Recorded on the NAB 2026 show floor — The LTO Show’s recap conversations with Spectra Logic, YoYotta, and Archiware about what’s new in tape storage, object-on-tape, and media archive workflows.

Spectra Logic — Mike Waters

I’m here with Mike Waters from Spectralogic Enterprise Sales, We’re here to talk a little bit about what’s new with Spectra at NAB, and we’re going to start with the Spectra Cube. Thanks Pete. Well, what we’re looking at is the actually the side of the Spectra Cube. This is our latest tape storage device. It holds over 16,000 cartridges and what that creates today is a density with the latest drives of over 50 petabytes and a little over the size of a 19-inch rack that sits in a data center. Yeah and I had to ask I said 1500 some slots in this chassis right here and that is a fact. Right. This holds up to 1500. 1670. That’s pretty good slot density.

Right, slot density. So the way it’s been engineered is to still use either the latest LTO10 tapes or the latest LTO10 tapes and drives or were very backwards compatible all the way to LTO6. So our customers can leverage the use of legacy media, legacy tape drives.

Mike, tell me a little bit about how Spectralogic is supporting object storage on tape. that’s a very big part of what we’re doing into tape devices right now. So we’ve created an abstraction layer so that the function that’s going on inside the tape library is not part of your day-to-day handling of the tapes. Gone are the days so…

I have to reach in, I have to find these things, I have to move them place to place. What we’re doing here is creating an S3 layer or an Amazon compatible S3 using the same API calls and everything That device also creates the ability for our customers be able to send data to the cloud, have them send data to other storage devices that are out here, and have that go to an AWS S3 compliant tape device that might be out here.

It used to be called the Black Pearl and we’ve evolved the naming conventions now and we’re calling everything the Spectre Storage Server. okay. Very and inside of this there’s a cache layer and a disk layer. Right. For I.O. and for resident projects. If you need to get some things quickly.

It will be up here in the cache. That will determine how big that device actually needs to be.

YoYotta — Martin Greenwood

All right, we’re here with Martin Greenwood, CTO at Yojata and we’re going to talk about some updates to Yojata’s backup software. Martin, thanks for joining us on the LTO show today. Hello, Pete. Good to see you. Good to see you again. And it’s the last day of NAB. It’s gone well for us. Yes. And what we’ve been showing is our new version five, which is the latest version of our Yojata software. Yeah.

And this can be used in on-set, in production and post for managing all of your digital assets. From camera cards all the way through to data tapes and utilizing maybe cloud storage as well. An app running on the Mac can control standalone LTO drives.

index rates, sands, cloud buckets, and we could also control tape libraries like the Magstore 1000, 2000, 4000. And you can see the new interface gives you a very graphic look with all the tapes loaded in there. And this is very helpful for people loading up tapes, taking tapes out, they know exactly what to put in.

And then our job interface allows you to set up jobs for large, small archives, large archives, moving things from one place to another. And of course it’s all checks unverified. Excellent. So a few things I want to focus in on with you today that are kind of new features, I think, in this version. Let’s start with this staging drive concept that you’ve introduced to make tape backup more efficient.

from your disk or what have you. So tape works at a high speed and it likes to maintain that high speed. will slow down or stop if necessary. But obviously it’s ideal if it keeps being fed a good stream of data. Now, if you’ve got fast storage, that’s…

That’s great. But quite often people are using shared storage now. Or they want to download from the cloud and move it from there and archive it. So in the past we’ve had to, we’ve got several different speeds. So the downloads from the cloud might be fast, but then someone else uses the link. The NAS might be running quickly. Someone else hits play on their grading system.

you lose your bandwidth here. That’s not ideal. So in the past we would have set up two jobs, a job to set it up and to stage it on a RAID, and then transfer that RAID. It’s all verified, so there’s no harm in doing that, but it’s an additional process. So now in version 5, we’re able to automatically utilize an SSD drive as like a staging here inside the software. So it means we can fill that up hopefully as fast as possible So before we were always trying to get matched source and destination. And now it means we can truly go from anything to anything, making it much easier to move data around. Yeah, and typically speaking, and a lot of people don’t realize, a tape drive wants a constant stream. That’s really the only way to max out the published spec.

for bandwidth and it’s also better for the drive because some of this back hitching and shoe shining that occurs when a drive is waiting for data add headwear and so on. then it gives tape a bad name.

customers doing small archives, large archives, people routinely running petabyte jobs to take. And they can predict exactly how long it’s going to take. They’ll probably be using RAIDs though, dedicated to the process. Now we simplify that. You can take it from where it is rather than having to move it into the data lab.

