PC background since the early 90's, I can say that Macs definitely need to be taken care of less for them to work, and it's not due to the hardware, but OS X. After all, they all use Intel parts now, anyway.

My overall point was that Apple should and cannot be seen as a vendor like Dell, HP, etc. because they are never shooting for that. They want computers to be just like any other appliance. That kind of thinking comes from German industrial designers such as Dieter Rams and architects such as Mies van der Rohe. This is what makes Apple a different company than the rest of them.

Now, I am not going to deny that Macs have their own issues, believe me, I know they have many issues...but usually it's less of a technical challenge to get it working than it is for a Windows based PC. I don't need access to BIOS settings or settings for primary/slave, etc etc, or bus speeds. I just want it to work.

Hopefully you understand my point.

However, it's great to see threads like these. I also don't think putting such a card will fry the PSU. That thing is 1,000 watts as far as I remember. But see, not thinking about watts or voltages is the fun part of Macs :) I don't miss my PC building days at all :)

You apparently do like fiddling with computers and by all means, do it. You might be better suited for Windows, since it is definitely more open when it comes to picking your own bits and pieces and upgrading down the line. I know you're sticking with Apple probably because you like OS X and the way the hardware is designed...but like I mentioned before, don't be completely pissed off at Apple when you can't upgrade too many things on their computers, because of the reasons mentioned above. There are plenty of other choices out there FAR cheaper and with more performance than Apple hardware...but not by far when you consider that the Mac Pro's use Xeon CPU's...I remember I did comparisons part by part back in 2008 when I was purchasing the Mac Pro, and it actually was more expensive to build my own workstation with the same Intel parts than it would be to get a Mac Pro.
 

slughead

macrumors 68040
Apr 28, 2004
3,107
237
It's one 6 pin cable pre-spec'd for 150w with two extra grounding wires using a 6-to-8 pin connection. You guys are terrified of this and it has crossed the threshold from 'odd' and has moved fully into 'weird'.

You still have no idea what the "specs" actually say, in spite of the fact that several people have told you. 6 pin standard only has FIVE wires, TWO of them 12 volt leads. Yes, we can all see there is a 3rd 12 volt lead on most 6 pin cables, but this is acting OUTSIDE the standard but does NOT imply the PSU or in this case the motherboard tracers are able to provide the 150watts of power that might be drawn. In fact, I have a PSU running my GTX670 right now with only two 12v leads.

I'll do you one better: I actually looked up the physical limitation of these 6 pin wires/plugs and found out that it's somewhere around 280watts per cable!

What does this mean? Absolutely nothing! I really don't give a crap about some $15 cable I got off E-Bay or what it's rated for. What really matters is which expensive components might get damaged.

Heck, I don't even care if the PSU can't handle it. The PSU has safeguards in place. I'm most worried about the motherboard tracers, which is an Apple-specific design flaw. These thick wires are more likely to heat up and damage surrounding components than to literally explode or pop.

Your anecdote of it working without frying is not the end-all, either. I've run too much current through circuits I've built before which have run many times before failing. The problem is with component longevity. Even if it'll take a year to run into the circumstances that will kill my Mobo, that's not acceptable.

You have presented no evidence saying you have any idea what the tolerances are for these tracers, and instead you have some unfounded ideas about the way power is distributed. At least, I think they're unfounded, being as how you've given no pictures/charts/graphs/links/sources/spiritual channellings otherwise.

It's very frustrating having a conversation with someone while 1) posting links they don't read 2) being told there are sources and not being shown them.

That said, after reading a couple reviews stating the TPD to be around 250, I think you're probably right that it's safe. The review I read initially had the TPD at around 275 peak, which is really pushing it. Would I risk it in a >$2500 Mac Pro ? No, but I think it's probably okay.

As for the tone of this "debate"... I happen to agree with a lot of people here:
Do you have a photo or other link to the layout of the Mac Pro motherboard? I haven't disassembled mine to the point at which I can see the traces on the motherboard. You keep saying you have docs from HP/Intel that prove what you're saying, do you have links to those? I'm genuinely interested in learning more about it. Where does the trace branch occur exactly?

