(Corrections): I did want to make a correction: As swap memory is available for M1 Macs (really all Macs/PCs), iPadOS seems to limit this functionality to actually "killing" the open apps to free up memory, or rather just handling this a bit differently than actual swap as we know it on a full operating system. I do believe this is possible and as more info comes out about memory utilization we could potentially see this functionality. This is a bit of a repeat based on some questions I've received about iPadOS 15. I did however find something I wasn't expecting. Testing will continue and I'm experimenting with some other beta released apps. I'll keep you posted. Go rock faces.
Amazing video. I have the 128gb 8gb ram model and this video made me feel a little more comfortable for not having to upgrade to the 16gb ram right away. Especially with the delays of m1x. Definitely gained a sub and cant wait for more videos.
Great to hear. And I know, it’s pretty awesome for a first beta. So far so good on my end. Glad to have you here with your workflow. Always helps to hear from others on how their experience is going. 👍
Thanks for the update, Kevin! I know you’ll keep us informed. BTW, the background music is perfect. It’s laid back and at the proper volume, so not to be distracting. 👍
I appreciate that. I’d spend more time researching tunes, but do try to keep it at the pace and feel of the video. It actually takes way more time than people think 😀 And yeah, just enough volume on the track to know it’s there, but not overwhelming to the vocal track. 🤙
12:19 Well that’s different! Thanks for the follow up. I haven’t checked, but I wonder if there are any new APIs in iPadOS 15 that would help speed this up? 🤔
I’m digging further, but if this is starting to become more a out communication protocol and not as much a out hardware, then it’s possible, but I still think hardware is a limiting factor, but I’m all for proving myself wrong.
great video Kevin, super happy to see an increase in transfer speed this also gives me hope that they can keep pushing it with more updates down the line, would you agree? looks like it hit a wall at 950mbs.
Yep, I be continued to test and haven’t gotten above that threshold, but heck of a lot better than 550. I really want to test the render and export on the SSD. I’m trying to see if I can find any information as to why this change occurred via the software.
@@KevinRossRN you are also the first to talk about this I also cannot find any additional information besides your video so well done!! appreciate the work
Very cool. This channel has been fantastic for me to get some data to base purchase decisions on. I definitely want to update my tablet game, but I’m very content to just wait a bit before taking the M1 iPad plunge. I’m enjoying the crap out of my M1 Mac Mini and it’s going to be awesome to see the development path that comes from all the Apple Silicon innovations!
Hey, you keep enjoying that mini. That thing is my rock when it comes to content creation. We’ve been through a lot together. Awesome to see you friend.
@@KevinRossRN I agree…I have only been able to get the fans to activate ONCE above baseline: I engaged ‘Flex Pitch/Flex Time’ on Logic on a VERY long stem. It only jumped 10 degrees C then promptly went right back down to its normal idle temp! 🤣😂 It’s a beast.
After the WWDC 21 show I staying with shooting on my Sony ZV1 and editing on my Mac Mini M1, so for the iPad Air 4 seems to work for me. Between the $200 and the $60 difference in AppleCare for me $279 with tax I don’t see enough of a difference with iPad Pro M1. Hopefully I agree after iPad OS 15 is available. Great videos. I am really I came across you channel.
That’s great to see the throughput results! Almost double the speed, that’s very encouraging- great thorough video again, love it! I did notice something odd on a TB4 hub I’m testing so this is great news!
I’m interested to hear what you’ve found on that hub. It seems that there are certain drives that aren’t recognized on the M1 IPP. I’ll be testing the hubs that I have. But yes, encouraging that the throughput may have been “fixed,” or at least better.
Thank you very much for your helpful reviews. Not only Thunderbolt 3 drives but USB 3.2 SSDs are slow too. On a Samsung T7 USB 3.2 10 Gbps SSD I got only 44 MBps copying a 5 GB file to the HFS+ formatted SSD using the original USB 3.2 and a TB3/USB4 cable. Did you see any speed improvement with USB 3.1 and USB 3.2 SSDs too?
I did actually. I’ll pin this comment to help others. So on a USB-C Gen 2 (SanDisk 1050MB/s read/write) I was getting a sustained 500MB/s, which is up from the previous ~250 - 300s.
It’s a passive cable that came with the enclosure. I’ve tested others. The cables that come with those enclosures are pretty sturdy…better than others.
Hi Kevin, Just a couple minutes into this video, I see your beautiful poster “Audio Apparatus”! I will watch the rest of the video now, hoping for a better angle/close up of the poster. 🎶👍
Really enjoying your very thorough work. Just subbed. I have one question. I am strictly a stills photographer, and I work almost entirely on my iPad Pro M1 512gb. What would you recommend as external storage? T5, T7 maybe, Sandisk Extreme or, something else?
