What's new
Carbonite

South Africa's Top Online Tech Classifieds!
Register a free account today to become a member! (No Under 18's)
Home of C.U.D.

Stuttering when using Premiere Pro & Photoshop

yepsy

Epic Member
Rating - 100%
46   0   0
Joined
Jun 7, 2012
Messages
1,777
Reaction score
150
Points
3,985
Location
Cape Town, Western Cape
I have been experiencing varying levels of stuttering and unresponsiveness when using Premiere Pro, Photoshop, and when gaming.

With Premiere it will come and go, depending on how the content has been buffered, but it is still annoying when my M1 Macbook Air provides a smoother playback experience. Running two SSDs and loads of RAM, I can't seem to figure out what is causing this exactly.

With Photoshop it will sometimes be severely worse, to the point you would struggle to use the application at all. Loads of excess RAM available, neither CPU nor GPU under much load, SSDs also not doing much.

Sometimes while launching games or switching in and out of them I also experience stuttering and a weird level of unresponsiveness, giving me the impression that some part of the chain is being bottlenecked. I keep an eye on task manager and can't see anything weird. The only thing that catches my eye is possibly slow read and write speeds being reported on the system drive (970 EVO Plus) of maybe a few MB/s, but I'm not sure if that is abnormal as it also varies with the workload. I have run benchmarks and diagnostics using Samsung Magician, with nothing strange being reported.

Any ideas as to what can be causing this issue?

My Specs:
ASRock X570 Phantom Gaming X AMD X570 Ryzen Socket AM4 ATX Desktop Motherboard
AMD Ryzen 9 3900X 12 Core 3.8GHz (4.6 GHz Boost) Socket AM4 Desktop CPU
2x G.Skill Ripjaws 16GB DDR4-3600MHz Black Desktop Memory
Palit GeForce RTX 2080 SUPER GameRock Premium Edition 8GB GDDR6 256-bit PCI-E Graphics Card
Corsair HX1200i 1200W 80 Plus Platinum Certified Fully Modular Desktop Power Supply
Samsung 970 EVO Plus 1TB NVMe M.2 PCIe Solid State Drive
Mushkin Helix-L 1TB M.2 PCIe NVMe Solid State Drive
 
Overheating CPU or overheating SSD?

Maybe try something like HWInfo and have a look at its reported values?

Anyone know what the VRM on that Asrock board is like? That could also cause the CPU to throttle.
 
Personal experience with DEVELOPMENT and SSDS.

When compiling code it DESTROYS just one area of your drive over and over. (Deletes compiled code and writes compiled code.)
Ends up making one section of the drive super slow..

For instance, system is slow.
Ok reformat windows. Sheesh super quick.
Install all softwares required. Damn super quick
Once you've installed everything and you're using the same sectors again its gets POEP slow again.

Not sure how SSD drives actually work in regards to reusing sectors at specific locations but this was my experience.
My laptop NVME 1TB
  1. Host Reads : 18450 GB
  2. Host Writes : 17267 GB
  3. Temperature : 42 C (107 F)
  4. Health Status : Good (48 %)


Now in your case, i suspect that TEMP files are created. And these boys are destroying your drive.
 
Last edited:
Personal experience with DEVELOPMENT and SSDS.

When compiling code it DESTROYS just one area of your drive over and over. (Deletes compiled code and writes compiled code.)
Ends up making one section of the drive super slow..

For instance, system is slow.
Ok reformat windows. Sheesh super quick.
Install all softwares required. Damn super quick
Once you've installed everything and you're using the same sectors again its gets POEP slow again.

Not sure how SSD drives actually work in regards to reusing sectors at specific locations but this was my experience.


Now in your case, i suspect that TEMP files are created. And these boys are destroying your drive.
Not sure if this will happen with damaged sectors on SSDs, but I seem to remember doing full formatting (not quick) of an HDD will detect bad sectors and write them to a table so that those sectors won't be used again, maybe, just maybe it would do the same with ssds? Would require a format or some sort of scan though at the very least.

Overheating CPU or overheating SSD?

Maybe try something like HWInfo and have a look at its reported values?