So in your software this ability to stage these jobs in a more simple and automated way it can use any type of disc for that but even a simple single SSD Yes, you do want a semi professional one and generally it means that the larger they are the bigger the heatsink it’s probably going to be better because SSDs are designed you know to The cheaper ones are designed to work very fast, short periods of time. So you put a few photos on, you run it for maybe five minutes maximum. And what happens is if it gets hot, it throttles the speed back. And you’ll notice that the smaller ones may get literally burning hot in your hands. So the professional versions have a larger heat sink to dissipate that. And then it means you can be using it.

all day long. What’s built into version five, which is new, we’ve always monitored the health of tape drives and tapes, but now we’re doing that for SSDs as well, there’s a technology called Smart, which looks at the sectors, the blocks that you’ve used in the SSD. SSDs are designed to, they do wear out. have wear-leveling technology built into them.

we’re able to monitor that. That’s handy for things like camera cards as well, so you can predict that they might be wearing out and fail if they’ve been used for a year or two. You might want to cycle them out. And the same thing, obviously, we can monitor our staging cache. You wouldn’t want to, for example, you wouldn’t want to be using the SSD inside your MacMilli or your laptop. You add one into the system, an external.

unit on Thunderbolt. Our whole engine runs things like copying verification around about 20 gigabytes per second is our upper limit. So we’ve got plenty of bandwidth to be able to work with faster and faster devices. With Thunderbolt 5 now increasing the speed. And there’s Thunderbolt 5 versions of tape backup.

Hardware coming out any day now. Yeah. So one other thing I want to chat with you about today, Martin, is your you now included support for encryption. Yes. On backups to tape. That’s right. How have you implemented that and what are some of the OK. Well, a little bit of history inside tape drives. The controller is able to encrypt. Yeah. But it’s a bit of an older generation of encryption. And what it would do is encrypt the whole tape.

And what that would mean, if you wanted to restore data from it, either the old tape or a few files, you’d be telling the drive to decrypt the data. So it would be unencrypted in the data lab, which is not really what you want. The idea of encryption is as you pass through different processes, as you go from cloud down to a drive to a tape, You want it to move, if possible, for certain projects that they want to be encrypted the whole time. And then the keys are made available to the end user who’s authorized to look at the content, the colorist, the compositor. And the data archivist or a scorer doesn’t need or want to see the actual content. So what we’ve built into version 5 is that in the middle of our copy, engine, we can switch in an encryption module and that could actually work on a file by file basis. So some files you may not want to encrypt like reports or documentation and that can go on the tape and that means they can load the tape and read that freely. Thumbnails that we put on the tape we may have another key so that someone can decode just the thumbnails. And then the actual content would be protected with the latest generation of encryption. I see. Very interesting. So you’ve implemented encryption through your software. That’s right. it’s software encryption. Instead of using the drive-based encryption. Exactly. But it gives, sounds like a lot more flexibility in a variety of use cases. And I think you told me in our prep that you still can enjoy the benefits of, so the key that you use for the encryption, how do you get it to a third party if necessary? Okay, well when we’re working on things like the Mac, we can generate, we can get the secure enclaves in here to generate a key, which is not even exportable, the private key, so it’s ultra secure, and we can set it up.

to do that on a number of machines for the people who are authorized to get the key. So that’s one great way of doing key management. And because our app runs on the Mac, it’s a good combination. So as I say, just to repeat that, the fact that it generates a private key and there’s a thing called a keychain, which is like a software thing, which is exportable. And then there’s also like a hardware chip in there that you literally cannot get.

into at all. So it means the key is completely secure. But we also can encrypt the keys and distribute them for you to store and use. And it’s industry standard encryption, so you can decrypt with other software. So with your system, I don’t need to set up a I think they’re called KMIP servers, right? I don’t have to set up a dedicated key server on my network and manage all that like I would with the hardware-based encryption. Yeah, yeah. And we also have neat tricks like you can share the full keys, but I can give you a key just to work on the audio if you’re the sound guy. And they can take the tape, which has got, you know, confidential photo.

camera material on there, but they won’t be able to look at that. Or vice versa, the offline editor will just be able to look at trans codes and so on. We’re using separate keys and we’re managing them here, but then obviously we distribute them to you, encrypted, then you have responsibilities of putting them in your own hardware store and security.

Is there a performance trade-off though to these design decisions you made to support encryption this way versus… No slowdown at all. We can do that in parallel with our verification and so on. And we’re actually verifying the original content which is obviously wrapped inside the encryption and also the actual encrypted file. So again, you can verify material coming off the drive without decrypting it.