Thanks. So as far as an external SSD, I like all of those options. The SanDisk Extreme is one that I use quite often as well as a DIY option that I reviewed recently from Sabrent. But honestly you’re not going to go wrong with any of those options. I’d search for the best price, especially now. Either T7 or SanDisk Extreme should give you slightly higher read and write on the iPad.
About the transfert speed, iOS 15 do it better than iOS 14. But TB3 allow 5GB/s right? So why we still lock to 960MB/s, that’s the question. As an owner of iPad Pro 2020 and care about speed transfert, I m reassured to keep my iPad. I will receive my SSD NVME this Thursday and will do some tests that I will share with you.
This has been the issue with these TB controllers on the M1's. I've actually got another video coming out on the M1 Mac with the latest operating system. Working on that today. There's going to be some bottleneck through the cable and enclosure along with the rated read/write speed and cache of the SSD, but yeah, 960 is still slower than it should be with this particular setup I have. Feel free to circle back around with your updates. I'm definitely curious.
@@KevinRossRN i send back my NVME Disc, I don’t know if it it the case or because I can’t update my disc on windows desktop. But I was restricted to 20-60MB/s while transferring in and from my iPad Pro 2020. So I think the NVME is not a good value and I bought an 1TB 2,5’ SSD Crucial MX500. I already get one and the speed sharing is up to 300-350MB/s. I wait here, keep what I have and hope a better iPadOS the next year. I wondering if I will bought a surface. Because I like the windows files management, better than iPadOS.
Well I’d say the surface is a whole different machine…basically. I think it’s a great device and would recommend it to anyone that wants that full operating system and familiar with Windows. Hopefully with the updates we’ll continue to get better performance. These chips are awesome, and will only get better.
For context, I have verified (using JazzDiskBench) that the read and write speeds on the M1, 256GB, 12.9 inch iPad Pro are almost exactly 3,000MB/s and 2,000MB/s, respectively. With that being said, these 2 things are confusing the hell out of me: 1. The iPad Pro (even in iPadOS 15) gets 5Gb/s speeds rather than 10Gb/s speeds via USB 3.1 Gen 2 2. The iPad Pro (even in iPadOS 15) gets 1GB/s speeds via Thunderbolt when it should be getting the full 2GB/s This is one of the situations where I’m just like… WTF?
Welcome to the head scratching club of M1. 😀 I’ll have to check out that benchmarking tool. I’ve got some additional tests that I’m running, however this also speaks to some of my other tests I did on the Mac where I could essentially get the same results in video rendering and export from a USB-C SSD vs the Thunderbolt enclosure with a very fast NVMe SSD 🤷♂️ The communication protocol is definitely off.
@@KevinRossRN Yeah, I’ve considered going the Thunderbolt enclosure route for my external M.2 SSDs, but they’re way more expensive than the 10Gbps ones and it seems there’s little real world benefit. Real world speeds on the 10Gbps enclosures are ~925MB/s, and I can’t image I’d need anything more than that anytime soon. I prefer the ASUS ROG STRIX Arion despite its obnoxious appearance. Reliable brand and great heat dissipation. Regarding the JazzDiskBench, I had to run it a few rounds before getting peak speeds. The first runs were a little bit slower for some reason.
Interesting on that benchmark. I’ll be mindful of it. And yes, the one benefit of the enclosure is that you get to choose the enclosure and/or the SSD, but agree that the real world difference is pretty insignificant in most use cases. I still use these Gen 2 SSDs all the time and recommend them often.
The blooming is definitely present, but it’s really a matter of whether it would bother you, I suppose. I see it if I’m looking for it when the room is dark, but personally it’s not something that I’m concerned with. But you can certainly save some money with the 2020. As far as performance, you’d likely not notice any difference.
Well it can easily be swapped on the SSD, but then there’s the “what if” factor that at some point an app is going to use more than 5GBs (per app) than what we already have. Appreciate you hanging on this one.
There really isn’t much. The RAM is app dependent, and more developers just need to make adjustments to utilize that RAM, but it won’t automatically happen in iPadOS15. As far as the Thunderbolt performance it too is pretty much the same. I’ll continue to test as more updates come in.
There is actually a website to get the dev beta profiles without being a developer. Google it using safari from the device you want to download it too :)
Yeah, I’ve seen that before. Didn’t see it this round, but I know it exists. I like having access to the developer forums and the other resources. Have you tried the beta, or thinking about it?