The op seems angry and excitable

Ya know, I've been on your side in this thread because I'm all for people experimenting with things that are out of the norm, but whenever someone has asked you for solid evidence you provide anecdotes and estimations. Posting a few links/pictures/etc.. to support your claims definitely wouldn't hurt your case.

- You post that it's working for you.
- Someone questions your math on the power draw (which was incorrect).
- I post that our standard recommendation is to use an external PSU for cards that require more than 225W.
- You start the name calling.

Do you have a link to the spec for the HP/Intel product? I'd be interested to see it.

So now that I'm kind of on your side (no thanks to anything that you've said, by the way), can you back up ANY of the claims you made? Where's the source on the tracer map? Where is the "spec" of how much current those tracers were designed to carry?
 
Last edited:

blackwoodfx

macrumors member
Original poster
Nov 15, 2010
76
3
Las Vegas, NV
I'll do you one better: I actually looked up the physical limitation of these 6 pin wires/plugs and found out that it's somewhere around 280watts per cable!

I posted on page 2: This is the *actual* spec of the MP PSU PCI wires: 21a @ 12v = 252w with contacts rated to 36a @ 12v = 432w. The now many years old 6 pin PCI specification someone earlier linked to that stated 5-wires is totally irrelevant to what's going on here. What matters is what's in the chassis and what Apple and Intel baked into the boards. That's a 6 lead strand with 3x 12v hots and 3 grounds. At 8 pin it becomes 5 grounds. Mathematically, perfectly well within reason to pull 150w (and more).

For the 7970, I'm plugging in a 6 pin (75w) an 8 pin (150w) and the PCI slot (75w) which provides 300w. The card uses < 250w. That's 50w under thresh and it's been non-stop warning stickers and chicken littles ever since...

So, I've got a question for you guys; Why are you all so insistent the LB or the BP can't handle this 'current' when you must know the very same boards handle much more than what we pull with the 6-to-8 pin adapter plugged in? I asked this earlier and everyone ignored it. The same LB everyone is concerned with already handles much more when routing power for the PCI lanes and the CPU sockets. Why are you all convinced it cannot do more, yet you also perfectly understand that a 3 hot PCI cable is absolutely built to pull 150w. Why wouldn't the PCB tracer also accept the same tolerances?

Do you somehow believe the PCB traces used for the PCI power connectors are of a different material than you know, all of the other ones on the board? Does that seem logical to you?

Specifically because, you know, I've proven this by successfully running multiple 7970's (on multiple Mac Pro's) at full loads and for extended burn-ins in a hard working environment and there is only ONE goddamn example anyone can find, on the entire internet and all of MacRumors, of someone actually killing their LB traces and that was doing something drastically over what I keep saying is safe here.

By the anti-7970 logic floating in this thread, you'd have to assume Apple+Intel (1) didn't bake in significant precaution when using PCB tracers on their boards for power, (2) ignored the thresholds of the standards they were using, (3) assumed under-providing tolerances would be better than bad press of melting Mac Pro towers, (4) Intel produced something 'mechanically' inferior to the same PCB traces they also used for their HP sever boards (and presumably others), (5) never anticipated people using PCI connector splitters to draw more power for additional cards, etc etc etc. That's YOUR logic (if you don't believe this works and is safe). What's mine? ...It's in spec, already on the board, well within thresh of the wires and what the PCB's already manage, and absolutely predictable from an engineering stand point.

OH... And Apple included the ****ing driver for this card. Weird huh?

I think you guys need to prove this doesn't work. Not sure why the onerous is on me to when I'm demonstrating pretty clearly it does.
 
Last edited:

SDAVE

macrumors 68040
Jun 16, 2007
3,573
599
Nowhere
Apple's got a large portion of my business wrapped around their little ProRes finger. The rest is a general enjoyment of the environment and the performance of their products. There's certainly a lot of forgiveness there and after-all, things are amazing today. iOS is just incredible and I really enjoy most of what Apple is doing. While these issues I listed (and many more like them) make the 'just works' bullet point pretty silly (and getting more ironic every product cycle), the fact remains, life with computers is more enjoyable than ever today and Apple is still the top of that heap.

