I never, never purchase any external SSD that does not have a detachable cable. A detachable cable is easily replaced if damaged. If a permanently installed cable is damaged you are screwed.
I would personally like if it had something like a captive lock for an included cable, so you can detach it but really lock it down, so it’s really solid when connected.
I have worked on computers since the late 80's. Unless you leave the ssd side always attached, I find the replaceable cable ones to be less reliable. The tiny ssd side connector will either lose the solder connection, or will rip out of the board. If the drive with a cable with proper strain relief is installed, that will last a good bit longer. At least in my experience it will, and I have serviced 1000's of professional clients who use external drives for various reasons. I have to replace or resolder my drive enclosures that I use for booting to Macs, on a pretty constant basis. Generally every year or 3. I have 2 WD usb drives with the built in cable and both are fine after 5 maybe 6 years.
@@otwer22or … they just sell branded separate cables that they advertise as the only way to get max speed? But you may use your own cables at “guaranteed reduced performance?” Not hard. Don’t defend planned obsolescence and lack of end user repair.
The Sabrent TB5 throttles down to 1.5 GB/s in extended use in their own YT vid ! The OWC Envoy TB5 throttles down to 1.8 GB/s You can actually get better sustained performance out of TB4 / USB4 drives with active cooling TB5 without active cooling is essentially a waste of money
@@siddhipriyamishra7210 For sustained performance (transferring 100s of GB frequently) a USB4 enclosure with a fan from the likes of Anyoyo, Yottamaster, Orico, Acasis (not their dual drive enclosure, that's only fast if you RAID the drives) coupled with a suitably fast PCIe 4 SSD (like above 5000MB/s) that you can re-use in a cooled TB5 enclosure later. Current TB5 drives and enclosures throttle down so hard when pushed that they're simply not worth the money IMO. They'll be very fast for lighter workloads though, but then there's the question wether you really need them or want to spend the extra $$$ on them. Bit disappointed with Max Tech for not illustrating the throttled TB5 performance, while other reviewers have shown it - even Sabrent themselves !
@@siddhipriyamishra7210 Well, if you don't use these drives for extended highspeed read/writes, max 100 GB, you can use these and they'll be very fast The OWC Envoy TB3 (yes THREE) is about 200 MB/s faster than their TB5 drive on sustained writes ! A USB4 enclosure with active cooling will be faster than either TB5 (or the OWC TB3) drive during sustained heavy use You can find these from Anyoyo / Acasis / Orico / and others Or like YenYun states, wait for actively cooled TB5 enclosures - they will be coming
Please do larger transfers (600-1000GB) and measure WHEN the drive slows down. Please also use a software where you can monitor the R/W-speed and get the average speed of the transfer. Always gets frustrated when newer drives that is supposed to be “faster ” actually performs worse over time than older “slower” models. Thanks!
All external ssds will eventually slow down during long writes due to thermals or when cache fills up. Only difference is that some handle it well and other become almost unusable. If a drive drops from 900MB/s to 500MB/s it’s fine! But from 900MB/s to 80MB/s or worse isn’t acceptable in time sensitive situations.
Sustained performance is mostly a function of the underlying drive (QLC vs TLC, DRAM cache, etc). The enclosure's controller mostly affects the peak speeds, though the cooling can have a big impact as well. Personally, I think the best play right now is an OWC 1m2 empty enclosure, with a 990 Pro or SN850x drive inside.
@joelv4495 personally I use the Seagate Firecuda 530, jerryrigged into an open USB 4.0 enclosure, and tied down with a Velcro strap. Its the fastest sustained speed I can find. 2.9GB all the way through, past SLC, unless overheating, but I squished a copper heat sink on top of a silicone one, with also some tape meant for even more heat dissipation. It's the proudest and fastest thing I've made, and I don't think about upgrading to TB5 enclosure yet, unless it comes down to a price of less than $2000
My massive sound libraries for music production are now reading at 5100++ MB/s versus my old Samsung T7's 900 MB/s. This results in less latency, etc. I love my Tb5 OWC Ultra.
Nice! I was thinking about buying one for this exact purpose! Really glad that it makes a difference! I am planning on getting a 2TB version of the Mac mini m4 pro then buying another 2TB external solely for virtual instruments/sample storage (I have about 1TB of that right now and my internal 2TB is nearly full).
Thanks for sharing! Glad to know that there is a good use case that actually helps compared to most other things! And even though its expensive, Apple also charges a ton and you don't get to use internal storage with multiple devices.
But, that average is no longer average. It’s getting harder to find new computers with SATA internal drives. Pro models have all gone NVMe because of the faster speeds. It has been years since Apple sold any internal SSDs that were SATA. I was shopping for internal drives and NVMe is essentially as cheap as SATA now, but is many times faster and smaller. 400MB/sec now seems quite slow to me, given all our current choices. The internal SSDs on the Macs in this test go 5000-7500MB/sec…15 times faster than a SATA SSD, and that kind of speed is standard on new PCs now too. There’s almost no justification to pay for slow SATA today.
when having SATA HD or SSD it's better to boot from an external drive: even over 10Gb USB... when not stuck with 5Gbps USB ports. Oh... desktops have expansion slots! :) And for Apple users: booting from an external drive over TB2/TB3 adapter makes a real difference.
I shopped w/in Thxgiving week, Gen4 SSD 4 TB inside Thunderbolt 4 aluminum enclosure now can reach 3400 read / 3400 write transferred to M4 MB Pro for total near $370. SSD being WD Black SN850X 4TB inside Qwiizlab Fanless 40Gbps USB4 Thunderbolt Aluminum SSD Enclosure. I consider this TB 4 combo valuable for years.
That Sabrent TB3 enclosure only supports 20Gb/s connections which makes the TB5 drive look disproportionally fast in comparison to what most TB3/USB4 enclosures are capable of. The value proposition of the TB5 drive makes absolutely zero sense once you factor that in, especially when you consider the thermal throttling that fast modern NVMe suffer without active cooling.
What if you’re a video editor or a large volume photographer and you mainly care about read speeds as that is what speeds up your workflow? Copying is only needed once, but reading is what actually matters for most people
Guarantee you that if you open up the OWC TB5, 95% of it is just cooling. It's similar to external 10Gb/s nic cards. 95% of the enclosure is just cooling cause you typically transfer a lot. It works, but TB4 SSDs is great.
@@MaxTechOfficial How many times per day are you wiping your project folder and re-copying it? the folder could be 4tb but if you only do it once at the start of a 200+h edit what does it even matter?
i use the ACASIS 40Gbps NVMe Enclosure w/ fan and I get roughly 3000MB/s Read & 2950+ MB/s Write on my Mini M2 Pro. This is on TB4 cable. I have a WD black 4TB drive installed as a boot drive.