@@KevinRossRN you are right my friend when you have that account you get more resources. I have iOS 15 on my phone and on the watch it’s pretty stable for a dev beta
The cable I did use was short. I’ve been testing these ports on the M1s since last year and used several variations throughout my testing. It’s a known issue 🤷♂️
This is a great point and something that had plagued the previous versions. So yes, iOS/iPadOS worked more in a virtual memory capacity and basically “killing” the apps to free up memory rather than a typical page in/page out scenario (with swap) that occurs on MacOS, or really any full operating system. The interesting thing here, and yes this would seem to be OS dependent, I agree, but the interesting piece here is that I’ve had apps opened for weeks and when going back to them whether suspended…I can’t say for certain, but they seem to still be active because they don’t auto refresh, even if I call up several other apps that seem to at least want to utilize as much of that max 5GBs of RAM. So while maybe the general use of the term swap may not be appropriate at this juncture until we get more details on this chip and how this OS is handling that unified memory, I’ll take the hit and correct it to say it could be that same old virtual memory as it has been, but the question still may remain as to how it keeps these apps suspended without a full refresh and/or killing the app…at least in most cases I’ve found because I know that iPadOS would likely need to reallocate that RAM, especially as additional multitasking is starting to become supported on iPadOS 15. I appreciate you posing this question though as I’m always thankful to have additional perspectives here. I’m continuing to look into this and will certainly provide a follow up. 😉🤘
Are the iPad Pro 2021, the Acasis enclosure and the WD SN750 all using 4 PCIe lanes? There is another cheaper TB3 enclosure the OWC Envoy Express but only supports 2 lanes.
Yes, the controller in the enclosure is Gen 3 (4 lanes) and the M1s will also have 4. I wasn’t aware of the details on the OWC. I’ll have to dig into that. Not sure which controller they’re using in that one.
They’re all the same friend. Apple is utilizing many of the same parts in these devices as a cost savings. Even those Intel Retimers. You also have to consider the communication protocol with each enclosure. This is why some work better than others, so definitely research what controller/chip on the board these manufacturers are using.
@@KevinRossRN How did perform the WD SN750 APFS formatted in the Acasis enclosure? Do you believe a faster NVMe with bigger DRAM cache could perform better on the iPP2021? I appreciate very much your efforts to figure this speed limit out.
The SN750 has performed quite well across all my M1s. Even though it’s a Gen 3 it performs as well as a Gen 4 SSD that I have from Silicon Power. You’re really only going to notice that cache when you’re transferring TBs of information (sustained). Unless you’re using the IPP as some kind of Server, like if you’re running some RAID array off of it, then you might consider a much larger cache. But, after that buffer at around 14GBs the write would typically drop from about 2800 to somewhere around 1600, however with that throughput I’m maxing out at around 900-950 on the IPP, and so far I haven’t been able to get any higher than that right now.
I have connected them, but it’s a pretty inefficient way because iPadOS and MacOS have this built in sync functionality that really makes it difficult for a drag and drop solution. It’s really not intuitive and more of a pain.
@@KevinRossRN Thanks, I agree. The sync functionality is not yippie… How about, still connected directly over Thunderbolt 3, setting up a shared folder that you would be able to read/write to and using an app like ‘FileBrowser’ on the iPad? The app has various connection methods. Not sure if the direct (Thunderbolt 3) connection would ‘force the dreadful sync functionality’ though. Still without the direct connection over Thunderbolt 3, a (direct) connection over Thunderbolt ethernet might be interesting to test, same app, ‘FileBrowser’? Only if you have the needed software/hardware of course. Don’t go spending any money just for this specific scenario ✅👍 Best,
I figured this t-bolt issue had to be a software problem. This is based on by knowledge as a lawyer, not any technical skill. Why? Because Thunderbolt is a proprietary brand of Intel. Brand standards are not the same as engineering standards. Engineers specify hardware performance. Brands specify performance to the ultimate consumer. E.g., if you buy a Burberry coat manufactured in any factory anywhere in the world you know the seams will be straight, drape consistent, stitching even and very tight. The same is true for tech in spades. neither Intel nor Apple will allow a product to carry their name brand without extensive qualification. This will always extend to the whole system to make sure the branded component will actually perform as advertised. While it's possible both the Intel and Apple engineers blew it, it's not the way to bet. So if it's not hardware, it has to be software. I'm betting that the power management throttled the transfer speed because due to the unique design the iPad was always operating on battery power. Also, if you are not doing anything other than transferring files, the power management (I think) assumes you're not actively using the ipad. this would explain (maybe) why the performance is better when you run other tasks along with file transfer--the power manager thinks "Hey, Kevin is really working hard. Need to boost the performance!." Be interested in your thoughts after you have more time. BTW, your channel is great for an ignorant person who is really trying to understand video. thanks.