Honestly, your work is pretty bad and I am not sure why you complain so much when you can spend the time that you're complaining with doing actual good work.

It's easier to criticize than be in the lion's den doing good work. Do you know how many high end studios run Apple computers and do amazing work? Yes, they have issues. Yes, but they get stuff done and don't spend time on a forum complaining about it.
 

blackwoodfx

macrumors member
Original poster
Nov 15, 2010
76
3
Las Vegas, NV
Apple's got a large portion of my business wrapped around their little ProRes finger. The rest is a general enjoyment of the environment and the performance of their products. There's certainly a lot of forgiveness there and after-all, things are amazing today. iOS is just incredible and I really enjoy most of what Apple is doing. While these issues I listed (and many more like them) make the 'just works' bullet point pretty silly (and getting more ironic every product cycle), the fact remains, life with computers is more enjoyable than ever today and Apple is still the top of that heap.
Honestly, your work is pretty bad and I am not sure why you complain so much when you can spend the time that you're complaining with doing actual good work.

It's easier to criticize than be in the lion's den doing good work. Do you know how many high end studios run Apple computers and do amazing work? Yes, they have issues. Yes, but they get stuff done and don't spend time on a forum complaining about it.

Comment of the year. That's flying your true colors. Now all we need is a nazi reference and this thread will be complete.

While it's almost certain you have seen at least some of my major work and/or contributions, it's equally certain you had no idea and you still don't. Whatever impression you got from Googling just now, I'm sorry it didn't live up to your lofty personal standards. Let's not judge a successful studio and all of the combined talent within by a business card, shall we?

Quoting me praising Apple while simultaneously suggesting I complain about them too much and should do something productive with my own 'bad work'... That's just classy.

You've all been a real pleasure and I appreciate everyone taking the time to comment. Peace.
 

Asgorath

macrumors 68000
Mar 30, 2012
1,573
479
Okay, I'm back, because I'm still trying to learn more about this issue. I'm just going to steer clear of the name calling, no matter what you say.

I posted on page 2: This is the *actual* spec of the MP PSU PCI wires: 21a @ 12v = 252w with contacts rated to 36a @ 12v = 432w. The now many years old 6 pin PCI specification someone earlier linked to that stated 5-wires is totally irrelevant to what's going on here. What matters is what's in the chassis and what Apple and Intel baked into the boards. That's a 6 lead strand with 3x 12v hots and 3 grounds. At 8 pin it becomes 5 grounds. Mathematically, perfectly well within reason to pull 150w (and more).

Source? Did you measure this yourself, or do you have documentation? That's all we're asking for. At this point, I believe that you may be correct. All we're asking for is some data to back up what you're posting here, because IT COMPLETELY CHANGES OUR UNDERSTANDING OF WHAT THE MAC PRO CAN SUPPLY. Forgive us for being a little cautious about throwing our understanding of what the system can provide out the window, just because you say so.

Personally, I did not know that the Apple 6-pin connector provided the 3 "hot" 12V wires, since the spec linked earlier only calls for 2. I was not aware of the fact that many (most?) 6-pin cables actually have the same number of hot leads as the 8-pin cables.

So, I've got a question for you guys; Why are you all so insistent the LB or the BP can't handle this 'current' when you must know the very same boards handle much more than what we pull with the 6-to-8 pin adapter plugged in? I asked this earlier and everyone ignored it. The same LB everyone is concerned with already handles much more when routing power for the PCI lanes and the CPU sockets. Why are you all convinced it cannot do more, yet you also perfectly understand that a 3 hot PCI cable is absolutely built to pull 150w. Why wouldn't the PCB tracer also accept the same tolerances?

Because nobody has seen a motherboard layout diagram or photo that shows the traces you're talking about? We've requested this numerous times. Forgive me for not taking your word for it, nobody is questioning the fact that the motherboard supplies 300W to the PCIe slots, but the 6-pin connections are at a completely different end of the motherboard. Surely there's a trace that branches off from the main line to the 6-pin connections on the motherboard? You seem to be suggesting that the connection can actually support 3 12V wires, which would imply that the traces leading up to it were designed to support all 3 being active.

