Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Both way works, however, IMO, de-lid is the e better way to do it, less risk to damage the CPU tray. Even though higher risk to damage the GPU and cost a bit more, but CPU is actually cheaper and easier to replace than the CPU tray now.

Hi thanks, that's what I was thinking too. Have you de-lidded your own? do you know any good sites that show you how to de-lid them?
 
Hi thanks, that's what I was thinking too. Have you de-lidded your own? do you know any good sites that show you how to de-lid them?

You're going to find opinions on both sides of this question. I read some credible opinions that the lid actually HELPS heat dissipation while others say it inhibits it.

I've watched several of the de-lidding videos on Youtube and it looks like a pretty scary process, much more risky (to me) than carefully installing with the lid on, which I did successfully on the first try. The skill required for successful de-lidding seems greater than that for finding the correct tightening tension of the heatsink over the lid. (The fan bracket clipping was pretty trivial.)

It's also not clear to me that - even with the lid off - you don't still need to be careful about how far you tighten down the heatsink. AFAIK there is no "stop" against which you can just crank it down tight. I DID add washers to the standoffs as a final "emergency stop" before I installed the heatsink over the lidded processors.

That said, if you are consistently running your CPUs max'd out for 3D rendering or video transcoding there might be some advantage in the greater direct contact between the heatsink and the die, but in conventional use I haven't even had the CPU fans ramp up.
 
Hi thanks, that's what I was thinking too. Have you de-lidded your own? do you know any good sites that show you how to de-lid them?

You can check out this eBay seller who offers De-Lidding for Xeon CPUs specifically for 2009 Mac Pros http://www.ebay.com/itm/171745993685 Some members here already tried his services and were satisfied with his work. He is also offering De-lidded CPUs. Better to have a professional do the work with tech support.
 
You're going to find opinions on both sides of this question. I read some credible opinions that the lid actually HELPS heat dissipation while others say it inhibits it.

I've watched several of the de-lidding videos and it looks like a pretty scary process, much more risky (to me) than carefully installing with the lid on, which I did successfully the first time. The skill required for successful de-lidding seems greater than that for finding the correct tightening tension of the heatsink over the lid. (The fan bracket clipping was pretty trivial.)

It's also not clear to me that - even with the lid off - you don't still need to be careful about how far you tighten down the heatsink. AFAIK there is no "stop" against which you can just crank it down tight. I DID add washers to the standoffs as a final "emergency stop" before I installed the heatsink over the lidded processors.

That said, if you are consistently running your CPUs max'd out for 3D rendering or video transcoding there might be some advantage in the greater direct contact between the heatsink and the die, but in conventional use I haven't even had the CPU fans ramp up.


Good to hear both sides of the story! To be honest I've never heard of anybody de-lidding a CPU before, which is why I was interested in hearing how easy/hard it is and the success-rate/results.

I've installed straight CPU's in a mac in the past, so doing it that way should be fine. I was wondering about lidded temps as well. So you don't notice your machine getting hotter than what it should?

What kind of washers did you use? just standard hardware store ones? Was there a guide you followed at all?

----------

You can check out this eBay seller who offers De-Lidding for Xeon CPUs specifically for 2009 Mac Pros http://www.ebay.com/itm/171745993685 Some members here already tried his services and were satisfied with his work. He is also offering De-lidded CPUs. Better to have a professional do the work with tech support.

Good to know, and will keep it in mind! Only thing is that's almost the same price I'm getting the CPU's for ($140 each), and as I live in Australia with postage cost to US and back, it would easily double my costs and not be worthwhile.

Maybe I can try and get good at it and offer the service in Aus :)
 
Problems with dual 130w CPU at MacPro 4.1 & 5.1?

I'm also considering about an upgrade to dual X5690!
The current config of my MP4.1 is:

MP4.1 flashed to 5.1
48 GB 1333Mhz RAM
2 x 2,93Ghz
GTX980
3 x 4TB Harddrives
1 SSD 512GB
PCIe 1TB SSD raid, Sonnet tempo
2 x optical drives.

Does anyone know the exact specification of the power supply?
Or had anyone of you problems with a dual 130W CPU, like X5690?

I don't want to burn my Mac ;-)

Regards
Zodi
 
Power Supply (614-0454, 661-5449)

I found this in the web:

Apple Mac Pro 2009/2010/2012
Delta DPS-980BB-2


12,2 Volt x 79 Ampere = 964W for the 12V line.
5 Volt x 5 Ampere = 25W for the 5V line.

is it maybe possible to measure the Amps in my current system?
with a nice app?
 

Attachments

  • $_57-1.JPG
    $_57-1.JPG
    151.3 KB · Views: 207
Hi all,

I finally have made the plunge and got my 2012 Mac Pro upgraded to 2x3.46 6 core (total 12 core) today and so far it seems to run nice.

But I am curious as to whether the temperature is running fine.

The CPU is seating at about 90% idle and I am running some rendering which seems to be offloading most of the work to the GPUs. But the temperature of the IOH Heatsink (What is that?) is highlighted red. Is this normal?