Appreciate you dropping by on this one. And yes, after extensive testing on the M1 Macs I kept hitting a wall, and waiting for software updates. After digging in (to your point) I did find that Intel Retimer in the tear down and figured they needed this because the controller being on the chip and a low powered one at that, they then had to put these near the port to equalize the throughput. And then further testing showed that this is also a communication protocol issue because when I was able to saturate the throughput running multiple tasks through a Thunderbolt hub, Boom! It started to perform much closer to advertised speeds. I’m working on additional tests on the M1 Macs with the beta version of Monterey to see if they’ve cleaned up this communication protocol and basically push that throughput when it’s called upon. 🤷♂️ But I love your perspective. As a healthcare provider I look at tech from a Anatomy and Physiology perspective, and then working around the pathophysiological issue. When I start mixing mitochondria and multi threading, watch out! 😂 Glad to have you around.
Thanks for the TB3 test, the performance is still restricted, most probably (1) to clearly delineate the IPad / IPadOS from the other M1 proper computer devices, (2) to limit the current draw from the TB3 poet and (3) share a common denominator with the USB3 / USB4 specs In the WWDC, Craig demonstrated, beyond all doubts, that APPLE intends that users have 3 devices, the M1 IPad Pro, the M1 MacBook Pro and the M1 iMac as their workflow using Universal Control features APPLE is just repurposing and reusing the M1 SOC in the IPad Pro from a cost and production point of view, not as a product differentiation Think of the power of M1, the 8 GB RAM and the TB3 on the iPad Pro to made it a “PRO” device without actually being used professionally….. it’s still for consumption and won’t cannibalise the sales of the other similarly specced M1 devices 3 product with 3 distinctive OS, ie IOS, IPadOS and MacOS, for 3 distinct categories of buyers for 3 types of different needs or use case Cheers
Yes on the points. And most definitely a yes on cost savings by using the same chip. I mean I’ve been in that position trying to figure out how to streamline and utilize similar pieces in multiple product lines. I think this WWDC solidified that Apple very much wants us in each Sandbox in their walled garden.
All we wanted was mactouchOS to iPad Pro not an accessory to “MacBook” with universal control, tablets is the future not laptops. We good if you got 2018 because M1 Pro superman with kryptonite belt
(Corrections): I did want to make a correction: As swap memory is available for M1 Macs (really all Macs/PCs), iPadOS seems to limit this functionality to actually "killing" the open apps to free up memory, or rather just handling this a bit differently than actual swap as we know it on a full operating system. I do believe this is possible and as more info comes out about memory utilization we could potentially see this functionality.
This is a bit of a repeat based on some questions I've received about iPadOS 15. I did however find something I wasn't expecting. Testing will continue and I'm experimenting with some other beta released apps. I'll keep you posted. Go rock faces.
I love the lack of histrionics on this channel. Keep it up 🤘🏻🔥 💻
Hey we could run some lines together if you want? 🎭 😀
Ha! Many thanks friend. I try to find that balance.
@@KevinRossRN Haha! The balance is there 😎
Amazing video. I have the 128gb 8gb ram model and this video made me feel a little more comfortable for not having to upgrade to the 16gb ram right away. Especially with the delays of m1x. Definitely gained a sub and cant wait for more videos.
Yeah, you are definitely fine with 8GBs of RAM. Enjoy your device and glad to have you here. 🤘👊
You’re running on the 8GB?
Great to hear. And I know, it’s pretty awesome for a first beta. So far so good on my end. Glad to have you here with your workflow. Always helps to hear from others on how their experience is going. 👍
Then I’ll be seeing you around. 😉
Many thanks.
New subscriber. Great job! Didn't really understood if the app still have the 5gb Ram cap. Thanks
Glad you’re hear. Maybe this cap will be lifted, eventually 🤷♂️
Thanks for the update, Kevin! I know you’ll keep us informed. BTW, the background music is perfect. It’s laid back and at the proper volume, so not to be distracting. 👍
I appreciate that. I’d spend more time researching tunes, but do try to keep it at the pace and feel of the video. It actually takes way more time than people think 😀
And yeah, just enough volume on the track to know it’s there, but not overwhelming to the vocal track. 🤙
The best review yet !
Thanks so much! I’ll keep them coming.
Thanks Kev. Great topic as usual.
