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 1,1-5,1 Strange Audio Issue on MP 4,1>5,1 Mojave 10.14.4
- Thread starter rocksarntpeople
- Start date
- Sort by reaction score
[doublepost=1554812486][/doublepost]
Is this dual E5520?
[doublepost=1554812562][/doublepost]
Same E5520?
[doublepost=1554812946][/doublepost]
Let's see the reports above. If so, there's one common theme: Gainestown processors. Believe only officially available in MP4,1 were E5520, X5550, and X5570. All others were upgrades to Bloomfield, which does not seem to be impacted and was officially an available processor option on MP5,1.
Why this presents in 10.14.4 and not 10.14.3 may be another issue...
Is this dual E5520?
[doublepost=1554812562][/doublepost]
Same E5520?
[doublepost=1554812946][/doublepost]
Let's see the reports above. If so, there's one common theme: Gainestown processors. Believe only officially available in MP4,1 were E5520, X5550, and X5570. All others were upgrades to Bloomfield, which does not seem to be impacted and was officially an available processor option on MP5,1.
Why this presents in 10.14.4 and not 10.14.3 may be another issue...
Last edited:
Welcome to the "Stuttering Audio In Mojave Club"
Already three of us with this issue.
Statistics:
2 cMP with Dual 2.26 Xeon - 1 with 2.66 Xeon
2 cMP with RX580 - 1 with nVidia GTX680
total amount of installed RAM is different
different SSD, HDD disks, connected via SATA or PCIe
sound sources are different too - built-in soundcard, additional pci-e ESI Maya44 XTe soundcard, USB DAC
We need to understand what is common in these three cases.
Last edited:
Looking like it might be Gainestown processor issue. Waiting for confirmation from the user reports.
[doublepost=1554814303][/doublepost]Open Terminal and enter:
Code:
sysctl -n machdep.cpu.brand_stringThis will tell you the type of processor being used.
Yes.
What about Xeon 2.93? is it Gainestown or Bloomfield? At previous page user MIKX reported he has no issues, i see 2.93 CPU in his signature.
As i see in the link, all Xeon's in cMP 4.1 are Nehalem.
Are Gainestown and Bloomfield is sort of subname or something so?
https://everymac.com/ultimate-mac-lookup/?search_keywords=MacPro4,1
AFAIK, there are 2.93 as both X5570 (Gainestown) and X5670 (Gulftown). Need report to which is being used.
Can 100% confirm this issue does not present with dual X5690 (2 x 3.47GHz) in authentic MP5,1. Originally was single tray W3680 (Gulftown) that I can swap in if further testing becomes necessary.
[doublepost=1554815367][/doublepost]
Take a look at this, likely helps a lot:
https://en.wikichip.org/wiki/intel/xeon
Sorry for not replying back to this, family emergency this weekend kind of put a damper on testing. I can tell you that yes, it appears the issues are with 4,1-5,1 updates with older Nehalem chips (as thats what I have). I will be able to isolate that tomorrow as my new Westmeres arrive and will be installed tomorrow evening.
Also I can say it is not a RAM issue (not that I thought that was the case either) as I have since installed 96 GB's of new RAM and nothing has changed (save for some bigger benchmark numbers)
As for the 4,1-5,1 update method, I followed this guide specifically: https://www.ifixit.com/Guide/How+to+Upgrade+the+Firmware+of+a+2009+Mac+Pro+41/98985
But I did verify this against various other "forum" guides and the methodology and software required is identical throughout so I believe this this be the supported method regardless.
As for Sierra - High Sierra - to Mojave, I followed this: https://forums.macrumors.com/thread...-bootrom-upgrade-instructions-thread.2142418/ and referenced the official apple specifications and upgrade recommendations.
While I dont know how others in this thread have done it, thats what I did and reading through lots of posts seems to be the accepted method for upgrading.