I actually opted for OWC's 1M4 enclosure, as it's beefy like the OWC drive here. The other nvme tb4 enclosures usually had fans that reviewers said were too loud, and I didn't get the warm fuzzy that they were reliable either. I opted to pay for OWC's enclosure (only on sale haha) and it easily maximizes the 970 Evo I put in there to run as a boot drive for an old iMac
I honestly don’t see a huge use case for one to transfer a ton a data from internal to external at those speeds. Typically if one does a transfer it’s to archive and the speed is less critical in that situation. We mostly have limited space on Macs due to cost. That means there likely isn’t a huge amount of files to transfer anyway. If we do happen to have 4 TB internal there is then much less need to use external unless it’s again to archive or backup. Buying a TB5 drive seems like a very expensive option for archiving. More likely we would buy a TB5 drive to actually work off of for the added speed which means a small chance we will transfer a large amount of data. From a direct camera recording drive it’s also more cost than necessary in my opinion. A USB4 drive is more than fast enough for external video recording and as a source drive to edit from. There is the transfer time to move that data to an edit drive if the camera drive needs to be used for another shoot. That might take a bit longer but again I don’t see that as a deal breaker at all. I have a hard time even justifying faster than USB4 from anything that would transfer data around.
It has been said that Thunderbolt 5 will be welcomed by every digital imaging technician (DIT) working TV and movie sets, because they do in fact need to transfer very large amounts of data daily under a deadline. And the cost of 10 Thunderbolt 5 SSDs will be one of the smallest line items on any of those productions.
You make some great points about the practical use cases for faster drives like TB5. For most workflows, speed is often limited by other bottlenecks, like the SD card itself. For example, even with my Sony CFexpress cards, the transfer speeds I get are just slightly above SATA SSD levels in real-world conditions. When it comes to editing, I’ve found that the difference between SATA and NVMe drives is noticeable only during specific tasks, like opening large projects or caching previews. Once you’re in the middle of a session-editing timelines, scrubbing footage, or color grading-the performance gap becomes negligible. In fact, editing off a NAS, which is fairly standard in many setups, feels almost identical for most tasks after the project loads. I agree that TB5 would be a game-changer primarily for those working directly off the drive where the speed really matters, or for scenarios with huge datasets requiring rapid transfers. For archival or standard external workflows, USB4 is more than capable and offers a much more cost-effective solution. It's all about matching the gear to the workflow, and in most cases, TB5 is likely overkill unless you’re dealing with extremely high-end, speed-sensitive tasks.
can't say thanks enough for not only this review, but alllll your content the last decade. you've single-handedly help me make the most informed decisions for my media company, helping underline the most important performance criteria that we prioritize. appreciate it immensely! question for you on this TB5 drive: would love your insights on specifically lightroom speeds. we all struggle with LR performance, even with highly optimized systems. curious how this performs hour+ into a LR editing session? I still struggle with Sandisk G40 (which has great r/w speeds) and it's just a matter of time before I need to optimize catalog/restart app/reboot computer etc to 'wake up' lightroom. when we get back from big shoots and have 1000s of images to edit, these speedbumps with LR can be not only frustrating but time-consuming. Would love any input you may have on how TB5 Envoy here could be helpful for us hybrid shooters, who do a lot of LR room for several hours at a time. thanks m8!
A lot of people missing out on the point. The TB5 drive (while it is indeed much more expensive now, which mostly all other new tech are), does not only shine on file transfers. With the price of upgrading your mac's internal storage, this is a great alternative. Process your files from this drive directly without having to worry about latency. Again, arguing that this is not worth it because you're not doing continuous file transfers can just use old, cheap, and reliable drives. Those who want to extend useable storage space for actual direct file consumption (photo, video editing, etc) would benefit using this. "Don't buy this" my ass.
Your testing is for a limited use case though? If you're using the SSD as a working drive for a heavy 4K (or 5,6,8k) timeline that speed advantage is definitely going to help. Sure, for transferring files, they're all reasonably fast...
I use it with the new Mac Mini Pro and for music production (All my software are stored on the OWC) and its just so much faster than my Thunderbolt 4 drive. A no brainer.
OWC is a great company. I bought an internal SSD from them for my 2012 16" MBPro a few yrs back, and it was night & day difference in speeds when booting up etc. after I installed it.
You pay a premium for this first TB 5 box with an SSD drive. Just wait until there is more competition and until then TB 4 with speeds around 3500 MB/s is not such a bad alternative.
Nope for the T5 right now. I've been a happy user of a Sabrent 1TB Rocket XTRM which is a T3 but so reliable, fast enough for video editing. I don't use the silicon case and I elevate it a little from the desk so even if it gets warm, it never crashed or caused any problems. It's connected all the time to my M1 Mac Mini. I believe I'll get another one that's a Sabrent Rocket XTRM Plus 2TB for the sake of space on it, it'll only be for editing actual video projects. You US peeps are happy for the size of your market, this SSD is $274 over there, in Sweden where I live it's bloody $354, "thanks" to Swedish import taxes ..
Great review. For me, I only rely on OWC external drives. I thought about getting a TB5 OWC Envoy Ultra with the intent of putting core content files on the external drive (to switch between a Mac Studio and the new MacBook Pro M4 Max. Two problems: 1. Real-world data transfer is incredible, but I think 3000+ MBs using the TB4 OWC Extreme may be good enough. 2. The short integrated cable doesn't work as I switch between systems (based on my cabling set up). Also, I prefer separate cables that I can easily replace (longer is better). And a third problem I can't explain: I used an external OWC 1M2 Express for core data (instead of the internal SSD). I set up Time Machine to do backups when plugged into Mac Studio desktop. After a couple of hourly backups to another backup external disk (also OWC), Time Machine failed. I rebooted, reformatted the Time Machine backup disk, and tried automated, background backups again. Failed!!! if I remove the @OWC Express from being part of the TM backup, it never fails. Any ideas???
I don't really need the super speed so I've been having good luck with the Samsung T7 drives which would compare with the Nano. Only crazy part is the price which changes all the time on Amazon in Canada. Can change up to $75 from week to week and even just between the 3 colors of the same size drive!!! Last week, the gray 1TB was $124 and the blue 1TB was $199.
If you want small andcheap, look at the Samsung T5 USB drive. Right know you can get 2TB for as little as 60$ on sale and that drive still packs 400MB/sec.
If the Sabrent Rocket Nano is a USB 3.2 Gen 2x2 drive, more than likely the speeds will still be capped to lower speeds, isn’t it? At least until the M3 generation Apple didn’t support 3.2 Gen 2x2. I’m unsure about the M4 generation. If that’s the case, buying a Samsung T7 that will have the same speeds and same storage for a cheaper price seems like a more reasonable approach.
