There is another way. Transcode to a media form that is easier on your CPU. Your media is a long GOP media which requires an awful lot of CPU to generate each and every frame. Cineform media which is much easier to edit is the way to go. Try the GoPro free version. It is a very high quality industry standard format, the only real disadvantage is that because you have each and every frame is there, your media files will be much larger. I do not have a link but just search these forums for help
Unfortunately, Bill, the Cineform encoder is no longer available for download directly from GoPro, as the program that included it (GoPro Studio) has been EOL'd three months ago. And the last version of GoPro's Quik desktop app that included GoPro Studio was version 2.3, which is also no longer available for download. Beginning with version 2.4, GoPro Studio is no longer included with Quik. And although GoPro will continue to support and update the Cineform encoder, no current or future GoPro software products will include it.
By the way, Premiere Pro CC 2015 and later (including CC 2018) includes a Cineform encoder, and that encoder should have been installed as part of the installation. Plus, since at least CC 2017, Premiere Pro should have given the OP the option to use proxies or transcode upon import (ingest). Cineform MOV presets are included with this feature.
So concerning Premiere only and based on the current footage i am guessing any higher configuration is overkill. I mean if we use proxied footage to edit there is no need for a high end cpu or gpu. If the actual benefits are only a 10% - 20% increase in exporting times and the process is not really time sensitive (delivery wise) any config like a mainstream cpu with a low class gpu like 1060 would be more than enough. Am i wrong on that?
Of course we are talking only for Premiere use on this system.
I am downloading the Benchmark right now and i will post soon.
When i try to run the vbs it says file not found.
I opened the project from both nvme drives and it showed something really interesting though weird.
The 960 Evo 256GB is plugged directly on the motherboard slot.
The 950 Pro 512GB is plugged in a Pci-x controller.
When i run samsung magician benchmark i get
1. When i opened the project from the 960 Evo
Export to 950 Pro took 25 seconds (Speed steady at 1.5GB/s)
Export to 960 Evo took 78 Seconds (Speed starts @350MB/s and goes down to 235Mb/s)
2. When i opened the project from the 950 Pro
Export to 950 Pro took 25 seconds (Speed steady at 1.5GB/s)
Export to 960 Evo took 78 Seconds (Speed starts @350MB/s and goes down to 235Mb/s)
I tried swapping the drives but the controller can't recognize the 960 Evo
Anyway now after all the plugging i tried exporting to 960 evo and it slowed down even more at 85 seconds
MPEG2-DVD Opened from 960 Evo (Warning replaced Fonts)
Export to 960 Evo - 32 Seconds (CPU @ 28%)
Epxort to 950 Pro - 31 Seconds (CPU @ 28%)
H.264 Test Opened from 960 Evo (Warning replaced Fonts)
Export to 960 Evo - 57 Seconds (CPU @ 65% - 100% mostly at 100% - GPU Max 5% on both GPUs)
Epxort to 950 Pro - 59 Seconds (CPU @ 100% - GPU Max 5% on both GPUs)
MPEG2-DVD Opened from 950 Pro (Warning replaced Fonts)
Export to 960 Evo - 32 Seconds (CPU @ 27%)
Epxort to 950 Pro - 31 Seconds (CPU @ 27%)
H.264 Test Opened from 950 Pro (Warning replaced Fonts)
Export to 960 Evo - 59 Seconds (CPU @ 45% to 100% mostly at 95% - GPU Max 5% on both GPUs)
Epxort to 950 Pro - 58 Seconds (CPU @ 45% to 100% - GPU Max 5% on both GPUs)
For Premiere use, a high-end CPU may be required, depending on the video codecs and video resolution used.
As for the GPU, the reason for the Radeon Pro Duo underperforming the GTX 1080 in exports has little to do with the GPU per se, but rather in the differences between the CUDA and OpenCL renderers. Adobe's OpenCL support in the MPE still isn't as robust or as mature as it's CUDA support.
Oh, by the way, for best performance the CPU and the GPU should be matched (relatively speaking). If one significantly underperforms the other, then that underperformance component will limit the higher-end component, and may cause the higher-end component to severely underperform even in situations which do not utilize the underperformance component.
And the worst possible combination for Premiere is an el-cheapo CPU and either an el-cheapo discrete GPU or only integrated graphics: The entire system will be excruciatingly sluggish all around.
"The 950 Pro 512GB is plugged in a Pci-x controller.". I assume you really meant PCIe bus with an adapter card, What one? What PCIe slot?
