I want to know is it worth moving my Mac Pro 2.8GHZ from - to
Intel
Xeon Processor
2.4GHz 6-Core E5645 ($21.00)?
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
Intel
Xeon Processor
2.4GHz 6-Core E5645 ($21.00)?
Probably not. That's a slower chip with a couple more cores. Almost everything benefits from single-threaded performance, so almost everything will be slower.
The rare software that makes good use of 6 cores will benefit from the extra 2 cores, but even then it will still be affected by the slower speed.
Personally I wouldn't bother. I think the purpose of that chip is the lower power use.
So the itch to upgrade is upon me, but I’m not sure it’s a noticeable upgrade I’m considering.
My specs (for the relevant part) are in my signature. I’m wondering if there is an observable increase in performance (without benchmarking tools) by replacing my cpu with an X5690.
I do very little video work, but I enjoy having a responsive system. Seriously, I don’t care, that I don’t need more performance for my mundane tasks. I just want to know if I would even be able to feel the difference.
My specs (for the relevant part) are in my signature. I’m wondering if there is an observable increase in performance (without benchmarking tools) by replacing my cpu with an X5690.
I do very little video work, but I enjoy having a responsive system. Seriously, I don’t care, that I don’t need more performance for my mundane tasks. I just want to know if I would even be able to feel the difference.
99% sure you can't feel it.
If you can feel that little difference. That means even with your current CPU, when you start to use more cores, the CPU just go one step down on the turbo boost. And you can feel the system responsiveness difference straight away.
To add to this, I updated a 4-core 2.93 to a 6-core 3.33. No "feelable" difference. Transcodes video faster though.
Last edited:
Thanks, if my math is correct I would gain only a 3.91% anyway, so yeah... prolly a wasted effort for me.
Well, from my experience. I started out with dual E5620s (2.4Ghz, 8 core) went to W5590s (3.33GHz, 8 core) then X5677 (3.46 GHz, 8 core). With each upgrade I could feel the difference. Maybe just in my head, but I don't think so!
Lou
Lou
That's the speed difference without turboboost. Only applicable when all cores are used (e.g. video encoding). For other normal use, you should compare the turboboosted frequency. So, the difference is even less.
Just a quick post to thank everyone who's contributed to this thread/forum. Upgraded yesterday (stock 2.66 to 3.46), went smooth as butter. Cheers
Do you notice any increase in fan speed (due to increased heat)?
Nope, if anything the machine runs cooler. Makes sense considering the stock cpu had never been touched. A good cleaning and new thermal paste is never a bad idea.
Edit: I also made some minor adjustments in the pci-e and psu fan speeds via macs fan control app. Some interesting info regarding temps and non-efi gpus in mac pros. Link:
https://forums.macrumors.com/thread...when-using-pc-non-efi-graphics-cards.1946672/
Last edited:
Ya can add the X5350 (qSpec is QUQD IIRC) to the Xserve1,1 list now if you want 3.33Ghz MacPro2,1 here I Come! (heres a GeekBench @ 2.66Ghz https://browser.geekbench.com/v4/cpu/8917013 note the score might be a bit lower then it would be, as I was running with no GPU acceleration so I had UI drawing overhead on the CPUs)
![upload_2018-7-6_16-44-27.png]( "upload_2018-7-6_16-44-27.png")
3.33Ghz MacPro2,1 here I Come! (heres a GeekBench @ 2.66Ghz https://browser.geekbench.com/v4/cpu/8917013 note the score might be a bit lower then it would be, as I was running with no GPU acceleration so I had UI drawing overhead on the CPUs)
You can add it too.
It's a wikipost. Anyone can add their own entries and/or help maintain it and improve it.
You can add it too.
It's a wikipost. Anyone can add their own entries and/or help maintain it and improve it.
alright I popped it in there hopefully without breaking anything... (I just copied and pasted the X5355 section and edited a few values...)
BTW can someone please remove the RAM section on the Pre Nehalem CPU sections?
those CPUs dont have any sort of control over the speed the RAM runs at as the memory controller is in the chipset not on the CPU
I recommend the RAM section on Pre Nehalem CPUs be replaced by a section on what speed the FSB of each CPU runs at that would be much more useful
I would do it, but I dont know BBCode tables for love-nor-money so I dont wanna balls anything up...
Hope someone can help.
I have a MP 1,1 flashed to 2,1 with x2 5150's working fine.
I tried installing a pair of E5345's (SLAEJ) but the mac will not boot.
With both CPU's installed, the CPU B error LED lights up. With one CPU installed, no CPU LED's light up but it still won't boot. The trickle and power LED's always light up but the GPU LED flashes with these CPU's
Any ideas would be greatly appreciated.
I have a MP 1,1 flashed to 2,1 with x2 5150's working fine.
I tried installing a pair of E5345's (SLAEJ) but the mac will not boot.
With both CPU's installed, the CPU B error LED lights up. With one CPU installed, no CPU LED's light up but it still won't boot. The trickle and power LED's always light up but the GPU LED flashes with these CPU's
Any ideas would be greatly appreciated.
the E5345 should Pass POST, OOB even if running stock 1,1 Firmware so one of your CPUs might be bad...
what I would do is test each CPU on its own first to figure out which one is bad (I Would also clean the contact pads on each CPU too with something like some Isopropyl)
its worth mentioning that on a MacPro1,1 the socket marked "Socket B" on the PCB is actually Socket 1/the main socket so if you running a 1,1 with Just 1 CPU make sure you have Just socket B populated (if you have just 1 CPU in socket A then the machine wont POST)
Thanks for the response
Interesting. So am I right in assuming the DIAG LED's still match to the corresponding CPU sockets?
FAIL-CPUB = socket B = main socket
FAIL-CPUA = socket A = second socket
Interesting. So am I right in assuming the DIAG LED's still match to the corresponding CPU sockets?
FAIL-CPUB = socket B = main socket
FAIL-CPUA = socket A = second socket
Just checked both CPU's in socket B and neither work. Is it in any way possible that a board can accept a dual core but not a quad core? Like a bad northbridge for example?
I have no other way to test these CPU's so i've bought another pair on ebay.
I have no other way to test these CPU's so i've bought another pair on ebay.
I've tried different quad cores and they won't work either so I conclude I have a board that won't accept quads!
^^^^Do a search it's been done a LOT!!!!!
Here's just one:
https://forums.macrumors.com/threads/mac-pro-5-1-2012-still-worth-it-in-2018.2125663/
Do some reading!
Lou
Here's just one:
https://forums.macrumors.com/threads/mac-pro-5-1-2012-still-worth-it-in-2018.2125663/
Do some reading!
Lou
Will the Mac Pro 3,1 work if I install two (2) L5408 (C0-SLAP5) in it, assuming I could find them? Electricity is very expensive here.
they should
but if low power is the goal then installing just 1 of these http://www.cpu-world.com/sspec/SL/SLANM.html will get you down to a 35W TDP at a great performance penalty of course (there is the L5215 as well with a 20W TDP however this CPU does not have a C0 counter part sadly)
however its worth noting the MP3,1 as a platform ignoring CPUs draws quite a lot of power
you may be better off with a low power Mac Mini or if you need a Mac Pro find a single CPU MP4,1/5,1 and kit it out with DDR3L and a low power CPU