The confusing element here is not knowing what Gen the SSD sticks are in each of the enclosures. I'm not sure how the TB converters in the enclosures interface with the PCI lanes on the SSDs. The number of lanes varies between the Gens (3,4 & 5).
compare it to a zyke drive with a fast Gen4 NVME drive or an TB4 casing with active cooling. Yes thermal throtteling is an issue. Also an issue is power consumption in sleep mode (closed lid e.g. over night)
While it's not there yet, value-wise, it does suggest you'll want Thunderbolt 5 in the next couple years. One more reason to choose the macbook m4pro over the m4.
I'd like to know how much the external SSD can replace internal SSD. For a laptop, it is inconvenient to carry an extra thing dongled to the machine, but the price of internal SSD on a macbook is high. More to the point though, let's talk about Final Cut Pro, or Logic Pro. Not small projects. It has been said in the past that having the project data on a separate SSD to the application is helpful for performance and I am wondering if that still holds true for internal speeds these days? Also noting that the internal speeds of SSD's vary a lot between machines. Anyway, can you do performance tests of these fast thunderbolt 5 SSD's in a real world scenario. In particular, with Final Cut Pro? See how much difference it makes compared to an internal-SSD-only project, or just to see how much SSD I really need internally if external SSD is fast enough for.... say .... a project in 4K60 with multiple layers and noise reduction, optical flow, whatever. The real world performance difference of SSD's with real world projects is very interesting, as compared to raw MBs specs.
Having purchased a base Mac Mini M4 Pro and wanting to put a 1TB external SSD for apps and games, what is better to get? An SSD with USB 3.2 or a USB with Thunderbolt?
What was the r/w speeds on the M4 Macbook to start with? if its a lower capacity SSD you would be throttled on how much data can be read from or written too
Very interesting test but I disagree with the conclusion. In the real world people don't just use these drives to transfer large files once in a while. People work off these drives and I can see how a slower one can be a real bottleneck. It would be interesting to see how these drives perform in a real world situation, like having a photo library on them, importing a bunch of RAW photos and do some editing and exporting.
How direct video editing from SSD drive effects drive? What are good speeds for drive for direct video editing from it? You mention that SABRENT NANO is enough for video editing, but how we can know that because footage type can be different ✌🏼
Hi. A bit of an unrelated question but im hoping you could help. I have an old 1tb external hard drive that i believe has a bad port. I have tried 2 different cables and sometimes works but keeps connecting and disconnecting. The cable has a bit of play in the hard drive port which i think is causing this. I have ordered new 4tb hard drive but i am wondering is there anything i can do to stabilise the connection so i can transfer the files? Sketchy or not. I just want to be able to get the files off it. Thanks.
Great video Max, BUT! Are there any differences in editing videos off of the new TB5 speeds?? I'm currently working from a Samsung T9 (2GB W/R p/Sec) - considering upgrading to the OWC Envoy Ultra TB5 SSD - putting the money on the side, how much of a difference would I get by editing off of it?
Both TB5 drives throttle down to below that speed in sustained operation Sabrent after about 100GB, they attribute it to drive cache in their own YT vid Look up YT reviews which test them over longer times, instead of these advertorials
Nonsense, the OWC is amazing and if you are editing High-end 8k RAW 16bit video there can be nothing else. Having said that of you never going to be doing that then I guess it's good advice.
What about iCloud Drive? Is the data from an external SSD being mirrored on iCloud Drive as well, the way it happens with the data on your internal SSD?
looks like there may be some disinformation here. Looks like the cable is not built in. Looking at the enclosure, it looks very similar to Angel Bird card readers. Meaning that the cable is meant to fit tight, so if you give it a firm pull, the cable should come out. Please report back to us V!
Why? That's longer than completely filling an empty drive to 100% capacity; what use case needs an external SSD to go from empty to full, then empty and back to full again in a 10-minute time period?
@@dontbekurt that really isn’t that important to me. I see the results. It gives me a good idea of how things will work for me. if he tested like for like. Then cool. But if not. Oh well.
@@redesignedlife777 That's just it.. I don't know if these tests are indicative of how it would really work because of the fact that all of those things are different.
In fact MaxTech and all reviewers need to stop saying 120Gb/sec when talking about data, because no one will ever sustain that for data. 120Gb/sec is an asynchronous burst mode.
@ it’s still annoying to do this every time, especially if you watch on auto play/playlist, on a device without capability to touch anything or if you just can’t touch it. And also the translation is so bad from the voice and the translation itself, so it’s not even a benefit for people which need this.
The Apple cost from 2 TB to 4 TB is more than the cost of one of these 4 TB drives. so having 2tb internal and 4tb external gives a larger storage pool for less money.
@@Tappits84 I see it is a $540 upgrade from 2 TB to 4 TB. Agree that just adding an external 4 TB would be better. Either get a more affordable drive for the overpriced Thunderbolt 5 one.