OH... And Apple included the ****ing driver for this card. Weird huh?

Apple (and NVIDIA before them) has been shipping a driver for the GTX 580, which also has an 8-pin connector, for a long time now. Doesn't mean it's officially supported in the Mac Pro.

Edit: Seconding Slughead's reservation about doing something that might destroy my motherboard after a 6 or 12 months of hard usage. I'm reminded of the 300MHz Celerons that could be overclocked to 450MHz, which many people did, and then complained when their systems started to get really unstable after a few months.
 
Last edited:

ClassObject

macrumors 6502
Mar 1, 2010
272
1
That's all we're asking for.

All we're asking for is some data to back up what you're posting here,...

We've requested this numerous times.

Who is this "We" you speak of? Have some balls and speak for yourself. If you think the OP claims are bogus, hook up a 7970 and prove you fried a motherboard. Otherwise, your objections are well understood and have become tiresome.
 

Asgorath

macrumors 68000
Mar 30, 2012
1,573
479
Who is this "We" you speak of? Have some balls and speak for yourself. If you think the OP claims are bogus, hook up a 7970 and prove you fried a motherboard. Otherwise, your objections are well understood and have become tiresome.

Forgive me for including DanielCoffey, slughead and others who have posted in this thread when asking for more details. I'm pretty sure they won't mind if I say "we" when asking the same questions that they've been asking. I'm talking about the commonly-accepted information that has been floating around the forums for as long as I've been reading them. I'm not making up the "225W limit", I learned about it from discussions on these forums. This knowledge has been propagated for a long time. That's what I'm talking about.

There was nothing in my last few posts that was an objection. I'm simply requesting more information, which I and others have been doing for the past few pages. I am not bashing the 7970, nor AMD, nor Apple, nor revelstudios. I'm (yes, I'm) simply asking for the source of the information he's posting, because it completely contradicts what is the widely-held understanding of the PCIe power situation for the Mac Pro. This is important stuff, and I'm (yes I'm) not content to just take his word for it. If that makes me a nanny or an *******, so be it.
 

slughead

macrumors 68040
Apr 28, 2004
3,107
237
[summarized] I assume the 6 pin coming off the motherboard can handle 150 watts based on nothing and refuse to provide sources saying so. I also concocted this "shared power" tracers scenario based on no information, comparing it to other server boards which you are to assume I have knowledge of but refuse to post sources for.

Maybe if you linked to the specs for the HP server boards--which I'm guessing don't even have PCIe 6 pin power--or even linked to ANY of the specs you refer to, we'd be satisfied.

We've been asking for JUST ONE SOURCE, ANY SOURCE for 2 pages now, and you keep eluding to the intel/HP specs which you have yet to show us. I've googled it and couldn't find anything. you linked to the Mac Pro's PSU specs, but nobody asked for that and it wasn't even a concern.

If Intel has created mobos in the way you've described before, please show us, it wouldn't be the best proof but it would work for me.


Who is this "We" you speak of? Have some balls and speak for yourself. If you think the OP claims are bogus, hook up a 7970 and prove you fried a motherboard. Otherwise, your objections are well understood and have become tiresome.


I would like to know as well, I've asked several times for a source to back him up.
 
Last edited:

ClassObject

macrumors 6502
Mar 1, 2010
272
1
Maybe if you linked to the specs for the HP server boards--which I'm guessing don't even have PCIe 8 pin power--or even linked to ANY of the specs you refer to, we'd be satisfied.

We've been asking for JUST ONE SOURCE, ANY SOURCE for 2 pages now, and you keep eluding to the intel/HP specs which you have yet to show us. I've googled it and couldn't find anything. you linked to the Mac Pro's PSU specs, but nobody asked for that and it wasn't even a concern.

If Intel has created mobos in the way you've described before, please show us, it wouldn't be the best proof but it would work for me.





I would like to know as well, I've asked several times for a source to back him up.

This is an Internet forum not a court of law. If you think OP is full of duke, put up or ...
 

slughead

macrumors 68040
Apr 28, 2004
3,107
237
This is an Internet forum not a court of law. If you think OP is full of duke, put up or ...