Thank you so much 😊
Thanks a lot for this very professional and useful serie! (testing M1 iPad Pro) :)
And thank you for hanging out. 🤙
12:19 Well that’s different! Thanks for the follow up. I haven’t checked, but I wonder if there are any new APIs in iPadOS 15 that would help speed this up? 🤔
I’m digging further, but if this is starting to become more a out communication protocol and not as much a out hardware, then it’s possible, but I still think hardware is a limiting factor, but I’m all for proving myself wrong.
great video Kevin, super happy to see an increase in transfer speed this also gives me hope that they can keep pushing it with more updates down the line, would you agree? looks like it hit a wall at 950mbs.
Yep, I be continued to test and haven’t gotten above that threshold, but heck of a lot better than 550.
I really want to test the render and export on the SSD.
I’m trying to see if I can find any information as to why this change occurred via the software.
@@KevinRossRN you are also the first to talk about this I also cannot find any additional information besides your video so well done!! appreciate the work
Many thanks friend. Testing is always fun, and even though the results aren’t always what I’d like to see, this one was a nice surprise. 👊🤘
Very cool. This channel has been fantastic for me to get some data to base purchase decisions on. I definitely want to update my tablet game, but I’m very content to just wait a bit before taking the M1 iPad plunge. I’m enjoying the crap out of my M1 Mac Mini and it’s going to be awesome to see the development path that comes from all the Apple Silicon innovations!
Hey, you keep enjoying that mini. That thing is my rock when it comes to content creation. We’ve been through a lot together.
Awesome to see you friend.
@@KevinRossRN I agree…I have only been able to get the fans to activate ONCE above baseline: I engaged ‘Flex Pitch/Flex Time’ on Logic on a VERY long stem. It only jumped 10 degrees C then promptly went right back down to its normal idle temp! 🤣😂 It’s a beast.
I know it’s so amazing at how quickly it can dial those thermals in without a jet engine fan. 🤷♂️
After the WWDC 21 show I staying with shooting on my Sony ZV1 and editing on my Mac Mini M1, so for the iPad Air 4 seems to work for me. Between the $200 and the $60 difference in AppleCare for me $279 with tax I don’t see enough of a difference with iPad Pro M1. Hopefully I agree after iPad OS 15 is available. Great videos. I am really I came across you channel.
Awesome to have you here and in my opinion, you’ve got a great set up already.
That’s great to see the throughput results! Almost double the speed, that’s very encouraging- great thorough video again, love it! I did notice something odd on a TB4 hub I’m testing so this is great news!
I’m interested to hear what you’ve found on that hub. It seems that there are certain drives that aren’t recognized on the M1 IPP.
I’ll be testing the hubs that I have. But yes, encouraging that the throughput may have been “fixed,” or at least better.
Awesome videos sorry I’m sooo late
Never late friend. Good to see you.
I learn a lot every time with your vids!!🌟
That’s the name of the game 🎓📚
Glad it helps friend. Appreciate you 🤘
Thank you very much for your helpful reviews. Not only Thunderbolt 3 drives but USB 3.2 SSDs are slow too. On a Samsung T7 USB 3.2 10 Gbps SSD I got only 44 MBps copying a 5 GB file to the HFS+ formatted SSD using the original USB 3.2 and a TB3/USB4 cable. Did you see any speed improvement with USB 3.1 and USB 3.2 SSDs too?
I did actually. I’ll pin this comment to help others. So on a USB-C Gen 2 (SanDisk 1050MB/s read/write) I was getting a sustained 500MB/s, which is up from the previous ~250 - 300s.
@@KevinRossRN I formatted the ssd to exFAT and the write speed increased to 70 MBps
How about APFS?
@@KevinRossRN Didn’t tried yet, because I’m running my Macs with El Capitan.
Oh, gotcha 😉
Great video and showcase of the performance, just out of curiosity which type of cable you are using for this testing? Thanks!
It’s a passive cable that came with the enclosure. I’ve tested others. The cables that come with those enclosures are pretty sturdy…better than others.
That’s awesome, glad they got that fixed.
Let’s hope we continue to get improvements.
Hi Kevin,
Just a couple minutes into this video, I see your beautiful poster “Audio Apparatus”!
I will watch the rest of the video now, hoping for a better angle/close up of the poster. 🎶👍
Ha! Maybe I’ll add that into a studio tour 😄
@@KevinRossRN Please do🥳✅👍 I will Google it and see if I can find it😉
Are you wanting to purchase it?
@@KevinRossRN will definitely want to see a close up. Thing is, it might not fit (size wise) my office wall…🧐✅👍
Check out PopChart Lab.
Really enjoying your very thorough work. Just subbed. I have one question. I am strictly a stills photographer, and I work almost entirely on my iPad Pro M1 512gb.
What would you recommend as external storage? T5, T7 maybe, Sandisk Extreme or, something else?