Start explaining the concepts you are testing with the benchmarks or it is hard to decide if the numbers bear any relevance to the user. As a lot of viewers in YT have no clue about the differences betwen bits and bytes, I would convert everything to one performance unit like MegaBytes per second (MB/s). (My nicknames in 1 & 2.) (1) Front-end speed is the speed of the controller in the storage component, the interface to the controller, the firmware in it and its cache memory. (2) Media speed is the speed of the long term storage memory behind the controller and cache. In and old S-ATA III magnetic platter HDD, the front end speed is 750 MB/s (megabytes per second - 6 Gigabit per second). The media speed - how fast a drive can write the data to the magnetic platter - may be only 50 MB/s. With a RAID 0 array you can mask this problem by having files striped across different physical drives, but you would need 15 (16 in practice) to get 750 MB/s sustained write speed with 50 MB/s media speed. (3) IOPS - Input/output Operations Per Second - references the controller and firmware's ability at another level. Controller/firmware has to do overhead for each new file and new block in there and the firmware has no unlimited capacity in time, either sequentially or in concurrently. (4) File-system efficiency. The file-system references how files are stored, catalogued, documented in the storage device. FAT is very simple while NTFS is more complex. In the case of NTFS, being a "journaling" system, your disk already changes when you switch the device on and no files got written - but the Journal (log) got opened and hence changed. While this is the operating system side of your I/Os and has its overhead time cost that translates into latency, it also relates to the IOPS. So when we test a device's performance, we try to get an insight into these different aspects, except we may not figure out where the bottlenecks are. In the case of Solid-State storage, as in flash memory cards or SSD, we have the major cost component of "memory cell quality". That quality defines top media speed as well as lifetime expectancy. Last time I compared, converting the TBW of a Samsung EVO SSD into P/E-cycles (program/erase) gave me the number 300. That means the cells in there would brown out until black out in 300 re-write times. A Samsung PRO at that same time gave me 3,000 P/E cycles. Well, the CFexpress Type B cards I use give me 30,000 and are a lot faster. These cards have faster front-end speed than my camera's internal I/O controller but at some point fall down to media speed, when I keep shooting top speed stills at maximum resolution. Where the bottlenecking happens? Don;t know. The processor and motherboard with sensor have incredible bandwidth and it's not there. So it can be between the camera's frame buffer and the I/O controller where the I/O controller is the limiting factor. While you measure speed, you use P/E cycles. We can use software to investigate a storage device's health - basically informing us where the memory cells are in the brown-our process slowly applied to them by each re-write. For the 300 P/E-cycle 25% health means you have 75 out of 300 left, for the 3,000 you have 750 left and for the 30,000 you have 7,500 left. Even better memory cells exist, by the way. An alternative way to raise the P/E cycles in a product is called "overprovisioning". In that case, your 1TB storage device may have 1.5TB media capacity. How does that work? If each of your writes is a file of 0.1 TB, and you erase the file after having copied it to some backup, then a 1TB drive with 1TB media has used 1 P/E cycle after 10 times, because the first time the file is written, erasing it only removes it from the user-end O(OS) of the file system, but it is not erased in the media. So the 1TB drive with 1,5 TB media gives you 15 times the 0.1TB file rewrites for 1 P/E cycle. That process is called "health management". As SSD (memory card) vendors have learnt that they get hardly any complaints about the longevity of their storage products, we see storage products with "strange" capacities. This follows from a vendor moving media capacity to the OS/user level. The 1TB card with 1.5TB media becomes 1.25 TB visible (on the label) with 1.5 media, and the price is increased for a product with less quality. The problem with benchmarking is that small files stay in the front-end performance ballpark. Copying a large directory of small files behaves differently than copying a single huge video file. Each has different (4) loads that need different (3) performance. We see something else now in cameras where formatting a card in the past was done by just erasing the operating system side of the card's internal tables - fast format. Now we can alternatively deep format a card where the media cells are all set back to 0, if not 0 already. The question is if that gives faster writes. It adds a P/E cycle to the used part of the card, though. All this applies to SSD of any type. "Wear levelling" that is done by "health management" was never present in HDD. There's no point in defragmenting an SSD as you cannot influence the wear levelling that does its own thing in the media.
You need to test properly. Just having one device plugged in does not reveal the issues. Load all the ports at once and see what speeds you get then! 'Cos for me, they fall back to USB3.1 speeds when I have three monitors plugged in... Also those speeds don't sustain. The utility you are using is not a speed test utility, it's a capability utility that does not stress the busses.
It's much much faster on a TB5 system (but costs more). No wholly smokes here for once! But it's big and bulky (no, no it's not). Not your best video. That drive you keep plugging looks like a toy, I expect better from you.
I never, never purchase any external SSD that does not have a detachable cable. A detachable cable is easily replaced if damaged. If a permanently installed cable is damaged you are screwed.
I would personally like if it had something like a captive lock for an included cable, so you can detach it but really lock it down, so it’s really solid when connected.
You realize the main reason for the attached cable was to ensure the quality of the cable to get consistent good performance.
I have worked on computers since the late 80's. Unless you leave the ssd side always attached, I find the replaceable cable ones to be less reliable. The tiny ssd side connector will either lose the solder connection, or will rip out of the board. If the drive with a cable with proper strain relief is installed, that will last a good bit longer. At least in my experience it will, and I have serviced 1000's of professional clients who use external drives for various reasons. I have to replace or resolder my drive enclosures that I use for booting to Macs, on a pretty constant basis. Generally every year or 3. I have 2 WD usb drives with the built in cable and both are fine after 5 maybe 6 years.
@@otwer22or … they just sell branded separate cables that they advertise as the only way to get max speed? But you may use your own cables at “guaranteed reduced performance?” Not hard.
Don’t defend planned obsolescence and lack of end user repair.
I think the cable on that OWC drive is detachable but you need to open the enclosure to do that.
The Sabrent TB5 throttles down to 1.5 GB/s in extended use in their own YT vid !
The OWC Envoy TB5 throttles down to 1.8 GB/s
You can actually get better sustained performance out of TB4 / USB4 drives with active cooling
TB5 without active cooling is essentially a waste of money
Which one do you recommend then
Seagate Firecuda 530 or 540 is my choice. Only slows to 3,800@@siddhipriyamishra7210
@@siddhipriyamishra7210 For sustained performance (transferring 100s of GB frequently) a USB4 enclosure with a fan from the likes of Anyoyo, Yottamaster, Orico, Acasis (not their dual drive enclosure, that's only fast if you RAID the drives) coupled with a suitably fast PCIe 4 SSD (like above 5000MB/s) that you can re-use in a cooled TB5 enclosure later.
Current TB5 drives and enclosures throttle down so hard when pushed that they're simply not worth the money IMO.
They'll be very fast for lighter workloads though, but then there's the question wether you really need them or want to spend the extra $$$ on them.
Bit disappointed with Max Tech for not illustrating the throttled TB5 performance, while other reviewers have shown it - even Sabrent themselves !
Wait for ACASIS TB501 with active fan
@@siddhipriyamishra7210 Well, if you don't use these drives for extended highspeed read/writes, max 100 GB, you can use these and they'll be very fast
The OWC Envoy TB3 (yes THREE) is about 200 MB/s faster than their TB5 drive on sustained writes !
A USB4 enclosure with active cooling will be faster than either TB5 (or the OWC TB3) drive during sustained heavy use
You can find these from Anyoyo / Acasis / Orico / and others
Or like YenYun states, wait for actively cooled TB5 enclosures - they will be coming
Please do larger transfers (600-1000GB) and measure WHEN the drive slows down. Please also use a software where you can monitor the R/W-speed and get the average speed of the transfer. Always gets frustrated when newer drives that is supposed to be “faster ” actually performs worse over time than older “slower” models. Thanks!
Not enough space on their test machines. They don't usually buy 8tb models.
All external ssds will eventually slow down during long writes due to thermals or when cache fills up. Only difference is that some handle it well and other become almost unusable. If a drive drops from 900MB/s to 500MB/s it’s fine! But from 900MB/s to 80MB/s or worse isn’t acceptable in time sensitive situations.
Sustained performance is mostly a function of the underlying drive (QLC vs TLC, DRAM cache, etc). The enclosure's controller mostly affects the peak speeds, though the cooling can have a big impact as well. Personally, I think the best play right now is an OWC 1m2 empty enclosure, with a 990 Pro or SN850x drive inside.