We've already posted the official tolerances of the 6 pin standard and shown they are not even close to the 150 watts the OP assumes. OP thinks that just because his card had an adapter in the box that allows him to run the card, everything is okay. He assumes that because Apple included a driver, it must be okay to use this card. He has actually said these things, which really speak to his credibility. As someone else pointed out, the GTX580 is also supported, and that certainly doesn't mean people should use 8 pin adapters to run it (even if they're included in the box!).

We also have posted links to people pushing the wattage towards the 150watt/connector limit and destroyed their motherboard.

Our position this whole time has been one of: People have messed things up this way, if you're suggesting people ignore this, please provide the spec tolerances saying it's okay for the wattage range you're indicating. He has provided nothing.

The reason we don't have reports flooding in of people doing this is that most people don't have the kind of money to risk a $3000 machine (or $1200 mobo) to save $60 on an external power supply. Every thread featuring a high-power card mentions whether or not it requires another PSU. Also, most MP owners stick with the Apple-provided cards.

We're also saying that even if it works for a while, that doesn't mean hardware longevity won't be impaired by doing this. We point to examples of over-clocking which is the same concept: running more current through a circuit than it's rated for, which can drastically decrease the life of computer components (would you like a source for that?)
 
Last edited:

All Taken

macrumors 6502a
Dec 28, 2009
780
1
UK
Actually if you read the threads about upgrading graphics cards on this forum, this information is pretty much a collective fact.

I've read and asked the question myself, the 'fact' of it all is that the community doesn't know either way. Many have taken to repeating the same statement about the connectors and traces etc etc but don't have any proof. Sure we have loose specification but until somebody such as the OP tries, we don't know. You could spin off a paragraph or two about how various forum members aka 'the community' have knowledge of this but when it comes to 'FACT' - we have nothing.

I'm all for caution but I get sick of hearing the same members repeating statements when they have no fact in them.

I'll leave you with this - Actually if you read the threads about upgrading graphics cards on this forum there are no facts, no real examples and a whole lot of mis-information for people to repeat. As you've proven.
 

blackwoodfx

macrumors member
Original poster
Nov 15, 2010
76
3
Las Vegas, NV
OP thinks that just because his card had an adapter in the box that everything is okay. He assumes that because Apple included a driver, it must be okay to use this card. He has actually said these things, which really speak to his credibility.

Um, no. No I didn't. But this comment sure speaks to your credibility.

This forum is a waste of time. Good riddance.
 

All Taken

macrumors 6502a
Dec 28, 2009
780
1
UK
We've already posted the official tolerances of the 6 pin standard and shown they are not even close to the 150 watts the OP assumes. OP thinks that just because his card had an adapter in the box that allows him to run the card, everything is okay. He assumes that because Apple included a driver, it must be okay to use this card. He has actually said these things, which really speak to his credibility. As someone else pointed out, the GTX580 is also supported, and that certainly doesn't mean people should use 8 pin adapters to run it (even if they're included in the box!).

We also have posted links to people pushing the wattage towards the 150watt/connector limit and destroyed their motherboard.

Our position this whole time has been one of: People have messed things up this way, if you're suggesting people ignore this, please provide the spec tolerances saying it's okay for the wattage range you're indicating. He has provided nothing.

The reason we don't have reports flooding in of people doing this is that most people don't have the kind of money to risk a $3000 machine (or $1200 mobo) to save $60 on an external power supply. Every thread featuring a high-power card mentions whether or not it requires another PSU. Also, most MP owners stick with the Apple-provided cards.

We're also saying that even if it works for a while, that doesn't mean hardware longevity won't be impaired by doing this. We point to examples of over-clocking which is the same concept: running more current through a circuit than it's rated for, which can drastically decrease the life of computer components (would you like a source for that?)

I think you're 'dumbing' down the OP a bit here. I doubt he literely thinks that it's okay because of a driver in beta + a cable in his graphics card box. Stretch? :p If he did think it was okay for that reason then crazy days but I'm guessing in actual fact he researched the tolerances and the TDP of the card and came to a conclusion based on fact not guess work.
 