Thanks. So as far as an external SSD, I like all of those options. The SanDisk Extreme is one that I use quite often as well as a DIY option that I reviewed recently from Sabrent. But honestly you’re not going to go wrong with any of those options. I’d search for the best price, especially now. Either T7 or SanDisk Extreme should give you slightly higher read and write on the iPad.
@@KevinRossRN Thank you very much, Kevin!
You got it 🤙
About the transfert speed, iOS 15 do it better than iOS 14. But TB3 allow 5GB/s right? So why we still lock to 960MB/s, that’s the question. As an owner of iPad Pro 2020 and care about speed transfert, I m reassured to keep my iPad. I will receive my SSD NVME this Thursday and will do some tests that I will share with you.
This has been the issue with these TB controllers on the M1's. I've actually got another video coming out on the M1 Mac with the latest operating system. Working on that today.
There's going to be some bottleneck through the cable and enclosure along with the rated read/write speed and cache of the SSD, but yeah, 960 is still slower than it should be with this particular setup I have.
Feel free to circle back around with your updates. I'm definitely curious.
@@KevinRossRN i send back my NVME Disc, I don’t know if it it the case or because I can’t update my disc on windows desktop.
But I was restricted to 20-60MB/s while transferring in and from my iPad Pro 2020.
So I think the NVME is not a good value and I bought an 1TB 2,5’ SSD Crucial MX500.
I already get one and the speed sharing is up to 300-350MB/s.
I wait here, keep what I have and hope a better iPadOS the next year.
I wondering if I will bought a surface. Because I like the windows files management, better than iPadOS.
Well I’d say the surface is a whole different machine…basically. I think it’s a great device and would recommend it to anyone that wants that full operating system and familiar with Windows.
Hopefully with the updates we’ll continue to get better performance. These chips are awesome, and will only get better.
Nice, hope that the thunderbolt speeds on the macbook will go up too😎 thought i‘d have to wait fir the next chip for that😅
I’ll definitely be looking into this. Keep you posted.
For context, I have verified (using JazzDiskBench) that the read and write speeds on the M1, 256GB, 12.9 inch iPad Pro are almost exactly 3,000MB/s and 2,000MB/s, respectively. With that being said, these 2 things are confusing the hell out of me:
1. The iPad Pro (even in iPadOS 15) gets 5Gb/s speeds rather than 10Gb/s speeds via USB 3.1 Gen 2
2. The iPad Pro (even in iPadOS 15) gets 1GB/s speeds via Thunderbolt when it should be getting the full 2GB/s
This is one of the situations where I’m just like… WTF?
Welcome to the head scratching club of M1. 😀
I’ll have to check out that benchmarking tool.
I’ve got some additional tests that I’m running, however this also speaks to some of my other tests I did on the Mac where I could essentially get the same results in video rendering and export from a USB-C SSD vs the Thunderbolt enclosure with a very fast NVMe SSD 🤷♂️
The communication protocol is definitely off.
@@KevinRossRN Yeah, I’ve considered going the Thunderbolt enclosure route for my external M.2 SSDs, but they’re way more expensive than the 10Gbps ones and it seems there’s little real world benefit. Real world speeds on the 10Gbps enclosures are ~925MB/s, and I can’t image I’d need anything more than that anytime soon. I prefer the ASUS ROG STRIX Arion despite its obnoxious appearance. Reliable brand and great heat dissipation.
Regarding the JazzDiskBench, I had to run it a few rounds before getting peak speeds. The first runs were a little bit slower for some reason.
Interesting on that benchmark. I’ll be mindful of it.
And yes, the one benefit of the enclosure is that you get to choose the enclosure and/or the SSD, but agree that the real world difference is pretty insignificant in most use cases.
I still use these Gen 2 SSDs all the time and recommend them often.
Does blooming affects notetaking on dark background to that extent that i should choose 2020 ipadpro???
The blooming is definitely present, but it’s really a matter of whether it would bother you, I suppose. I see it if I’m looking for it when the room is dark, but personally it’s not something that I’m concerned with. But you can certainly save some money with the 2020. As far as performance, you’d likely not notice any difference.
If it’s less than 50% then that shows right there for this performance we need the 16GB. Thanks for the test and the video.
Well it can easily be swapped on the SSD, but then there’s the “what if” factor that at some point an app is going to use more than 5GBs (per app) than what we already have.
Appreciate you hanging on this one.
@@KevinRossRN 👍
I would like to see what the numbers are now that iOS 15 is officially out.
There really isn’t much. The RAM is app dependent, and more developers just need to make adjustments to utilize that RAM, but it won’t automatically happen in iPadOS15. As far as the Thunderbolt performance it too is pretty much the same. I’ll continue to test as more updates come in.