"When i try to run the vbs it says file not found." This happens when you run the script from other than the location of the export files..
i can see no logical explanation for those weird results. The only unusual situation is that I have never seen operating with both a CUDA GPU and an AMD GPU at the same time. If you could pull out that Radeon for a test to see if that disk exporting weird action goes away (doing that frees up PCIe lanes).
Your source footage being H264 is the bottleneck and not Premiere. Transcode that footage to something less CPU intensive (like DNxHD) and your editing should be smooth.
I don't have both cards in there. I wrote about the GTX 1080 because i did use it extensively until now and i know how it performs on this system. Only the Radeon Duo Pro (Polaris) 32 GB is on the PCI-E 1 x16 right now. Then the Adapter for the NVME Drive 950 Pro is on PCI-E 2 X8, PCI-E 3 is Free because it shares bandwidth with PCI-E 1, PCI-E 4 X8 10Gbe Lan.
It seems there is a problem with the write on the 960 Evo. I tried copying the Disk Test.avi file.
960 Evo -> 950 Pro = 1.5GB/s steady throughout the copy
950 Pro -> 960 Evo = Starts at 1.5GB/s and after 10 seconds it drops to 330MB/s and stays there throughout the copy.
I will try moving the 950 Pro on the motherboard slot to confirm that the problem is on the 960 Evo and not the motherboard but i think the latest evo firmware might have done the trick.
What i meant as cheap cpu and graphics would be a Ryzen 1700 and a GTX 1060 not anything lower than that. If transcoding the footage is the only option here based on the stupid DJI files especially, then this combination should be enough for using premiere even if that means you get slight worse exporting times. The real time editing with the transcoded footage would be pretty much the same unless i were to use to many video channel etc.
My original question was actually if it is worth it to move to Threadripper 1950x for real time editing. If i still need to transcode this footage there is no actual benefit as there is no absolute need for that.
Ok so i asked on another forum about the 960 Evo and they told me that after 20GB the 960 Evo drops to 300MB because it doesn't sustain the write speeds. That's why the DiskTest.avi and the simple copy fails.
So the only option i have here is:
850 Evo -> Boot Drive
960 Evo -> Source Footage (Original + Transcoded)
950 Pro -> Projects, Exports, Previews
So i first copy all the footage from the NAS (550MB/s Read) to 960 Evo and start from there.
And at some point if i feel comfortable (money wise) i uprgade to Threadripper 1950x to be able to max out my specs for use with the rest of Adobe CC Collection as well...
I mean Adobe Dimension CC shows 45% utilization on GPU 0 when using it prior to render and then the render is mostly CPU.
Why Adobe hasn't figured out on how to utilize the GPUs?
I can only think that it would make the countless Intel CPU product lines obsolete. Although i believe that would straighten out and simplify the Market.
First of All i don't understand a single word. This is already enabled but the GPUs don't utilize compared to the cpu and of course there is no problem playing back a single file even with lumetri (live changing while playback) and 3 effects on top (depending on the effect)
I think you are thermal throttling your 960 EVO, a condition I have not seen on my M.2 devices as you can see form my storage test results. I do have a small fan blowing air across my 960 PRO 1 TB motherboard mounted M.2 and then the air goes past the adapter board mounted 960 PRO 512 GB M.2. For the following test I pu
t my 960 EVO on another adapter card and it also has a littlle air flow past it. As you can see from the chart from my PPBM website plot the export time for those drives I never have seen any throttling on my units.
So I went one step farther and found a way to monitor the temperature of the M.2 drives it is HWiNFO below you will see below that under the Disk I/O export my 960 EVO reach a maximum of 62 degrees C
It took awhile to find these numbers in HWiNFO but if you select Sensors it helps
The Write Drop starts much earlier than when the temperature starts to rise. But i does get higher and higher while trying to copy. Can you confirm the firmware you are using. I have the 3B7QCXE7
Mine is the original (never upgraded) 287QCXE7
We do not know where that temperature sensor is so one of the several chips on that device could be causing the problem but it may have some thermal lag to reach the sensor.
Looks like last night's escapade wasn't as bad as shown above
It appears the firmware update required a reboot and due to the late hour I only did so this morning...
Note; when I mentioned the firmware update was done unintentionally, it was due to the "refresh" icon on the top left of Samsung's Magician's window. It isn't a "refresh" - unless one accepts refresh means updating the firmware
Below are the results as of now...
Also attached is a screen clip showing the temperature of the 960 Pro as it was running the "4K Write" cycle. The actual temperature reached was 51 degrees Celsius. It's the line at the very bottom, which is partly cut off...
In this instance, the newer firmware appeared to have reduced the sequential read speeds slightly.
Ok so Samsung really messed up with our drives. We just have to wait for an update? Is there a way to contact support for the ssd's?