Anyone know what the VRM on that Asrock board is like? That could also cause the CPU to throttle.
I agree to use HWInfo, the max and minimum values are a big help in troubleshooting.
 
Personal experience with DEVELOPMENT and SSDS.

When compiling code it DESTROYS just one area of your drive over and over. (Deletes compiled code and writes compiled code.)
Ends up making one section of the drive super slow..

For instance, system is slow.
Ok reformat windows. Sheesh super quick.
Install all softwares required. Damn super quick
Once you've installed everything and you're using the same sectors again its gets POEP slow again.

Not sure how SSD drives actually work in regards to reusing sectors at specific locations but this was my experience.
My laptop NVME 1TB
  1. Host Reads : 18450 GB
  2. Host Writes : 17267 GB
  3. Temperature : 42 C (107 F)
  4. Health Status : Good (48 %)


Now in your case, i suspect that TEMP files are created. And these boys are destroying your drive.
The previous owner of the PC did tell me that he used it for some machine learning. I’m not sure if that falls under that category.

I will keep an eye on HWInfo and report back. Appreciate all the reponses.
 
When compiling code it DESTROYS just one area of your drive over and over. (Deletes compiled code and writes compiled code.)
Ends up making one section of the drive super slow..

This should not be happening to an ssd, the OS doesn't know which nand cell the data is stored in (there is an internal data mapping table the ssd controller uses), and the ssd itself will apply wear levelling to distribute write operations.
 
This should not be happening to an ssd, the OS doesn't know which nand cell the data is stored in (there is an internal data mapping table the ssd controller uses), and the ssd itself will apply wear levelling to distribute write operations.
From my experience this has happened on multiple machines.

My current machine has 32gb ram and i don't go over 24 gb but it feels like the ssd is getting destroyed by compiling.

49% health after only 28tb.
 
I used to have a very similar desktop build. I experienced the same then I switched to a MacBook Pro and Photoshop and Lightroom do much better. I built mine new so doubt it had anything to do with ssds etc. It could be some random driver or hardware issue or something but I think PS/LR just run better on macOS. Also I don’t think I’ve seen either of them crash on macOS but definitely did on windows.
 
49% health after only 28tb.
If I'm not mistaken, the % health status (or the "remaining life") is calculated using the drive's built-in value for its TBW rating.

E.g. If a drive is rated for 100 TBW, then each terabyte written will subtract 1% from the reported health.

So if your drive is reporting 49% health after 28 TBW, that means the drive was only rated for around 57 TBW to begin with.

Unless you're using a different metric for health.
 
If I'm not mistaken, the % health status (or the "remaining life") is calculated using the drive's built-in value for its TBW rating.

E.g. If a drive is rated for 100 TBW, then each terabyte written will subtract 1% from the reported health.

So if your drive is reporting 49% health after 28 TBW, that means the drive was only rated for around 57 TBW to begin with.

Unless you're using a different metric for health.
My drive has 5 year warranty and 200TBW

Oh and is only 2.2 years old
 
Make sure you're not in power saving/balanced power mode. Small thing that can make quite a difference.
 
Crystal Disk Info. Don't have an idea on how they determine that
I think when you run a report using CDI it gives you a screen with all the parameters it uses to calculate that score.

Like it'll list TBW, reallocated sectors, etc.

Might be worth posting a screenshot, then it might be possible to determine what exactly is going on with your drive and whether there's a way to prevent it from aging prematurely.

Also what is the make and model of the drive?
 
I think when you run a report using CDI it gives you a screen with all the parameters it uses to calculate that score.

Like it'll list TBW, reallocated sectors, etc.

Might be worth posting a screenshot, then it might be possible to determine what exactly is going on with your drive and whether there's a way to prevent it from aging prematurely.

Also what is the make and model of the drive?