And then when it’s decrypted, it’s checked yet again that the content inside is valid. And I don’t need more RAM or CPU to… No, the great thing, I that’s one of the reasons why we focus on Mac is Apple Silicon has really transformed a lot of workflows. Archive systems tend to run for long periods of time connected to things like the Mac Store tape library over there. And Mac minis run very cool.

low power and high performance on all the models have got enough power. Some of the, you you may want a few more ports. That would be a reason to get a particular model to connect Thunderbolt ports to more devices. you literally we can use, I mean, we’re up to with version five, we control up to four tape libraries with eight tape drives. So that’s like enterprise software, which would be, you know, in a different ballpark price-wise. We’re far more advanced than that at a lower price model. So Martin, and we’ve kind of gone over on time, so we’re going to try to keep it a bit brief, but you’ve implemented some really exciting support for S3 storage. tell us about That’s right. So before we had some S3 in our version 4, some support, but it would…

upload and download as a separate process because of the reasons I said before about speed matching. We have clients who say, well, I’ve got a 10 gig link and they forget about the fact that obviously other departments are using that same. The great thing is it’s shareable. That’s a bad thing, it’s shareable. You really want your dedicated 10 or 25 gig link, but you’re not going to get that. But now, integrating it into the software so that you can pull your files from S3 and go direct to tape if required. And one last thing is we have the ability to conform options. So we can put in editorial timelines because obviously once you’ve done the offline edit, you don’t need most of the content. You just need selected files. So we can pull some of those files from a cloud bucket, some of them from a drive, a raid, and then pass them on. So this conforms, explain that a bit more because I think it’s really important because it does sound like it makes the whole workflow related to your storage use and efficiency a lot more efficient. if you think of it, yeah. it’s content aware based on what platform it was, what the file types are. Exactly, it has to look at file types because if you just looked at the clip name, you might be picking up the offline, so an HD version of the content. There might be times when you want that, to pull that back to give to an editor. But there will be other times when you want the original camera material or where you want the EXRs, have been made, image sequences that have been made from the, which you’ve had some other process. So, and the digital asset manager doesn’t have necessarily all that metadata.

Whenever we index a drive, a tape, or a bucket, we extract the metadata and build that up with optional thumbnails as well. So we know the codec, the resolution, all the time codes, and so on. And then we grab the editorial timeline, which could be EDLs, XMLs, et cetera. So we’re supporting Avid, Premiere, Final Cut, all the big workflows. So you grab all your timelines.

which is the recipe for cutting your 100 hours of material down to your 15 minute program. And that’s what’s in your conformed database that you’re maintaining. And it drops it in here and it’s like a very advanced shopping list and it goes through and works out. We need those red files, those Sony files, different stills or whatever. And it will find out where they are. Some of them might be on a raid locally or on shared storage, so it will retrieve those. Other ones will be on tape. Well, they’ll all be on tape, but for reasons of speed and parallel, you can restore. And we can then generate a restore list to go to your drive, your tape drive. And if you’ve got multiple tape drives, we can actually make separate restore lists and run in parallel.

So we can restore using multiple tape drives. That’s very interesting. but again, with this capability, you can be very selective about what you might restore from cloud. Yes. Right. Which is the egress fees or you know what kill you. that this is this is new to this implementation of your S3 support. Right. Yes. Yeah. So that’s very powerful. Yes.

Well, that’s a great overview, Martin. I really appreciate it. Very exciting. always fun to see you and your team here at NAB. Yeah. Thanks for coming by our booth. Yeah. Thanks for joining us on the LTO show.

Archiware — David Fox

We’re here with David Fox of ARCQR. Going to talk a little bit about the newest version of ARCQR P5 and what people can look forward to. David, nice to see you again. Hey Pete, nice to be here. Thanks for coming to chat with us. So, yeah, 2026, another NEB show. It’s version 8 that we have had out since earlier this year.

we have a replication tool called P5 Synchronize. That previously was able to synchronize between on-prem storage hosts, maybe across WANs, different sites, replicating data. So that can now also go to S3 storage buckets. So, know, Amazon, Google, Backblaze, Wasabi, et cetera. So you can now have like infrastructure where you’ve got like a Denver, Michigan office, you can send data between those two, and then one of them could send all of that up to the cloud. You can have lots of different nodes doing different things in different directions. You can orchestrate it all around schedules. Powerful. So that’s quite nice. And then of course, we’ve got our traditional P5 archive, people using that with LTFS to push stuff to tape, free up space on disk. We now have the ability to work with EDL files, so you can take…

export your timeline from your editor. P5 can use that to know what high res files to restore from LTO. So if you’ve got a workflow where you archive everything upfront, get stuff, the high res off the storage, do your edit and then at the end do your conform. We can now work in that kind of environment. We’ve got the ability now when customers are archiving.

they may be some kind of regulatory compliance over the archive, which means that they can store it for two years and then they’ve got to delete. So for example, maybe body cam, security footage, or enforcement, that kind of stuff. So what you can now do with our archive is you can archive stuff off to tape, you can set a two year retention, and then you can delete stuff from our index. it will indicate when a tape can be recycled.