Also I can say it is not a RAM issue (not that I thought that was the case either) as I have since installed 96 GB's of new RAM and nothing has changed (save for some bigger benchmark numbers)
As for the 4,1-5,1 update method, I followed this guide specifically: https://www.ifixit.com/Guide/How+to+Upgrade+the+Firmware+of+a+2009+Mac+Pro+41/98985
But I did verify this against various other "forum" guides and the methodology and software required is identical throughout so I believe this this be the supported method regardless.
As for Sierra - High Sierra - to Mojave, I followed this: https://forums.macrumors.com/thread...-bootrom-upgrade-instructions-thread.2142418/ and referenced the official apple specifications and upgrade recommendations.
While I dont know how others in this thread have done it, thats what I did and reading through lots of posts seems to be the accepted method for upgrading.
Last edited by a moderator:
@rocksarntpeople - it may be helpful for others to change the title of this post to "Strange Audio Issue on MP4,1>MP5,1 Mojave 10.14.4". Your machine ID's as MP5,1 after the hack/update, but it is not a true MP5,1 machine.
----
Open Terminal and enter:
This will tell you the type of processor being used.
This may not be a Nehalem issue, but isolated to the few CPUs that were used in MP4,1 stock configurations that were not carried into MP5,1 stock configurations. This includes E5520, X5550, and X5570.
----
Open Terminal and enter:
Code:
sysctl -n machdep.cpu.brand_stringThis may not be a Nehalem issue, but isolated to the few CPUs that were used in MP4,1 stock configurations that were not carried into MP5,1 stock configurations. This includes E5520, X5550, and X5570.
Title changed as requested. Thanks.
[doublepost=1554816256][/doublepost]I believe that the issue stems from the processors being used as the base hardware of the 4,1 and 5,1 are identical save for the chips and the firmware lock (hence being able to swap procs, etc between them).
To my knowledge, Mojave is only supported on 5,1 machines and therefore only the newer Westmere chips.
Again, just speculation, but I will be able to fully test this theory tomorrow and report back if the problem goes away.
[doublepost=1554816415][/doublepost]Not currently in front of the machine, will check when I get home, but here is the serial number lookup with the processor data:
CPU speed: - 2 x 2.26 GHz
CPU Type: - Intel Xeon (Nehalem)
CPU Model: - E5520
Cores: - 4 per CPU.
Bus Speed: - Up to 3.2GHz (QPI)
Cache: - 8 MB L3 cache Per CPU
64-bit: - Yes
Turbo Boost: - Up to 2.53 GHz
Mid-2010 basic single CPU quad core 2.93GHz is Nehalem, have one in front of me now, other models are Nehalem too.
Various MP5,1 models came from factory with Nehalem Xeons. Only Mac Pros with hexa cores you can say for sure that the CPU is Westmere.
So that is two confirmed reports of the E5520 processor presenting this audio issue. Suspect the issue will impact E5520, X5550, and X5570.
If any additional reports, please confirm your CPU model.
As bsbeamer stated above though, the 2.26 (E5520) was never carried into the 5,1 production line so perhaps that is the missing piece of the puzzle? Again just speculation, but could that also mean potential problems for ANY CPUs not officially supported during production, ie 5680,5690s?
Afaik, according the links 2.93 in MP 4.1 could be single W3540 Bloomfield or dual X5570 Gainestown. But in signature of user MIKX i see Dual Quad Core 2.93. So high probability they are X5570 Gainestown. But he reported yesterday there no audio issue in his machine. It would be great if MIKX would inform us with more detailed information about CPUs.
https://everymac.com/systems/apple/...-quad-core-2.93-early-2009-nehalem-specs.html
https://everymac.com/systems/apple/...eight-core-2.93-early-2009-nehalem-specs.html
CPUs in MP 5.1 are dual six core X5670 Westmere.
