Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
MP All Models Mac Pro CPU Compatibility List
- Thread starter ActionableMango
- WikiPost WikiPost
- Start date
- Sort by reaction score
- Status
has anyone ever tried a Core i7 995X in a single CPU Mac Pro? 3.6Ghz 6 Core beast of a LGA1366 CPU heh (i didn't even know it existed till i came across a CPU Z shot of one)
That CPU is not listed in Intel ARK.
I searched around for information and it appears that Intel planned it and made some, but never released it. It sounds like at best you might be able to find an Engineering Sample and I wouldn't have any confidence that it would work in a MP.
Just figured id mention it as i saw it come up on CPU Zs site recent validations (I check that page for odd systems that show up 
) http://valid.x86.fr/dlb8st
) http://valid.x86.fr/dlb8st
Just installed an x5677 in a 2009 MP 4,1. Upgraded firmware to 5,1 and installed 16gb of 1333MHz ram. Runs great. Cinebench score for the cpu is 552. I didn't run it with the old 2.66ghz processor so I can't say how much of an improvement it is,, but it feels quicker. I used Arctic MX-4 thermal paste. Temperatures sit a around 92ºF at idle, and when running cinebench never got hotter than 124ºF. The GPU card is the stock NVIDIA GeForce GT 120 and only managed 12.61 fps in cinebencz. The x5677 offers the highest possible clock speed in a cMP for a bargain basement price of only around $50. You sacrifice two cores in not going with the 5690, but the 5690 currently sells for almost $260 on eBay.
Last edited:
Can someone with a 4,1 and x5680/90 let me know their CPU temps? I'm getting 49/39 Celsius and up to 74/70 at high load in logic, its winter right now and I'm worried this will go above 80 in summer. I've read that max temp for these CPUS is 78 degrees.
When I installed the CPUs I used some washers that where a bit taller than the 2mm that was recommended. I think it was 2.1-2mm. Should I reinstall removing some washers and go 1.8mm?
-Thanks
When I installed the CPUs I used some washers that where a bit taller than the 2mm that was recommended. I think it was 2.1-2mm. Should I reinstall removing some washers and go 1.8mm?
-Thanks
Your temperature is normal. If you still worry about that, simply install Macs Fan Control to create your own fan profile, or spin up the fan to a higher speed.
That 78 is Tcase max temperature (a temperature that you can't measure in the cMP), not the max Tdiode temperature (the temperature that softwares can read from the CPU), don't worry about that. My W3690's Tcase max is 68, and it can run at 85 Tdiode 24/7 without thermal throttling.
For 74C, the cMP may not even spin up the fan, because it's still quite far away from thermal throttling, the Apple engineers decided that you prefer a quieter Mac Pro, rather than a cooler CPU. So, it seems that they programmed the fan stay at idle until the CPU reach about 80C, then spin up the fan to make it stay at around 85C.
At least this is the native fan behaviour on my cMP with the W3690. If my 68 Tcase max W3690 can survive with this fan setting and temperature, your 78 Tcase max X5690 should not be a problem at all.
REcently was able to purhcase a twin processor 2010 model to future proof myself for a while from my already reasonable w3690 upgraded single core 2010 model.
Super happy with the purchase and the price! The first thing I did was buy two x5690's.
I ran benchmarks on:
My w3690 single core with 32gb ram (about 12700 benchmark geekbench4)
2.4 ghz twin processor 32 gb ram (about 12666 benchmark) (with 5870 graphics, but not sure if a factor in CPU testing?)
x5690 twin processor 32 gb ram.
The benchmark on this was "only" 19879.
I understand that benchmarks are multifactorial, but I thought there should be a bigger difference between the dual CPU models.
My last question would be:
Is there a way to assess the quality of the other hardware - motherboard fan's etc. I would prefer to keep the better condition or newer machine if possible. My machine has been faultless and cleaned reasonably regularly, so perhaps its better the devil you know??
Super happy with the purchase and the price! The first thing I did was buy two x5690's.
I ran benchmarks on:
My w3690 single core with 32gb ram (about 12700 benchmark geekbench4)
2.4 ghz twin processor 32 gb ram (about 12666 benchmark) (with 5870 graphics, but not sure if a factor in CPU testing?)
x5690 twin processor 32 gb ram.
The benchmark on this was "only" 19879.
I understand that benchmarks are multifactorial, but I thought there should be a bigger difference between the dual CPU models.
My last question would be:
Is there a way to assess the quality of the other hardware - motherboard fan's etc. I would prefer to keep the better condition or newer machine if possible. My machine has been faultless and cleaned reasonably regularly, so perhaps its better the devil you know??