SDAVE

macrumors 68040
Jun 16, 2007
3,573
599
Nowhere
Comment of the year. That's flying your true colors. Now all we need is a nazi reference and this thread will be complete.

While it's almost certain you have seen at least some of my major work and/or contributions, it's equally certain you had no idea and you still don't. Whatever impression you got from Googling just now, I'm sorry it didn't live up to your lofty personal standards. Let's not judge a successful studio and all of the combined talent within by a business card, shall we?

Quoting me praising Apple while simultaneously suggesting I complain about them too much and should do something productive with my own 'bad work'... That's just classy.

You've all been a real pleasure and I appreciate everyone taking the time to comment. Peace.

tumblr_m6ur5dqCCm1rtznyyo1_400.jpg
 

slughead

macrumors 68040
Apr 28, 2004
3,107
237
I think you're 'dumbing' down the OP a bit here. I doubt he literely thinks that it's okay because of a driver in beta + a cable in his graphics card box. Stretch? :p If he did think it was okay for that reason then crazy days but I'm guessing in actual fact he researched the tolerances and the TDP of the card and came to a conclusion based on fact not guess work.

You tell me what he meant by this. Maybe I'm wrong. After a long rant about how it just simply MUST be the case that what he's doing is okay, he says this:

OH... And Apple included the ****ing driver for this card. Weird huh?

As far as him reaching conclusions based on facts, if it's based on facts I could be swayed. All he'd have to do is show the MB tracers could handle the wattage, and extrapolate the 6 pin power draw from the TPD, and show that the latter was less. He basically just assumed that he can run up to 150 watts off of one of the 6 pin cables (a standard which is only 75 watts), then when he extrapolates the wattage he claims it's less based on that. He has provided ZERO evidence the MP can support 300watts for a card, and by pointing out the standards we have to assume Apple is following and showing the consequences of failing to follow that standard, we are presenting counter-evidence.

Now after clarifying the TPD of the 7970 (which we'll say peaks at 250 watts) I've even said that I think the 7970 is probably okay to use, and will likely not cause long term damage. However, I base this on a lot of assumptions about Apple/Intel's "wiggle room" which is totally a guess. As such I recommend just dropping $60 on a freaking external PSU which takes 15 minutes to setup. OP takes a more hardline view based on silly assumptions like "Apple provided a driver [therefore it must be okay]" and "the 6 pin has the same number of 12v wires, [therefore it must be okay]" and "I've run this thing for a little while [therefore it must be okay]".

It's not only misleading, but it's know-it-all unfounded nonsense. Again: Please don't misunderstand, if I were given evidence, any evidence at all that the MP could safely run 300watts/PCIe card using the combination of power sources running through the motherboard, I would be fine. Heck, even if I were just presented with evidence of the MP shutting off when those tolerances were reached I would have fewer trepidations. I've read 2 reports of the MP getting damaged from this and 0 of it shutting down before that damage occurs.

If I'm wrong please let me know, but as far as I can tell it is that ridiculous.
 
Last edited:

All Taken

macrumors 6502a
Dec 28, 2009
780
1
UK
You tell me what I meant by this. Maybe I'm wrong. After a long rant about how it just simply MUST be the case that what he's doing is okay, he says this:



As far as him reaching conclusions based on facts, if it's based on facts I could be swayed. All he'd have to do is show the MB tracers could handle the wattage, and extrapolate the 6 pin power draw from the TPD, and show that the latter was less. He basically just assumed that he can run up to 150 watts off of one of the 6 pin cables (a standard which is only 75 watts), then when he extrapolates the wattage he claims it's less based on that.

If I'm wrong please let me know, but as far as I can tell it is that ridiculous.

He's using a little bit of overhead or a whole lot of the overhead on those traces, or none at all? which one? We need a definitive test, I can take donations if anyone wants the absolute limit?
 

slughead

macrumors 68040
Apr 28, 2004
3,107
237
He's using a little bit of overhead or a whole lot of the overhead on those traces, or none at all? which one? We need a definitive test, I can take donations if anyone wants the absolute limit?