There is actually a website to get the dev beta profiles without being a developer. Google it using safari from the device you want to download it too :)
Yeah, I’ve seen that before. Didn’t see it this round, but I know it exists. I like having access to the developer forums and the other resources.
Have you tried the beta, or thinking about it?
@@KevinRossRN you are right my friend when you have that account you get more resources. I have iOS 15 on my phone and on the watch it’s pretty stable for a dev beta
@@KevinRossRN I’ll probably get it when I get my iPad thanks a lot for your review this helps a lot
That’s awesome to hear. Haven’t tried those yet and I’m trying to get my Mac ready for Monterey, but researching all the bugs first.
Happy to help, and sending my best for when you get it 🤙👊
This is their very first beta. I expect a lot more changes.
Yes, first beta. As I update I’ll certainly keep people posted.
Have you tried the short Apple Thunderbolt 3 cable yet?
The cable I did use was short. I’ve been testing these ports on the M1s since last year and used several variations throughout my testing.
It’s a known issue 🤷♂️
correct me if i'm wrong, but ipad and ios does not have swap as far as i remember, means there's no caching the memory into the ssd..
This is a great point and something that had plagued the previous versions.
So yes, iOS/iPadOS worked more in a virtual memory capacity and basically “killing” the apps to free up memory rather than a typical page in/page out scenario (with swap) that occurs on MacOS, or really any full operating system.
The interesting thing here, and yes this would seem to be OS dependent, I agree, but the interesting piece here is that I’ve had apps opened for weeks and when going back to them whether suspended…I can’t say for certain, but they seem to still be active because they don’t auto refresh, even if I call up several other apps that seem to at least want to utilize as much of that max 5GBs of RAM.
So while maybe the general use of the term swap may not be appropriate at this juncture until we get more details on this chip and how this OS is handling that unified memory, I’ll take the hit and correct it to say it could be that same old virtual memory as it has been, but the question still may remain as to how it keeps these apps suspended without a full refresh and/or killing the app…at least in most cases I’ve found because I know that iPadOS would likely need to reallocate that RAM, especially as additional multitasking is starting to become supported on iPadOS 15.
I appreciate you posing this question though as I’m always thankful to have additional perspectives here.
I’m continuing to look into this and will certainly provide a follow up. 😉🤘
Are the iPad Pro 2021, the Acasis enclosure and the WD SN750 all using 4 PCIe lanes? There is another cheaper TB3 enclosure the OWC Envoy Express but only supports 2 lanes.
Yes, the controller in the enclosure is Gen 3 (4 lanes) and the M1s will also have 4.
I wasn’t aware of the details on the OWC. I’ll have to dig into that. Not sure which controller they’re using in that one.
@@KevinRossRN But is the M1 iPad Pro also using 4 lanes? Maybe only M1 Macs and MacBooks?
They’re all the same friend. Apple is utilizing many of the same parts in these devices as a cost savings. Even those Intel Retimers.
You also have to consider the communication protocol with each enclosure. This is why some work better than others, so definitely research what controller/chip on the board these manufacturers are using.
@@KevinRossRN How did perform the WD SN750 APFS formatted in the Acasis enclosure? Do you believe a faster NVMe with bigger DRAM cache could perform better on the iPP2021? I appreciate very much your efforts to figure this speed limit out.
The SN750 has performed quite well across all my M1s. Even though it’s a Gen 3 it performs as well as a Gen 4 SSD that I have from Silicon Power.
You’re really only going to notice that cache when you’re transferring TBs of information (sustained). Unless you’re using the IPP as some kind of Server, like if you’re running some RAID array off of it, then you might consider a much larger cache.
But, after that buffer at around 14GBs the write would typically drop from about 2800 to somewhere around 1600, however with that throughput I’m maxing out at around 900-950 on the IPP, and so far I haven’t been able to get any higher than that right now.
iOS is not cut out doing MacOS's job on the iPad.
That is correct. Just unfortunate we have a desktop/laptop class chipset and the apps can’t even take advantage of it.
@Kevin Ross
Have you tried connecting the iPad directly to the Mac mini and transfer files? So no external ssd’s, just the Mac mini ssd.
Best,
I have connected them, but it’s a pretty inefficient way because iPadOS and MacOS have this built in sync functionality that really makes it difficult for a drag and drop solution. It’s really not intuitive and more of a pain.
@@KevinRossRN Thanks, I agree. The sync functionality is not yippie… How about, still connected directly over Thunderbolt 3, setting up a shared folder that you would be able to read/write to and using an app like ‘FileBrowser’ on the iPad? The app has various connection methods.
Not sure if the direct (Thunderbolt 3) connection would ‘force the dreadful sync functionality’ though.