@joelv4495 personally I use the Seagate Firecuda 530, jerryrigged into an open USB 4.0 enclosure, and tied down with a Velcro strap. Its the fastest sustained speed I can find. 2.9GB all the way through, past SLC, unless overheating, but I squished a copper heat sink on top of a silicone one, with also some tape meant for even more heat dissipation. It's the proudest and fastest thing I've made, and I don't think about upgrading to TB5 enclosure yet, unless it comes down to a price of less than $2000
ALSO several gigs of extremely small data like LUTs or Lightroom presets. Even fast SSDs slow down when doing thousands of tiny tiny files.
My massive sound libraries for music production are now reading at 5100++ MB/s versus my old Samsung T7's 900 MB/s. This results in less latency, etc. I love my Tb5 OWC Ultra.
Nice! I was thinking about buying one for this exact purpose! Really glad that it makes a difference!
I am planning on getting a 2TB version of the Mac mini m4 pro then buying another 2TB external solely for virtual instruments/sample storage (I have about 1TB of that right now and my internal 2TB is nearly full).
Thanks for sharing! Glad to know that there is a good use case that actually helps compared to most other things! And even though its expensive, Apple also charges a ton and you don't get to use internal storage with multiple devices.
keep in mind that an average SATA SSD in a desktop is around 400MB/s tops read. The little mini drive out performs that.
People still use SATA SSDs?
But, that average is no longer average. It’s getting harder to find new computers with SATA internal drives. Pro models have all gone NVMe because of the faster speeds. It has been years since Apple sold any internal SSDs that were SATA. I was shopping for internal drives and NVMe is essentially as cheap as SATA now, but is many times faster and smaller. 400MB/sec now seems quite slow to me, given all our current choices. The internal SSDs on the Macs in this test go 5000-7500MB/sec…15 times faster than a SATA SSD, and that kind of speed is standard on new PCs now too. There’s almost no justification to pay for slow SATA today.
when having SATA HD or SSD it's better to boot from an external drive: even over 10Gb USB... when not stuck with 5Gbps USB ports. Oh... desktops have expansion slots! :) And for Apple users: booting from an external drive over TB2/TB3 adapter makes a real difference.
I shopped w/in Thxgiving week, Gen4 SSD 4 TB inside Thunderbolt 4 aluminum enclosure now can reach 3400 read / 3400 write transferred to M4 MB Pro for total near $370. SSD being WD Black SN850X 4TB inside Qwiizlab Fanless 40Gbps USB4 Thunderbolt Aluminum SSD Enclosure. I consider this TB 4 combo valuable for years.
SN850X is way overkill you can use that for a Thunderbolt 5 enclosure too
@@Ryzen-4090-i3n Too bad TB5 is still new, no independent enclosure at reasonable price yet.
Yes, I'm waiting on some TB5 enclosures to make my own.
@@briang6652 Yeah too high price but when they get cheaper you could upgrade the enclosure without changing the SSD
Enclosures., especially with cooling fans is overkill but the way to go!
I get 3.2 GB/s read and write on my base M4 mini with the OWC 1M2 external USB 4 enclosure and Samsung 990 Pro 4TB.
That Sabrent TB3 enclosure only supports 20Gb/s connections which makes the TB5 drive look disproportionally fast in comparison to what most TB3/USB4 enclosures are capable of. The value proposition of the TB5 drive makes absolutely zero sense once you factor that in, especially when you consider the thermal throttling that fast modern NVMe suffer without active cooling.
What if you’re a video editor or a large volume photographer and you mainly care about read speeds as that is what speeds up your workflow? Copying is only needed once, but reading is what actually matters for most people
Guarantee you that if you open up the OWC TB5, 95% of it is just cooling. It's similar to external 10Gb/s nic cards. 95% of the enclosure is just cooling cause you typically transfer a lot. It works, but TB4 SSDs is great.
I guarantee you you're wrong.
Yeah I don't think I'll ever need to transfer 121GB of data (except time machine) in a day
Most of our youtube video projects are 150-300GB, and some can be 600GB lol.
@@MaxTechOfficial Do you keep all your video projects? If so you must have petabytes of local storage
@@MaxTechOfficial How many times per day are you wiping your project folder and re-copying it? the folder could be 4tb but if you only do it once at the start of a 200+h edit what does it even matter?
i use the ACASIS 40Gbps NVMe Enclosure w/ fan and I get roughly 3000MB/s Read & 2950+ MB/s Write on my Mini M2 Pro. This is on TB4 cable. I have a WD black 4TB drive installed as a boot drive.
FYI TB5 supports 120gb for display only, 80gb for drives and other devices
Great test.
I was just missing a comparison with one of the fastest TB4 enclosures from Orico or equivalent, with a high quality PCIe 4.0 NVMe SSD.
I actually opted for OWC's 1M4 enclosure, as it's beefy like the OWC drive here. The other nvme tb4 enclosures usually had fans that reviewers said were too loud, and I didn't get the warm fuzzy that they were reliable either. I opted to pay for OWC's enclosure (only on sale haha) and it easily maximizes the 970 Evo I put in there to run as a boot drive for an old iMac
Wow! That's faster *externally* than my M2 Pro MacBook Pro is *internally!* Whoa!
This is a great test, was waiting for this.
CES is around the corner. More TB5 products will probably be showcased. OWC is solid though.
I honestly don’t see a huge use case for one to transfer a ton a data from internal to external at those speeds.
Typically if one does a transfer it’s to archive and the speed is less critical in that situation. We mostly have limited space on Macs due to cost. That means there likely isn’t a huge amount of files to transfer anyway.
If we do happen to have 4 TB internal there is then much less need to use external unless it’s again to archive or backup.
Buying a TB5 drive seems like a very expensive option for archiving.
More likely we would buy a TB5 drive to actually work off of for the added speed which means a small chance we will transfer a large amount of data.
From a direct camera recording drive it’s also more cost than necessary in my opinion. A USB4 drive is more than fast enough for external video recording and as a source drive to edit from.
There is the transfer time to move that data to an edit drive if the camera drive needs to be used for another shoot. That might take a bit longer but again I don’t see that as a deal breaker at all. I have a hard time even justifying faster than USB4 from anything that would transfer data around.
It has been said that Thunderbolt 5 will be welcomed by every digital imaging technician (DIT) working TV and movie sets, because they do in fact need to transfer very large amounts of data daily under a deadline. And the cost of 10 Thunderbolt 5 SSDs will be one of the smallest line items on any of those productions.
You make some great points about the practical use cases for faster drives like TB5. For most workflows, speed is often limited by other bottlenecks, like the SD card itself. For example, even with my Sony CFexpress cards, the transfer speeds I get are just slightly above SATA SSD levels in real-world conditions.