----------------------------------------------------------------------------
CrystalDiskInfo 9.1.1 (C) 2008-2023 hiyohiyo
Crystal Dew World: Crystal Dew World
----------------------------------------------------------------------------

OS : Windows 11 Professional [10.0 Build 22621] (x64)
Date : 2023/07/28 13:04:34

-- Controller Map ----------------------------------------------------------
+ Standard NVM Express Controller [SCSI]
- MKNSSDDE1TB-D8
- Microsoft Storage Spaces Controller [SCSI]
- Xvdd SCSI Miniport [SCSI]
- Microsoft VHD Loopback Controller [SCSI]

-- Disk List ---------------------------------------------------------------
(01) MKNSSDDE1TB-D8 : 1000.2 GB [0/0/0, sq] - nv

----------------------------------------------------------------------------
(01) MKNSSDDE1TB-D8
----------------------------------------------------------------------------
Model : MKNSSDDE1TB-D8
Firmware : EGFM52.3
Disk Size : 1000.2 GB
Interface : NVM Express
Standard : NVM Express 1.3
Transfer Mode : PCIe 3.0 x4 | PCIe 4.0 x4
Power On Hours : 6841 hours
Power On Count : 81197 count
Host Reads : 18475 GB
Host Writes : 17303 GB
Temperature : 39 C (102 F)
Health Status : Good (48 %)
Features : S.M.A.R.T., TRIM, VolatileWriteCache
Drive Letter : C:


My power on count is concerning. But i think thats how this wootbook handles the drive.
 
I don't think it's the drive.
Storage will hardly ever run at sustained full capacity unless you are benchmarking.
It also wouldn't cause stuttering like he is describing... It would just be slow at loading assets.

I'd start with GPU or chipset drivers.
 

----------------------------------------------------------------------------
CrystalDiskInfo 9.1.1 (C) 2008-2023 hiyohiyo
Crystal Dew World: Crystal Dew World
----------------------------------------------------------------------------

OS : Windows 11 Professional [10.0 Build 22621] (x64)
Date : 2023/07/28 13:04:34

-- Controller Map ----------------------------------------------------------
+ Standard NVM Express Controller [SCSI]
- MKNSSDDE1TB-D8
- Microsoft Storage Spaces Controller [SCSI]
- Xvdd SCSI Miniport [SCSI]
- Microsoft VHD Loopback Controller [SCSI]

-- Disk List ---------------------------------------------------------------
(01) MKNSSDDE1TB-D8 : 1000.2 GB [0/0/0, sq] - nv

----------------------------------------------------------------------------
(01) MKNSSDDE1TB-D8
----------------------------------------------------------------------------
Model : MKNSSDDE1TB-D8
Firmware : EGFM52.3
Disk Size : 1000.2 GB
Interface : NVM Express
Standard : NVM Express 1.3
Transfer Mode : PCIe 3.0 x4 | PCIe 4.0 x4
Power On Hours : 6841 hours
Power On Count : 81197 count
Host Reads : 18475 GB
Host Writes : 17303 GB
Temperature : 39 C (102 F)
Health Status : Good (48 %)
Features : S.M.A.R.T., TRIM, VolatileWriteCache
Drive Letter : C:


My power on count is concerning. But i think thats how this wootbook handles the drive.
Well shit. I'm officially stumped. Based on that though you only have 17 TBW currently, which is even more concerning.

It might be the power on count, yeah. But we've exceeded the limit of my knowledge on this.

Out of interest. What does HWINFO64 say for host writes and remaining life? I've always just used HWINFO's stats.
 
Crystal Disk Info. Don't have an idea on how they determine that


----------------------------------------------------------------------------
CrystalDiskInfo 9.1.1 (C) 2008-2023 hiyohiyo
Crystal Dew World: Crystal Dew World
----------------------------------------------------------------------------

OS : Windows 11 Professional [10.0 Build 22621] (x64)
Date : 2023/07/28 13:04:34

-- Controller Map ----------------------------------------------------------
+ Standard NVM Express Controller [SCSI]
- MKNSSDDE1TB-D8
- Microsoft Storage Spaces Controller [SCSI]
- Xvdd SCSI Miniport [SCSI]
- Microsoft VHD Loopback Controller [SCSI]

-- Disk List ---------------------------------------------------------------
(01) MKNSSDDE1TB-D8 : 1000.2 GB [0/0/0, sq] - nv

----------------------------------------------------------------------------
(01) MKNSSDDE1TB-D8
----------------------------------------------------------------------------
Model : MKNSSDDE1TB-D8
Firmware : EGFM52.3
Disk Size : 1000.2 GB
Interface : NVM Express
Standard : NVM Express 1.3
Transfer Mode : PCIe 3.0 x4 | PCIe 4.0 x4
Power On Hours : 6841 hours
Power On Count : 81197 count
Host Reads : 18475 GB
Host Writes : 17303 GB
Temperature : 39 C (102 F)
Health Status : Good (48 %)
Features : S.M.A.R.T., TRIM, VolatileWriteCache
Drive Letter : C:


My power on count is concerning. But i think thats how this wootbook handles the drive.