But that is not the question of this topic.
Would and AMD Threadripper 1950x improve my workflow based on the current footage we edit or would i need to ingest with that CPU as well?
Would i see the GPU utilizing more with the threadripper?
Yes, that became evident after the original speed tests with the firmware as it came out of the box.
There was also an issue as soon as the firmware was updated the ASUS mother board no longer recognized bot the 950 and 960 as PCIe Gen 3 x 4. This was eventually resolved with an ASUS Bios upgrade. However, there was that loss in speed you noted.
Of interest, not upgrading the 950 or 960 bios allowed these devices to function without issues...
It may be the NVMe driver, see the clip below, out of the three Samsung 9xx pros only the one with the higher Read rate is not shown as a NVMe device -
I am waiting for a reply from Samsung about the problems we are having. I also tried crystal disk mark and HWInfo showed that the 960Evo was going to 95 Degress Celsius therefore i stopped it.
Anyway i overclocked the Core i-6850k to 4.3Ghz with higher voltage and it gave me 4 seconds less on the H264 test.
Then i used my 4k DJI Footage on a 1080p Timeline and applied some directional blur effects.
While playback and realtime effects changing i saw the CPU staying at 20% load and the GPU0 at 55%.
Though when exporting to h264 CPU was at 100% load and GPU at 1% (only my monitor was giving the load i guess)
What the hell is up with that?
I also tried exporting to MPEG2 Bluray and the CPU stayed at 30% load while the GPU at 1% therefore the exporting time was increasing.
I am very confused. Am i missing something?
I am see you Speccy results above and i believe it is the Speccy tool just has not caught up with the newer technology.
Below that what is the ambient temperature with no load on that 960 EVO and where is is mounted?
I really think you should send that 960 EVO to Samsung for testing!
The 960 Evo is mounted onto the motherboard slot. At idle the temperature is at about 42 C.
I have already send Samsung my specs and configuration with some screenshots where a simple file copy drops and i am waiting for their reply. Although the Evo is not my main concern here. Unfortunately based on how i've seen it work so far i have to jump to Threadripper at some point
That 960 EVO ambient temperature is about 10° C hotter than my motherboard mounted unit, Of course I said I have a sperate fan at that point.
Well jumping to Threadripper will not solve that specific problem.
I see above that you have run 3 out of the four PPBM tests the last one is the export of the MPEG2-DVD timeline without GPU Acceleration i.e. CPU only. What is your score on it so I can see how much improvement might be possible?
It took 76 seconds to export with CPU only
Sorry but something must have been set up wrong that number is impossible
An i7-6850K at 4.4 GHz normally scores about 300 seconds on that test
It must be the Space Shuttle!
I open the PPBM12-DE.prprj from "New Folder" on my 950 Pro
i select to the MPEG2-DVD test timeline.
I go to FIle -> Export Media and i leave everything untouched. I add the export to the queue and select Render "Software Only"
Am i doing something wrong?
Edit: I used the export directly without media encoder but changed the project settings to software only and it took the same amount of time to export.
Actually Maximum Render Quaitiy didn't check itself.
Now that i've checked it it takes 313 seconds on CPU only and 32 seconds with GPU Acceleration and the CPU stays at 26-29% load
Edit: I redid the h264 test it got 73 seconds.
Note: i currently have on open Teamviewer session on idle and chrome opened while doing the above tests
This is what Samsung has replied:
The 960 EVO is a TLC based device which has something called Turbo Write which uses a cache that is placed on the drive to provide the fast speeds. When the cache is full your writing to the NAND directly which results in lower speeds. However this temporarily and once the cache is empty the write speeds go back up. Going based off the results of these screen shots there is no present error occurring with the drives and this is completely normal. The numbers in Samsung Magician also show consistent results.
So this what your Output.csv file would look like:
These are all fairly good numbers, both the 73 seconds and the 313 seconds possibly would go down slightly with fewer processes running. Your GPU assisted MPEG2-DVD s core seem high to me as I tested a Superclocked one and got 15 seconds and would expect about maybe 18-20 seconds Was it running 99% usage?
See if you can get that idling temperature down. I have never seen thermal throttling in my system. As I said you may have to install a small fan to cool you M.2 devices
The GPU assisted MPEG2-DVD takes 32 seconds and the CPU stays at max 29% load.
I will install the gaming drivers for AMD Radeon Pro Duo as well and test.
Edit: Seems like the only card that doesn't have the Gaming feature is mine! Yey!
Thanks for posting the Samsung Feedback
In my case, after the firmware update, it explains the need to reboot before the read / write speed results recovered