I'd be in favor of that. However, the reality is that I'm not too worried about the mobo actually catching fire or anything under these tolerances. If you look at the problems overclockers experience, they're a lot more subtle like errors, instability, and decreased hardware lifespan. Therefore I'd be much more interested in what the physical tolerances of the tracers are, which is easy to calculate based on the cross-sectional area of the tracer and the material it's made of (likely copper)--or we could just ask Intel.

edit: btw I added more to my above post
 
Last edited:

theSeb

macrumors 604
Aug 10, 2010
7,466
1,893
none
Edit: Seconding Slughead's reservation about doing something that might destroy my motherboard after a 6 or 12 months of hard usage. I'm reminded of the 300MHz Celerons that could be overclocked to 450MHz, which many people did, and then complained when their systems started to get really unstable after a few months.

Hahaha. I was going to post this. I am one of those people that ran my 300 MHz Celeron at 450 MHz by moving a jumper on my motherboard. It was great... except in the summer when the computer would overheat in games and either crash or restart. Then after about a year the motherboard went poof.

The point here is pretty simple: if you are running things beyond the operating specifications, then don't be surprised when it all goes horribly wrong. The fact that the card has been "running perfectly for weeks" proves very little to me.

----------

Um, no. No I didn't. But this comment sure speaks to your credibility.

This forum is a waste of time. Good riddance.

Why are you so angry? People here are advising caution to other readers based on facts and how that electrical stuff works. You then become angry, start calling people names and then you have the audacity to accuse people of calling you names? Seriously, tone it down because your posting style does not give you any credibility. They actually remind me of a baby throwing their toys out of the cot.
 

All Taken

macrumors 6502a
Dec 28, 2009
780
1
UK
Hahaha. I was going to post this. I am one of those people that ran my 300 MHz Celeron at 450 MHz by moving a jumper on my motherboard. It was great... except in the summer when the computer would overheat in games and either crash or restart. Then after about a year the motherboard went poof.

The point here is pretty simple: if you are running things beyond the operating specifications, then don't be surprised when it all goes horribly wrong. The fact that the card has been "running perfectly for weeks" proves very little to me.

----------



Why are you so angry? People here are advising caution to other readers based on facts and how that electrical stuff works. You then become angry, start calling people names and then you have the audacity to accuse people of calling you names? Seriously, tone it down because your posting style does not give you any credibility. They actually remind me of a baby throwing their toys out of the cot.

Just stop. No facts are available and you're talking like you're advising people with them. At least Slughead realises it is an area of unknown. I would genuinely like to know the true limit here instead of listening to misinformed users spouting off about a limit they know nothing about.

----------

In terms of a true test what would we need?

Back Pane logic board for a 2010 Mac pro?
A couple of cards 250w+ at load? (In order to verify results)

I won't say it will work with (X) wattage on the traces as nobody really knows, it hasn't been tested. I also won't adopt the crush and refute approach to new posts on the matter as seen across the boards. Until we know we shouldn't form the opinion that a couple of people have repeated to be fact!!

Anyone have anything constructive in terms of a test setup?
 

Inconsequential

macrumors 68000
Sep 12, 2007
1,978
1
Let's just put this 2/3 12V wire thing to bed.

Based on these sources:
http://www.pcisig.com/developers/ma...c_id=fa4ec3357012d69821baa0856011c665ac770768
http://www.playtool.com/pages/psuconnectors/connectors.html#pciexpress8

The 6 pin is 3x12V and 3x Grounds.
The 8 pin is 3x12V and 5x Grounds.

The difference in current carrying capacity is APPARENTLY the rating of the 12V wires, at 2.083A and 4.167A respectively, now if Apple have used higher gauge wire from the PSU to the backplane traces then we can safely say that between the PSU and the backplane we are within specs.


The questions we need answering are:

1. The rating of the backplane traces
2. The rating of the mini 6-pin adapter on the backplane.
3. If there is any redundancy in the backplane's traces.
 

theSeb

macrumors 604
Aug 10, 2010
7,466
1,893
none
Just stop. No facts are available and you're talking like you're advising people with them. At least Slughead realises it is an area of unknown. I would genuinely like to know the true limit here instead of listening to misinformed users spouting off about a limit they know nothing about.


I am sorry. I have forgotten that I have no idea what I am talking about.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.