Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
No they're not silent - and the ATi cards run hotter/noiser than the nvidia ones.

I may actually record mine one day and put the clip up...
 
Mmmm yeah, wow! Excellent powers of observation!

Yes, I upgraded from the 5150 (x2) to the 5355 (x2).

5150 http://ark.intel.com/Product.aspx?id=27218 65W
5160 http://ark.intel.com/Product.aspx?id=27219 80W
5355 http://ark.intel.com/Product.aspx?id=28035 120W
5365 http://ark.intel.com/Product.aspx?id=30702 150W


I asked about this at the time on the Apple support forums and everyone told me
the parts were identical and not to worry about it.

2006
Mac Pro Quad 2.0GHz
Mac Pro Quad 2.66GHz
Mac Pro Quad 3.0GHz

2007
Mac Pro Quad 2.0GHz
Mac Pro Quad 2.66GHz
Mac Pro Quad 3.0GHz
Mac Pro 8-core 3.0GHz​
I just assumed from that, that Apple was using the beefiest HS for all models to
make production easier/cheaper or something. How sure are you they're different?

http://www.welovemacs.com/6614458.html

http://arstechnica.com/hardware/reviews/2006/08/macpro.ars/4


Well....the heatsinks between the original 2.66Ghz quad and the Penryn 3.2Ghz are most definitely different. My Google-fu can't turn up any more info than that. It's something you may want to look into.

Oh, and since you re-used the heatsinks, you did remember to remove the old paste with alcohol or something like that and add new stuff, right?
 
http://www.welovemacs.com/6614458.html

http://arstechnica.com/hardware/reviews/2006/08/macpro.ars/4


Well....the heatsinks between the original 2.66Ghz quad and the Penryn 3.2Ghz are most definitely different. My Google-fu can't turn up any more info than that. It's something you may want to look into.

Oh, and since you re-used the heatsinks, you did remember to remove the old paste with alcohol or something like that and add new stuff, right?

Not alcohol tho! I'm a pro! I use acetone! :) hehehe...

Anyway, all of my original heat testing and concerns stemmed from an unmodified MP with the original dual X5150 procs installed. That and looking at all of the other posts at the time over in the Apple support and developer forums. The x5355 procs run about two degrees hotter than the X5150 ones did under all conditions. So like bozz, I think it's not something to worry about. Thanks for the links BTW!!!

And you're right the 2008 ones clearly show more heat-pipes.

I think my next machine will likely look like this tho:

l_24b11351fe52b934e9eea56f04d5e911.jpg

http://aquamac.proboards.com/index.cgi?board=hack1&action=print&thread=460
 
Right now at default Ambient: 27c CPU: 34

If my room temp gets into the 33c range then the cpu jumps up to about 55-60 idle but that is where Smcfan comes in handy. :)

One question I do have is why doesn't the 09 MP ramp up the fan speeds by itself? I've seen it get near 70c and still at the default fan speeds!
 
If it's like the 06, 07, and 08 Mac Pros it actually does ramp up but not enough to make any actual difference. You can see a 2006 MP doing it here between "B" and "D" on the bottom graph:

Hardware_Monitor_Fan-Temp_Relation.jpg


Apple wants to wait for the CPUs to hit 75C before they take any action at all and then their goals seem to be ONLY to keep the CPU below 80C. As long at your CPU temps are below 75C Apple will not increase fan speed. And as you can see in that graph once they think the temp is stable and not increasing they will even attempt to reduce fan speed again. So their goal is to keep your CPU at 75C. LOL. This requires such a slight increase in fan speed to achieve you probably can't notice that the fans have indeed actually increased.

2009 may be different I dunno...


EDIT: They do not seem to consider drive temp at all and I question if they're considering RAM temps. Tho they probably are for the RAM. They do also consider PSU temps as I've seen in other graphs not shown here (tho you can kinda see it here too).
 
Apple wants to wait for the CPUs to hit 75C before they take any action at all and then their goal seem to be ONLY to keep the CPU at below 80C. As long at your CPU temps are below 75C Apple will not increase fan speed. And as you can see in that graph once they think the temp is stable and not increasing they will even attempt to reduce fan speed again. So their goal is to keep your CPU at 75C. LOL. This requires such a slight increase in fan speed to achieve you probably can't notice that the fans have indeed actually increased.

2009 may be different I dunno..

Thanks, always a wealth of knowledge!

The 2009 is the same at least in my experience. Our AC was out and it got into the 90's inside and the MP was registering 35c Ambient and 69c CPU and STILL didn't increase the fans. I couldn't believe it so I quickly turned on SMCFAN and only made the defaults near 1000rpm for BOOSTA and Intake/Exhaust and it dropped down to 45c.

I never feel comfortable having my CPU running anywhere near 70c
 
Quick question about smcfan control: can you set a new minimum that the OS will be able to increase if need arises? Or is it that, once you go manual, you have to watch the temps yourself?

Thanks

Loa
 
Quick question about smcfan control: can you set a new minimum that the OS will be able to increase if need arises? Or is it that, once you go manual, you have to watch the temps yourself?

Thanks

Loa

Yes, that's basically all smcFanControl is doing. Setting a higher minimum value.

But if that value is over 800 or 900 then I don't think OS X will ever do anything to it simply because it doesn't think it's warranted. But yeah, if you set 1000 RPM and then your CPU(s) hit 75C then Apple will increase it more. At least as I understand it - this is the case.
 
Yes, that's basically all smcFanControl is doing. Setting a higher minimum value.

Good. I'll give it a try as I should receive my 4 RE3 disks early next week and I want to keep everything cool even during the summer heat.