The settings are set to Fahrenheit for you guys to see (I assume most of you guys are in the US). Thanks for all those who have shared their information on this site.

https://forums.macrumors.com/attachment.php?attachmentid=524657&stc=1&d=1421160829

https://forums.macrumors.com/attachment.php?attachmentid=524658&stc=1&d=1421160829

Which software you use?
 
MacPro 4.1 heat test

Handbrake at work 1350% Cpu and Valley benchmark together, see
the measurements PSU 12V 33 Amp !
 

Attachments

  • heat test.jpg
    heat test.jpg
    358.3 KB · Views: 295
Have a matched pair of x5680's on the way! Was wondering about thermal paste and what people recommend these days?

I still have some arctic silver somewhere, but have been reading about Liquid Pro which is supposed to be the best, anyone tried it?

http://www.coollaboratory.com/en/products/liquid-pro/

Anyone got any thoughts on Liquid Pro? anyone tried it? or should I stick with something like Arctic Silver? what are people using?
 
Anyone got any thoughts on Liquid Pro? anyone tried it? or should I stick with something like Arctic Silver? what are people using?

I don't think anyone who uses normal, quality paste has complained about high CPU temps, unless they did not apply it correctly.

Not to say you may get a few degrees cooler with the expensive Liquid Pro, but it's not really necessary. With dual x5680's, my idle temps are around 36-40C.
 
Last edited:
Handbrake at work 1350% Cpu and Valley benchmark together, see
the measurements PSU 12V 33 Amp !

I made a little bit harder test to show the max out, but with x5680.

MacPro 5,1 2x3,33 GHz with GTX 980

Used: CPULoad (2350 %) and Furmark together

attachment.php


With a 680 GTX i got up to 40 Amp
 

Attachments

  • Bildschirmfoto 2015-05-10 um 17.51.47.png
    Bildschirmfoto 2015-05-10 um 17.51.47.png
    99.6 KB · Views: 1,065
I installed X5675 and 1333MHz RAM

The new Cpu are running well and silence.
From the power supply point of view i've the feeling the X5690 are also not a
problem for the Mac Pro 2009, because my new 6 core 3.06Ghz CPU's need less power than the 2.93GHz CPU's before.

Left: old CPU right: new CPU

also interesting is the fan speed !
 

Attachments

  • heat test.jpg
    heat test.jpg
    358.3 KB · Views: 267
  • Cinebench.jpg
    Cinebench.jpg
    1.3 MB · Views: 283
also interesting is the fan speed !

Good to see that your new CPU works well.

However, if you really want to compare it to the old system. You need to run them in the same ambient condition. Otherwise, the temperature / fans speed comparison is meaningless.

Also, it takes time to warm up the CPU. You may run handbrake + another CPU stress software together for about 15min to achieve the peak temperature / fan speed.
 
The new Cpu are running well and silence.
From the power supply point of view i've the feeling the X5690 are also not a
problem for the Mac Pro 2009, because my new 6 core 3.06Ghz CPU's need less power than the 2.93GHz CPU's before.

Left: old CPU right: new CPU

also interesting is the fan speed !

Great to see! I'm about to do a very similar upgrade.

What thermal paste did you use?

Also, did you de-lid your cpu's?
 
Good to see that your new CPU works well.

However, if you really want to compare it to the old system. You need to run them in the same ambient condition. Otherwise, the temperature / fans speed comparison is meaningless.

Also, it takes time to warm up the CPU. You may run handbrake + another CPU stress software together for about 15min to achieve the peak temperature / fan speed.

Ok, you are right I'll do it today!

I also tested my Mac with Win7 / 3D Mark and it works well.
faster than 97% of all PC's :)

----------

Great to see! I'm about to do a very similar upgrade.

What thermal paste did you use?

Also, did you de-lid your cpu's?

I ordered this kit:

https://www.dncomputers.com/parts-accessories/processors/mac-pro-2009-4-1-processor-upgrade-kit-to-12-core-3-06ghz-xeon-x5675.html
 
Good to see that your new CPU works well.

However, if you really want to compare it to the old system. You need to run them in the same ambient condition. Otherwise, the temperature / fans speed comparison is meaningless.

Also, it takes time to warm up the CPU. You may run handbrake + another CPU stress software together for about 15min to achieve the peak temperature / fan speed.

The temp and fan speed is similar to my old CPU, but the conversion
speed in Handbreak is 80% higher :)
 

Attachments

  • X5675 Handbreak & Valley.jpg
    X5675 Handbreak & Valley.jpg
    1.8 MB · Views: 241
Hey guys, this may be a little off topic. I've completed the upgrade of my 2009 Mac Pro with 2x Six-Core 3.33 GHz with RAID 0 SSD, with GTX 680 GPU and 48GB of RAM. Geekbench score of around 30,000 multi-core 64bit.

My question is regarding running Windows 7. Would you rather install Windows via Bootcamp, or VM such as Parallels? In my 2011 Macbook Pro, I used Bootcamp so I wouldn't split system resources. But now that I have a beast of a machine, would it just make sense to install Windows via VM?

I plan on running FL Studio with VSTs, and probably install my Steam games such as Street Fighter IV, Payday, Beyond Earth.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.