IMO, Logic board is hard to tell, just pick the one that looks better (clearner, the copper still looks golden yellow but not dark brown, etc.)
For the fan, you may develop your own test. e.g. Place a dB meter at a constant distance, spin up the fan to a constant value to measure the noise. Or slowly spin up the fan to a value that your ear can notice the difference. Then pick the lower noise / lower RPM one.
Or just check the manufacture period, and then pick the newer one.
There should be quite a few different objective way to help you to choose the "better" one.
The difference between the dual CPU models is directly related to CPU speed. The X5690 at 3.46GHz is about 50% more than the 2.4GHz model hence the roughly 50% increase in benchmark figures.
Yep, in fact there are services on Ebay and from OWC for doing exactly that and I believe a few people here have already done so.
Nice. I was thinking of running a hex-core i7 for gaming and windows, and maybe a lower clock rate pair of Xeons for photoshop and daily use.
From my understanding those KP were/are related to stepping and Microcode updates in the CPU. The Mac firmware update would also update the Microcode in the Xeons with older steppings (SLAC4, SL9YM), so they'd work fine afterwards. The latest stepping (SLAEG / G0) would already include those Microcode updates and thus run fine with a MP1,1 firmware (confirmed by own experience). Besides, the stepping is also important for idle power consumption and thus the need for manual fan speed control (or the absence thereof), see my old post. Thus I propose to add stepping information in the MP 1,1/2,1 section as well.
Apart from that: Wonderful idea and well done - thanks!
Photoshop benefit more from clock speed than core count. The 2nd Xeon should be just idling for most of the time.
For normal daily use, a Hex core CPU (with HT) should be more than enough.
Hi everyone!
I have lurked these forums MANY times before. They were great in so far keeping our 2007 mac pro still working and highly upgraded as far as the video card goes.
Anyway, I am thinking of whether to get a 2012 5,1 model and upgrade it or buy a whole new computer (it would then not be a mac because the can is the least pro computer ever). I was looking for some information in geekbench, and came accross something that is really interesting: someone posted both a 3,1 and 5,1 score... with an i7 6700K!!! Here is a link and a couple of screenshots.
https://browser.primatelabs.com/v4/cpu/489074
https://browser.primatelabs.com/v4/cpu/1378300
Anyone here has any idea if this is possible? And even more so, how to do it? His geekbench score shows clearly he is indeed using a mac 5,1 motherboard.
Thanks!
I have lurked these forums MANY times before. They were great in so far keeping our 2007 mac pro still working and highly upgraded as far as the video card goes.
Anyway, I am thinking of whether to get a 2012 5,1 model and upgrade it or buy a whole new computer (it would then not be a mac because the can is the least pro computer ever). I was looking for some information in geekbench, and came accross something that is really interesting: someone posted both a 3,1 and 5,1 score... with an i7 6700K!!! Here is a link and a couple of screenshots.
https://browser.primatelabs.com/v4/cpu/489074
https://browser.primatelabs.com/v4/cpu/1378300
Anyone here has any idea if this is possible? And even more so, how to do it? His geekbench score shows clearly he is indeed using a mac 5,1 motherboard.
Thanks!
Attachments
I think what you are seeing could be a hackintosh. I could be wrong, but I think part of getting OSX to work on non Apple hardware is to trick it into thinking that it sees an Apple motherboard.
Not possible. People can spoof the information on their computers that gets sent to Geekbench.
That's just what GeekBench reads off of the computer / SMBios.
Like the photo below is mine.
The Hack is the "iMac 14,2" and this is just the compute test on GeekBench4.
You can see how a GTX980Ti compares to my MacPro D500, 5k iMac M395X and new tMBP with Intel550.
Can also see even with zero overclocking the 6700k in the Hack scores a good 2,000 more than the 6700k in the latest iMac 5k, which scores slightly more than the 6c nMP.
Like the photo below is mine.
The Hack is the "iMac 14,2" and this is just the compute test on GeekBench4.
You can see how a GTX980Ti compares to my MacPro D500, 5k iMac M395X and new tMBP with Intel550.
Can also see even with zero overclocking the 6700k in the Hack scores a good 2,000 more than the 6700k in the latest iMac 5k, which scores slightly more than the 6c nMP.
Attachments
Does the 6 core Dual Xeon X5690 work in my stock single processor 6 core 3.33 5,1 Mac Pro? I see install services only offering x3690 processors. Is one more compatable than the other?