When it comes to editing, I’ve found that the difference between SATA and NVMe drives is noticeable only during specific tasks, like opening large projects or caching previews. Once you’re in the middle of a session-editing timelines, scrubbing footage, or color grading-the performance gap becomes negligible. In fact, editing off a NAS, which is fairly standard in many setups, feels almost identical for most tasks after the project loads.
I agree that TB5 would be a game-changer primarily for those working directly off the drive where the speed really matters, or for scenarios with huge datasets requiring rapid transfers. For archival or standard external workflows, USB4 is more than capable and offers a much more cost-effective solution. It's all about matching the gear to the workflow, and in most cases, TB5 is likely overkill unless you’re dealing with extremely high-end, speed-sensitive tasks.
Would love to see some M.2 cards in thunderbolt enclosures/hubs.
That's basically how most external SSDs are made.
@@Tappits84 I’m talking about different brands of M.2 and enclosures… not a complete SSD that is all one piece.
This would be perfect for a mac mini in standard configuration!
It's okay with the M4 Mini, but the M4 Pro Mini is quite a bit faster. Microcenter sells the M4 Pro entry level Mini for $1,200.
can't say thanks enough for not only this review, but alllll your content the last decade. you've single-handedly help me make the most informed decisions for my media company, helping underline the most important performance criteria that we prioritize. appreciate it immensely!
question for you on this TB5 drive: would love your insights on specifically lightroom speeds. we all struggle with LR performance, even with highly optimized systems. curious how this performs hour+ into a LR editing session? I still struggle with Sandisk G40 (which has great r/w speeds) and it's just a matter of time before I need to optimize catalog/restart app/reboot computer etc to 'wake up' lightroom. when we get back from big shoots and have 1000s of images to edit, these speedbumps with LR can be not only frustrating but time-consuming. Would love any input you may have on how TB5 Envoy here could be helpful for us hybrid shooters, who do a lot of LR room for several hours at a time.
thanks m8!
A lot of people missing out on the point. The TB5 drive (while it is indeed much more expensive now, which mostly all other new tech are), does not only shine on file transfers. With the price of upgrading your mac's internal storage, this is a great alternative. Process your files from this drive directly without having to worry about latency. Again, arguing that this is not worth it because you're not doing continuous file transfers can just use old, cheap, and reliable drives. Those who want to extend useable storage space for actual direct file consumption (photo, video editing, etc) would benefit using this.
"Don't buy this" my ass.
ACASIS for the win, OWC is funny for their prices.
Your testing is for a limited use case though? If you're using the SSD as a working drive for a heavy 4K (or 5,6,8k) timeline that speed advantage is definitely going to help. Sure, for transferring files, they're all reasonably fast...
I use it with the new Mac Mini Pro and for music production (All my software are stored on the OWC) and its just so much faster than my Thunderbolt 4 drive. A no brainer.
OWC is a great company. I bought an internal SSD from them for my 2012 16" MBPro a few yrs back, and it was night & day difference in speeds when booting up etc. after I installed it.
You pay a premium for this first TB 5 box with an SSD drive. Just wait until there is more competition and until then TB 4 with speeds around 3500 MB/s is not such a bad alternative.
What an amazing effort to cover all the updated tech in the tech world, Thankyou once again covering these topics. 🥰🥰😃😃😄😄
Nope for the T5 right now. I've been a happy user of a Sabrent 1TB Rocket XTRM which is a T3 but so reliable, fast enough for video editing. I don't use the silicon case and I elevate it a little from the desk so even if it gets warm, it never crashed or caused any problems. It's connected all the time to my M1 Mac Mini.
I believe I'll get another one that's a Sabrent Rocket XTRM Plus 2TB for the sake of space on it, it'll only be for editing actual video projects. You US peeps are happy for the size of your market, this SSD is $274 over there, in Sweden where I live it's bloody $354, "thanks" to Swedish import taxes ..
2:33 durability*
TB5 smokes but should be on iPad Pro and iPhone Pro too going forward
For expanding Macbook and Mac Mini storage without paying apple ext*rtion price, 400-600 USD for 2 -4TB is a bargain.
ya.. just bought OWC 1M.2 & 4TB 990 pro for $405 including taxes delivered.
I just picked up the Sandisk Pro G40 and getting 2850/2900 on my Mac mini m4 pro.
Great drive, it’s practically that fast on my M1 MB air too
Great review.
For me, I only rely on OWC external drives. I thought about getting a TB5 OWC Envoy Ultra with the intent of putting core content files on the external drive (to switch between a Mac Studio and the new MacBook Pro M4 Max.
Two problems:
1. Real-world data transfer is incredible, but I think 3000+ MBs using the TB4 OWC Extreme may be good enough.
2. The short integrated cable doesn't work as I switch between systems (based on my cabling set up). Also, I prefer separate cables that I can easily replace (longer is better).
And a third problem I can't explain: I used an external OWC 1M2 Express for core data (instead of the internal SSD). I set up Time Machine to do backups when plugged into Mac Studio desktop. After a couple of hourly backups to another backup external disk (also OWC), Time Machine failed. I rebooted, reformatted the Time Machine backup disk, and tried automated, background backups again. Failed!!! if I remove the @OWC Express from being part of the TM backup, it never fails. Any ideas???
Maybe a corrupted file or directory on the OWC Express. Have you tried to repair it with Disk Utility "First Aid"?
@ yes
@@SoftwareManiacLSM Weird. What if you back up the OWC express separately, to a TM volume of its own?
@@jpdemer5 I DID have an apfs corruption!!! Not sure how. Now everything works!
*_Спасибо, чувак_*
_Spasibo, chuvak_
*ThanX dude!*
thunderbolt 5 is overkill it's good that we have in macs for screens, docks and it's much more futureproof when TB5 becomes a standard
I don't really need the super speed so I've been having good luck with the Samsung T7 drives which would compare with the Nano. Only crazy part is the price which changes all the time on Amazon in Canada. Can change up to $75 from week to week and even just between the 3 colors of the same size drive!!! Last week, the gray 1TB was $124 and the blue 1TB was $199.
If you want small andcheap, look at the Samsung T5 USB drive. Right know you can get 2TB for as little as 60$ on sale and that drive still packs 400MB/sec.
They will do 500-550MB/s if your cables are ok.
Sabrent has a Thunderbolt 5 Drive also, will you test it?
If the Sabrent Rocket Nano is a USB 3.2 Gen 2x2 drive, more than likely the speeds will still be capped to lower speeds, isn’t it? At least until the M3 generation Apple didn’t support 3.2 Gen 2x2. I’m unsure about the M4 generation.
If that’s the case, buying a Samsung T7 that will have the same speeds and same storage for a cheaper price seems like a more reasonable approach.