I wouldn't blindly rely on crystal disk, it's an agnostic tool that seems to only read smart data (great for hdd) and the drive controller may or may not fully disclose all the relevant info on their smart report (though I don't know it well enough to determine how accurate or inaccurate it is), you should also try ensure you use the latest version of crystal disk as drive manufacturers may release newer firmware that changes how ssd controllers behave and do their reporting. Preferably you should check with the manufacturers software like samsung magician for samsung ssd, etc, but failing if no software is available then you should refer to the individual smart attribute values if they are present, this short article will explain,


A high power on count shouldn't matter for an ssd, it might be indicative that your pc tends to put the ssd into "standby/sleep mode" and then power on when you use it.
 
Well shit. I'm officially stumped. Based on that though you only have 17 TBW currently, which is even more concerning.

It might be the power on count, yeah. But we've exceeded the limit of my knowledge on this.

Out of interest. What does HWINFO64 say for host writes and remaining life? I've always just used HWINFO's stats.

Self-Monitoring, Analysis and Reporting Technology (S.M.A.R.T.)
Available Space Below Threshold: OK
Temperature Exceeded Critical Threshold: OK
Device Reliablity Degraded: OK
Media In Read Only Mode: OK
Volatile Memory Backup Device Failed: OK

Drive Temperature: 34 °C
Warning Temperature Threshold: 75 °C
Critical Temperature Threshold: 80 °C
Time Above Warning Temperature Threshold: 0 minutes
Time Above Critical Temperature Threshold: 0 minutes

Spare Capacity Available: 100%
Device Health: 48%
Power Cycles: 81198
Power On Hours: 6841 hours
Unsafe Shutdowns: 62
Media Errors: 0
Total Host Reads: 18480 GBytes
Total Host Writes: 17307 GBytes
 
Self-Monitoring, Analysis and Reporting Technology (S.M.A.R.T.)
Available Space Below Threshold: OK
Temperature Exceeded Critical Threshold: OK
Device Reliablity Degraded: OK
Media In Read Only Mode: OK
Volatile Memory Backup Device Failed: OK

Drive Temperature: 34 °C
Warning Temperature Threshold: 75 °C
Critical Temperature Threshold: 80 °C
Time Above Warning Temperature Threshold: 0 minutes
Time Above Critical Temperature Threshold: 0 minutes

Spare Capacity Available: 100%
Device Health: 48%
Power Cycles: 81198
Power On Hours: 6841 hours
Unsafe Shutdowns: 62
Media Errors: 0
Total Host Reads: 18480 GBytes
Total Host Writes: 17307 GBytes
Bizarre. It must be the power cycles then, or maybe power on hours. I have a 5 year old laptop and SSD with only 9 TBW (not used in a write-heavy context) but I only have 912 power cycles - drive is at 98% health. So 81 000 power cycles seems insane. Might be worth investigating that
 
Well shit. I'm officially stumped. Based on that though you only have 17 TBW currently, which is even more concerning.

It might be the power on count, yeah. But we've exceeded the limit of my knowledge on this.

Out of interest. What does HWINFO64 say for host writes and remaining life? I've always just used HWINFO's stats.

If the report is actually accurate, seems to indicate a faulty/defective drive. I would still refer to the individual raw attribute data to determine what is what.
 
Back on topic, I would check temperatures. Components maybe overheating and the system is downclocking.
 
PowerCycle are sus, maybe stress test the drive and see if it stutters at all.

You could also try disabling Sysmain service, previously known as SuperFetch and see if this helps at all.

And have you installed any malware or bitcoin miners that come with some of the free software you can download?
 

Users who are viewing this thread

Latest posts

Back
Top Bottom