[doublepost=1554819755][/doublepost]
If you look at this list, you will see that none of the 4.1 CPUs was never carried into 5.1, not just 2.26 or 2.66...
bsbeamer's conclusions are more concerned with Gainestown vs Bloomfield (both are Nehalem).
https://everymac.com/systems/apple/mac_pro/index-macpro.html
[doublepost=1554820409][/doublepost]
We are trying to locate the problem, please kindly confirm your CPU model.
Enter in terminal this code:
sysctl -n machdep.cpu.brand_string
Thank You!
Last edited:
Due to the nature of the issue and trying to troubleshoot, let's be 100% clear when referring to CPUs in this thread. Let's not speculate about certain machines/users until those users can confirm which CPU model they are using in their machines.
AFAIK, there are 2.93 as both X5570 (Gainestown) and X5670 (Gulftown). For this particular troubleshoot, it makes a big difference.
W3520, W3540, W3580, E5520, X5550, and X5570 were stock configs for MP4,1 (Bloomfield & Gainestown)
W3530, W3565, W3680, E5620, X5650, and X5670 were stock configs for MP5,1 2010 (Bloomfield & Gulftown)
W3565, W3680, E5645, X5650, and X5675 were stock configs for MP5,1 2012 (Bloomfield, Gulftown, Westmere-EP)
E5520, X5550, and X5570 are Intel Xeon 5500 series (Gainestown) and only used in MP4,1.
W3520, W3540, and W3580 are Intel Xeon 3500 series (Bloomfield). This series was also used in MP5,1 2010 and 2012.
No issues with X5690 in my 5,1 machine right now with 10.14.4 18E226.
I understood.
But in that list i see only two different 2.93 in MP 4.1 (not 5.1): single W3540 Bloomfield or dual X5570 Gainestown, but no X5670.
Anyway, hope MIKX will report.
Last edited:
That list above is original, OEM, STOCK CONFIG. Several users have upgraded/updated their CPUs from stock configs since original purchases in 2009/2010/2012. Just because the machine was originally an MP4,1 does not mean it is still using MP4,1 issued CPUs or MP4,1 generation/series. Many have upgraded to MP5,1 series CPUs via various methods.
There are posts from @MIKX on this forum showing previous CPU and/or tray issues, so it is best not to assume anything until it is confirmed. The user was reporting no audio issues on their machine in post #19.
There are now numerous reports of 2 x E5520 with this audio issue. Can debate the cause and/or possible solutions, but it's enough reports to indicate there is an issue with at least Mojave 10.14.4 and this processor. Would caution anyone using E5520 to stay away from 10.14.4 18E226 at this time.
If ALL of the reports can easily correct this with a downgrade to Mojave 10.14.3, then it may indicate a software change that was part of the HDMI/DP audio fixes for certain systems/GPUs that was implemented as part of the 10.14.4 update.
Also assume a CPU swap to supported models in MP5,1 would fix.
Unintended consequence of the multiple MP5,1 firmware updates on an incompatible machine, or something else entirely? Unsure if or how this can be reported to Apple as an issue or bug when the E5520 was never part of a machine that is officially Mojave-compatible.
If ALL of the reports can easily correct this with a downgrade to Mojave 10.14.3, then it may indicate a software change that was part of the HDMI/DP audio fixes for certain systems/GPUs that was implemented as part of the 10.14.4 update.
Also assume a CPU swap to supported models in MP5,1 would fix.
Unintended consequence of the multiple MP5,1 firmware updates on an incompatible machine, or something else entirely? Unsure if or how this can be reported to Apple as an issue or bug when the E5520 was never part of a machine that is officially Mojave-compatible.
I will be installing my X5680's tomorrow so I will be able to at least say if THOSE particular processors have any affect on the issue.
[doublepost=1554830709][/doublepost]
While not what I would consider a "solution" to the issue directly, if a processor upgrade on an older 4,1 firmware upgraded machine does indeed solve the issue, these chips are cheap enough that for anyone looking to stay as current as possible software wise, it may be considered a solution...will have that definitive answer tomorrow when the chips arrive provided they arent DOA