The confusing element here is not knowing what Gen the SSD sticks are in each of the enclosures. I'm not sure how the TB converters in the enclosures interface with the PCI lanes on the SSDs. The number of lanes varies between the Gens (3,4 & 5).
compare it to a zyke drive with a fast Gen4 NVME drive or an TB4 casing with active cooling. Yes thermal throtteling is an issue. Also an issue is power consumption in sleep mode (closed lid e.g. over night)
While it's not there yet, value-wise, it does suggest you'll want Thunderbolt 5 in the next couple years. One more reason to choose the macbook m4pro over the m4.
Absolutely. I’m about to spring for a Mac Mini Pro for the same reason. Today’s latest and greatest will be tomorrow’s normal.
IRW you should test the LR pano and HDR stitch as well as preview build and exports for these SSD.
Error- the max bandwidth of TB 5 is only 80 GBs when transferring data. The 120 GBs bandwidth only works with high res, high refresh displays.
You can't do drive tests the same way you do cpu benchmarks. Use huge files, loads of small files, simultaneous read and write etc
I'd rather have larger and heavier if it means a cooler SSD. It's only going to sit on my desk, so I'd be fine with one the same size as the Mini.
The price will drop as other vendors start making TB 5 drives
I'd like to know how much the external SSD can replace internal SSD. For a laptop, it is inconvenient to carry an extra thing dongled to the machine, but the price of internal SSD on a macbook is high. More to the point though, let's talk about Final Cut Pro, or Logic Pro. Not small projects. It has been said in the past that having the project data on a separate SSD to the application is helpful for performance and I am wondering if that still holds true for internal speeds these days? Also noting that the internal speeds of SSD's vary a lot between machines. Anyway, can you do performance tests of these fast thunderbolt 5 SSD's in a real world scenario. In particular, with Final Cut Pro? See how much difference it makes compared to an internal-SSD-only project, or just to see how much SSD I really need internally if external SSD is fast enough for.... say .... a project in 4K60 with multiple layers and noise reduction, optical flow, whatever. The real world performance difference of SSD's with real world projects is very interesting, as compared to raw MBs specs.
Cool video. How hot does it get?
Did you try formatting the SSD into APFS? I’ve noticed a big increase in speed on a regular usb c SSD
I suppose M.2 RAID drive with TB5 will be available in the middle of next year.
I’d love to just get the nano but the M4 mini doesn’t have USB . Will an adapter kill the speed ?
TB5 is new but it to expensive + you'd need a TB5 Mac which the masses won't have.
TB4 is good enough for a long time.
just buy a 8TB 7200 rpm internal sata drive and put it in a usb 3 dock .. is 2 or 3 min that long to wait?
8tb is under $200
Having purchased a base Mac Mini M4 Pro and wanting to put a 1TB external SSD for apps and games, what is better to get? An SSD with USB 3.2 or a USB with Thunderbolt?
Don’t get usb 3.2 Mac’s don’t support it.
What was the r/w speeds on the M4 Macbook to start with? if its a lower capacity SSD you would be throttled on how much data can be read from or written too
Cloud storage?
Very interesting test but I disagree with the conclusion. In the real world people don't just use these drives to transfer large files once in a while. People work off these drives and I can see how a slower one can be a real bottleneck. It would be interesting to see how these drives perform in a real world situation, like having a photo library on them, importing a bunch of RAW photos and do some editing and exporting.
And in the real world, say working on Photoshop files, is there any real difference in speed / performance?
How direct video editing from SSD drive effects drive? What are good speeds for drive for direct video editing from it? You mention that SABRENT NANO is enough for video editing, but how we can know that because footage type can be different ✌🏼
Hi there! possible to do a test with 500GB? I feel most video project files are not gonna be 128GB. And so the throttle will make a bigger difference.
Hi. A bit of an unrelated question but im hoping you could help.
I have an old 1tb external hard drive that i believe has a bad port. I have tried 2 different cables and sometimes works but keeps connecting and disconnecting. The cable has a bit of play in the hard drive port which i think is causing this.
I have ordered new 4tb hard drive but i am wondering is there anything i can do to stabilise the connection so i can transfer the files? Sketchy or not. I just want to be able to get the files off it.
Thanks.
Got to have at least 4tb for any NVME and it be over 3000mb/s, so needs a thunderbolt controller.
Great video Max, BUT!
Are there any differences in editing videos off of the new TB5 speeds??
I'm currently working from a Samsung T9 (2GB W/R p/Sec) - considering upgrading to the OWC Envoy Ultra TB5 SSD - putting the money on the side, how much of a difference would I get by editing off of it?
Both TB5 drives throttle down to below that speed in sustained operation
Sabrent after about 100GB, they attribute it to drive cache in their own YT vid
Look up YT reviews which test them over longer times, instead of these advertorials
There should be no difference, even with RAW footage between the two drives as they are more than fast enough.
Nonsense, the OWC is amazing and if you are editing High-end 8k RAW 16bit video there can be nothing else. Having said that of you never going to be doing that then I guess it's good advice.
That’s stupid fast! 3:29
Not enough. My Satechi USB4 Enclosure + SN850X manages 3800MB/s.
What about iCloud Drive?
Is the data from an external SSD being mirrored on iCloud Drive as well, the way it happens with the data on your internal SSD?
How is the speed by 5000 jpg files to backup the photo library?
wait is m4 air wont have tb5? should I get m3 deal in black friday?
Everything is expensive when first came out. Give it time and price will drop.
it’s look like you are doing promotional showcase …
you should put a disclaimer
Nope bought myself.
looks like there may be some disinformation here. Looks like the cable is not built in. Looking at the enclosure, it looks very similar to Angel Bird card readers. Meaning that the cable is meant to fit tight, so if you give it a firm pull, the cable should come out.
Please report back to us V!
OWC themselves call it a "built-in" Thunderbolt cable. That doesn't sound removable to me.
Anyone know what apps they use to monitor performance on Mac?
Nvme ssd in enclosure is better in any way
Pls test the speed for a certain duration say like 10 minutes. For example transfer gigs of data with multiple files and see the average speed.
Why? That's longer than completely filling an empty drive to 100% capacity; what use case needs an external SSD to go from empty to full, then empty and back to full again in a 10-minute time period?
Feels like a flawed test, as each drive has different flash chips inside them.
What is flawed about the test? If you test the devices, and got the results….it is what it is…..
@ you would test the same flash chip in each interface on the same laptop.
@@dontbekurt that really isn’t that important to me. I see the results. It gives me a good idea of how things will work for me. if he tested like for like. Then cool. But if not. Oh well.
@@redesignedlife777 That's just it.. I don't know if these tests are indicative of how it would really work because of the fact that all of those things are different.
Try the tropic colour space shuttle please
What about USB4??
USB 4 and Thunderbolt 4 are essentially the same spec. If they tested Thunderbolt 4, look at those numbers for your USB 4 result.
idk why you kept on plugging and unplugging. You do know there are three ports.