Still without the direct connection over Thunderbolt 3, a (direct) connection over Thunderbolt ethernet might be interesting to test, same app, ‘FileBrowser’? Only if you have the needed software/hardware of course. Don’t go spending any money just for this specific scenario ✅👍
Best,
I figured this t-bolt issue had to be a software problem. This is based on by knowledge as a lawyer, not any technical skill. Why? Because Thunderbolt is a proprietary brand of Intel. Brand standards are not the same as engineering standards. Engineers specify hardware performance. Brands specify performance to the ultimate consumer. E.g., if you buy a Burberry coat manufactured in any factory anywhere in the world you know the seams will be straight, drape consistent, stitching even and very tight. The same is true for tech in spades. neither Intel nor Apple will allow a product to carry their name brand without extensive qualification. This will always extend to the whole system to make sure the branded component will actually perform as advertised. While it's possible both the Intel and Apple engineers blew it, it's not the way to bet. So if it's not hardware, it has to be software. I'm betting that the power management throttled the transfer speed because due to the unique design the iPad was always operating on battery power. Also, if you are not doing anything other than transferring files, the power management (I think) assumes you're not actively using the ipad. this would explain (maybe) why the performance is better when you run other tasks along with file transfer--the power manager thinks "Hey, Kevin is really working hard. Need to boost the performance!." Be interested in your thoughts after you have more time. BTW, your channel is great for an ignorant person who is really trying to understand video. thanks.
Appreciate you dropping by on this one. And yes, after extensive testing on the M1 Macs I kept hitting a wall, and waiting for software updates.
After digging in (to your point) I did find that Intel Retimer in the tear down and figured they needed this because the controller being on the chip and a low powered one at that, they then had to put these near the port to equalize the throughput.
And then further testing showed that this is also a communication protocol issue because when I was able to saturate the throughput running multiple tasks through a Thunderbolt hub, Boom!
It started to perform much closer to advertised speeds.
I’m working on additional tests on the M1 Macs with the beta version of Monterey to see if they’ve cleaned up this communication protocol and basically push that throughput when it’s called upon. 🤷♂️
But I love your perspective. As a healthcare provider I look at tech from a Anatomy and Physiology perspective, and then working around the pathophysiological issue.
When I start mixing mitochondria and multi threading, watch out! 😂
Glad to have you around.
Thanks for the TB3 test, the performance is still restricted, most probably (1) to clearly delineate the IPad / IPadOS from the other M1 proper computer devices, (2) to limit the current draw from the TB3 poet and (3) share a common denominator with the USB3 / USB4 specs
In the WWDC, Craig demonstrated, beyond all doubts, that APPLE intends that users have 3 devices, the M1 IPad Pro, the M1 MacBook Pro and the M1 iMac as their workflow using Universal Control features
APPLE is just repurposing and reusing the M1 SOC in the IPad Pro from a cost and production point of view, not as a product differentiation
Think of the power of M1, the 8 GB RAM and the TB3 on the iPad Pro to made it a “PRO” device without actually being used professionally….. it’s still for consumption and won’t cannibalise the sales of the other similarly specced M1 devices
3 product with 3 distinctive OS, ie IOS, IPadOS and MacOS, for 3 distinct categories of buyers for 3 types of different needs or use case
Cheers
Yes on the points. And most definitely a yes on cost savings by using the same chip. I mean I’ve been in that position trying to figure out how to streamline and utilize similar pieces in multiple product lines.
I think this WWDC solidified that Apple very much wants us in each Sandbox in their walled garden.
Have you got 1 really big 50GB file instead of a folder full of files?
Plenty. It spikes to about 950 for a quick second and settles at around 920-925MB/s for a sustained transfer/copy.
All we wanted was mactouchOS to iPad Pro not an accessory to “MacBook” with universal control, tablets is the future not laptops. We good if you got 2018 because M1 Pro superman with kryptonite belt
Yeah I think this was Steve’s wish to never have that functionality. Must be written in the Apple archives somewhere 🤷♂️😉
@@KevinRossRN At this point only a fool would buy the M1 iPad Pro now or next year version 2022, when the 2018 works just the same, lol.
The 2018 is still plenty iPad for most. Or really, all. 😄
@@15heartz not really...If you're a heavy gamer, you can tell the performance difference between the 2018 and the 2021
Are you rocking any games right now? Always looking for new stuff to test?
Stop talking with your hands!!!
Looks like we need a hug over here 👆people.
@@KevinRossRN no you need to stop talking with your hands!
A hug is on the way friend. Sending my best. I’ll head over to your videos and provide some loving support. What topics to you primarily cover?