Loa
 
yes and half the latency. :)

i dont think its that linear though, because most of plugins oversample and when you have 192khz you dont have to oversample that much/at all << less cpu load

although i do believe that given the goal medium 4x sample rate is still a bit overkill.

Well 192Khz is pretty much. But if you could handle 96Khz in Audio projects, that would be very nice. I never noticed, that the more sample rate you chose, the less latency you'll have... Normally I fix it with the buffersize as you know.
 
Yeah, if you want it in a Mac case it seems like a bit of a project. ;) I dunno tho... I might just build my own case. I was a cabinet maker in a former life and something like below looks pretty sweet to me!


shintomod.jpg
dragoncase-thumb.jpg
2056828740_ff3f1afcff_o.jpg

damnation-pc-case-mod.jpg
001.jpg
01russwork1.jpg

final09.jpg
166_s.jpg
woodcasemod.casemodblog.com.jpg

There's a buttload of very interesting ideas if you're up for a project:

I'm not ready to start yet tho. I guess in about 8 months to a year. I wanna watch the market for awhile longer. I keep waiting for something totally new to arise. Like quantum computing or something. :D

I guess I can make something that will run completely silent and still be very cool. ;)
 
Yeah, if you want it in a Mac case it seems like a bit of a project. ;) I dunno tho... I might just build my own case. I was a cabinet maker in a former life and something like below looks pretty sweet to me!


shintomod.jpg
dragoncase-thumb.jpg
2056828740_ff3f1afcff_o.jpg

damnation-pc-case-mod.jpg
001.jpg
01russwork1.jpg

final09.jpg
166_s.jpg
woodcasemod.casemodblog.com.jpg

There's a buttload of very interesting ideas if you're up for a project:

I'm not ready to start yet tho. I guess in about 8 months to a year. I wanna watch the market for awhile longer. I keep waiting for something totally new to arise. Like quantum computing or something. :D

I guess I can make something that will run completely silent and still be very cool. ;)



WOW!

2 or 3 of those are very beautiful, thanks for posting them up :)
 
Yeah, the bottom left two I've read about, that guy does amazing work.

I also like that pizza-box shaped one.

Tess - you wouldn't happen to be into steampunk now would you? :)
 
I think this is not so good. ;) Even if things don't become unstable you essentially have two little heaters in there heating up the rest of the system. I guess if you're exhaust fans aren't
drawing this heat away efficiently then it'll add 5 or 10 degrees to some of the other components.

Is your ambient temperature higher than about 5 degrees above room temperature? It shouldnot be if it is. If any of these things bug you just install scmFanControl and mess around with it till you're comfortable with the noise to temperature ratio and save it as a preset called idle.
Then maybe create another preset with the fans at a slightly faster speed (+500?) and call it Working. Then you can select between the two as needed.

Thanks Tesselator - I've installed smcFanControl on the 2008 3.2 8-core and have set the CPU/MEM fan at 900RPM, and the remaining at 800RPM. Now, I see that most of the rams are hovering around 55 degrees celsius. (Out of the 8, I see 2 staying at 50 and another one staying at 58).

I believe the default fan speed without smcFanControl sets the CPU/MEM at 600RPM. So far... 600RPM vs 900RPM fan noise isn't bothering much and good seeing the ram staying much cooler than before hovering around 70-75 doing nothing!!!

The other interesting thing is that the ram (6x1gb) on the 2009 MP 2.66 is hovering between 38-42 degrees celsius with the default fan settings without running smcFanControl.

Something terribly wrong with the 3.2? or is it the damn third-party ram installed there? Yes, the 3.2 is using all 8 slots vs 6 slots on the 2.66. Any further input?
 
Yeah, if you want it in a Mac case it seems like a bit of a project. ;) I dunno tho... I might just build my own case. I was a cabinet maker in a former life and something like below looks pretty sweet to me!
...
2056828740_ff3f1afcff_o.jpg
...
I'm not ready to start yet tho. I guess in about 8 months to a year. I wanna watch the market for awhile longer. I keep waiting for something totally new to arise. Like quantum computing or something. :D

I guess I can make something that will run completely silent and still be very cool. ;)

Who needs quantum computing when you can be running applications to unfold proteins and having coffeea at the same time?! Best part; spill-proof. So many laptops have suffered that horrible fate ;):D
 
Sounds good James. Looks like you've found a solution that works.

Pure speculation (of course) but I guess the hotter RAM is running hotter just because of it's spec and manufacturer. I would guess that if you look up the detailed specs you'll maybe notice some differences. :) Could be anything tho really... including where you have the machines placed in the room, etc. I can shave 5°C off my temps just by cracking the door nearest the machine open a foot or so. ;)


.
 
The other interesting thing is that the ram (6x1gb) on the 2009 MP 2.66 is hovering between 38-42 degrees celsius with the default fan settings without running smcFanControl.

Something terribly wrong with the 3.2? or is it the damn third-party ram installed there? Yes, the 3.2 is using all 8 slots vs 6 slots on the 2.66. Any further input?

The discrepancy comes from the fact that the RAM in the 2009 mac pro are completely different than the RAM in your 2008 machine. The RAM in your 2008 machine run very very hot compared to normal RAM. Nothing to be concerned about, it's just the way it is.
 
I believe the default fan speed without smcFanControl sets the CPU/MEM at 600RPM. So far... 600RPM vs 900RPM fan noise isn't bothering much and good seeing the ram staying much cooler than before hovering around 70-75 doing nothing!!!

Wow, in smcFanControl on my 2009, it won't let me set the CPU fan less then 1516rpms...you can set yours at 600? Everything else is 600 aside from boosta which is 800, but cpu is 1516...anything less and I get a "fell short of minimum" warning.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.