AI translated audio is such a pain to hear. Please don't do that.
Not sure this is that much better than FW400.
this is not correct. You must tested ssd thunderbolt 5 VS ssd nvme gen 4.0 or 5.0 by plug thunderbolt 5
My god that's a lot of porn
compare please m3 pro and m4
apple says 120gb/s and in reality it took 120gb/minute. Who's fault is that, Apple or the SDD or the cables?
That is spec for the full TB5, not what is able to be used for data.
In fact MaxTech and all reviewers need to stop saying 120Gb/sec when talking about data, because no one will ever sustain that for data. 120Gb/sec is an asynchronous burst mode.
hi
KKKKKKKKKKKKK Your brasilian portuguese Ai dubled voice sounds like a little girl
Do never forever ever use translation again... It's AWFUL
U can choose to disable it
Please disable this horrible RUclips auto-translation. They are painful!
Just tap the “CC” symbol on the upper right portion of the video.
@ it’s still annoying to do this every time, especially if you watch on auto play/playlist, on a device without capability to touch anything or if you just can’t touch it.
And also the translation is so bad from the voice and the translation itself, so it’s not even a benefit for people which need this.
I’m so sick of this channel being all about what they can sell. It’s sad how desperate they are to create click bait.
Then go away, it’s that simple. No one asked your opinion on the channel, please contribute in a positive way or not at all.
I wish I would have purchased my MacBook Pro with 4 TB instead of only 2 TB. The extra cost is no different than getting this external drive
The Apple cost from 2 TB to 4 TB is more than the cost of one of these 4 TB drives. so having 2tb internal and 4tb external gives a larger storage pool for less money.
@@Tappits84 I see it is a $540 upgrade from 2 TB to 4 TB. Agree that just adding an external 4 TB would be better. Either get a more affordable drive for the overpriced Thunderbolt 5 one.
Now in Portuguese! 👏🏼👏🏼👏🏼
com voz feminina rs
Start explaining the concepts you are testing with the benchmarks or it is hard to decide if the numbers bear any relevance to the user. As a lot of viewers in YT have no clue about the differences betwen bits and bytes, I would convert everything to one performance unit like MegaBytes per second (MB/s).
(My nicknames in 1 & 2.)
(1) Front-end speed is the speed of the controller in the storage component, the interface to the controller, the firmware in it and its cache memory.
(2) Media speed is the speed of the long term storage memory behind the controller and cache.
In and old S-ATA III magnetic platter HDD, the front end speed is 750 MB/s (megabytes per second - 6 Gigabit per second). The media speed - how fast a drive can write the data to the magnetic platter - may be only 50 MB/s. With a RAID 0 array you can mask this problem by having files striped across different physical drives, but you would need 15 (16 in practice) to get 750 MB/s sustained write speed with 50 MB/s media speed.
(3) IOPS - Input/output Operations Per Second - references the controller and firmware's ability at another level. Controller/firmware has to do overhead for each new file and new block in there and the firmware has no unlimited capacity in time, either sequentially or in concurrently.
(4) File-system efficiency. The file-system references how files are stored, catalogued, documented in the storage device. FAT is very simple while NTFS is more complex. In the case of NTFS, being a "journaling" system, your disk already changes when you switch the device on and no files got written - but the Journal (log) got opened and hence changed.
While this is the operating system side of your I/Os and has its overhead time cost that translates into latency, it also relates to the IOPS.
So when we test a device's performance, we try to get an insight into these different aspects, except we may not figure out where the bottlenecks are.
In the case of Solid-State storage, as in flash memory cards or SSD, we have the major cost component of "memory cell quality". That quality defines top media speed as well as lifetime expectancy. Last time I compared, converting the TBW of a Samsung EVO SSD into P/E-cycles (program/erase) gave me the number 300. That means the cells in there would brown out until black out in 300 re-write times. A Samsung PRO at that same time gave me 3,000 P/E cycles. Well, the CFexpress Type B cards I use give me 30,000 and are a lot faster. These cards have faster front-end speed than my camera's internal I/O controller but at some point fall down to media speed, when I keep shooting top speed stills at maximum resolution. Where the bottlenecking happens? Don;t know. The processor and motherboard with sensor have incredible bandwidth and it's not there. So it can be between the camera's frame buffer and the I/O controller where the I/O controller is the limiting factor.
While you measure speed, you use P/E cycles. We can use software to investigate a storage device's health - basically informing us where the memory cells are in the brown-our process slowly applied to them by each re-write. For the 300 P/E-cycle 25% health means you have 75 out of 300 left, for the 3,000 you have 750 left and for the 30,000 you have 7,500 left. Even better memory cells exist, by the way.
An alternative way to raise the P/E cycles in a product is called "overprovisioning".
In that case, your 1TB storage device may have 1.5TB media capacity. How does that work? If each of your writes is a file of 0.1 TB, and you erase the file after having copied it to some backup, then a 1TB drive with 1TB media has used 1 P/E cycle after 10 times, because the first time the file is written, erasing it only removes it from the user-end O(OS) of the file system, but it is not erased in the media. So the 1TB drive with 1,5 TB media gives you 15 times the 0.1TB file rewrites for 1 P/E cycle. That process is called "health management".
As SSD (memory card) vendors have learnt that they get hardly any complaints about the longevity of their storage products, we see storage products with "strange" capacities. This follows from a vendor moving media capacity to the OS/user level. The 1TB card with 1.5TB media becomes 1.25 TB visible (on the label) with 1.5 media, and the price is increased for a product with less quality.
The problem with benchmarking is that small files stay in the front-end performance ballpark. Copying a large directory of small files behaves differently than copying a single huge video file. Each has different (4) loads that need different (3) performance.
We see something else now in cameras where formatting a card in the past was done by just erasing the operating system side of the card's internal tables - fast format. Now we can alternatively deep format a card where the media cells are all set back to 0, if not 0 already.
The question is if that gives faster writes. It adds a P/E cycle to the used part of the card, though.
All this applies to SSD of any type. "Wear levelling" that is done by "health management" was never present in HDD. There's no point in defragmenting an SSD as you cannot influence the wear levelling that does its own thing in the media.
You need to test properly. Just having one device plugged in does not reveal the issues. Load all the ports at once and see what speeds you get then! 'Cos for me, they fall back to USB3.1 speeds when I have three monitors plugged in... Also those speeds don't sustain. The utility you are using is not a speed test utility, it's a capability utility that does not stress the busses.
Thank for the translation I did'nt know you was a woman :D
Pertama woe
It's much much faster on a TB5 system (but costs more). No wholly smokes here for once! But it's big and bulky (no, no it's not). Not your best video. That drive you keep plugging looks like a toy, I expect better from you.