I see. So now can that be set for any period of time, it could be ad hoc. Okay. So you can just go into the index, take a folder from 2022, delete it from the index, and then you can look at your tapes and we’ll show you which tapes are now completely empty because they were only storing 2023 data.

Yeah. So it means that your archive, whilst obviously it’s great to archive stuff and work on the basis that you’re keeping it forever. But sometimes you’re not. Yeah. need to it over. It’s a regulatory thing. Particularly in the video surveillance market, there’s a use case for that. So DLM now allows you to graph and analyze archive storage by client or any other metadata query. So that’s digital lifecycle management, presume. That’s it. It’s data lifecycle management. Data lifecycle Which we already had, which does verification and migration of data. Maybe to tape to cloud, the other way around, tape to tape, but changing the generation. So that now allows you to do some analytics based on your customers.

Maybe you could use it to do some chargebacks see who’s consuming the So you know what your cost is for the archive, but you’ve got 10 clients. Maybe one of them is the monster that’s using all of the archive storage. so the dashboard is where you can track that. And the final one is LTO hardware. And that’s a big one.

It is. It’s huge. So how do you guys implement that? The encryption, we basically have now a key management section in the product. So you can either bring in keys, which is our preference, because we’d sooner you already have a key that you have stored outside of P5. Right. can bring that in. So if I have a key management server, my own keys, it’ll work with your software. You can bring those keys in. Or if you don’t have one, you can generate keys inside of us, but then you need to export them because they obviously need be stored externally. And then it’s a pool based encryption. So we work on the basis of storage pools, which is a collection of tapes for a particular purpose. So you just turn encryption on on a pool, everything gets encrypted from that point onwards. And that’s using the hardware encryption on the drive or software encryption? 256-bit AES encryption in the drive. you take that tape, send it to your client to deliver something, send them the key via some secure.

means, not a post-it note on the tape. Hopefully not. And then they’ll be able to put the key in and read the tape. Excellent. that’s a big one. Because a lot of our customers have been asking for that. That’s a update, David. Thank you. You’re welcome. Anything else we should take as a takeaway from ARCUWARE? Just that we like the fact at the moment that we have a really easy way in a browser to browse a customer’s disk-based storage.

and then they can choose what they want to move to LTO. So they’re going to be freeing up space on storage, which is difficult to expand in the current climate and expensive in today’s current climate. So yeah, so it’s not easy to add another 50 terabytes even If you add an LTO, P5, you can use the simple web interface. You can just move stuff and so it doesn’t have to be thought of so much as a long-term archive, you can also think of it as shuffling around to be able to work within the confines of the amount of disk storage that you have. So I think that’s a way that tape is going to be used much more in the coming months until things level out again. I would tend to agree. You know, you guys were here at NAB, course, Media and Entertainment, and you guys excel in this domain, but you have so many features that seem enterprise friendly. Yeah.

How is the enterprise business for you guys? Are you still primarily selling to &E customers? We’re expanding a little bit further out. I work primarily with customers in the UK. And we’ve had quite a lot of education sales. kind of sciencey faculties within universities where there’s a lot of… Well, you have an API, which I find very intriguing for those kind of folks. And let’s face it. Linux people.

Yeah, and the future is API to API. That’s the future. And obviously, S3 is a big part of that API interplay between different vendors’ products. And the other thing I just want to restate is that we have the ability for tape to be addressed as an S3 bucket, which means that any software or workflow that can send objects to an S3 bucket and work within the deep archive mechanism where objects are flushed to tape and they’re going to take a couple of hours to bring back, then we can do that. So we can do exactly what AWS Deep Archive does, but on-prem customer owns the storage, they know where the data is, it’s in the building, they can do their own redundancy, et cetera, cetera. And on that note, that’s why we are coming to you today from the on-prem zone. That’s right. We’re in the on-prem zone.

Yeah, which is kind of a cool. surrounded by a bunch of products which are all about, you know, maybe repatriating data that’s in the cloud or just working more with what you’ve got on the front. Yeah, we’re going to have to as an industry work to expand that concept a little bit as a coalition, perhaps, federation. Well, David Fox, thanks so much for spending time. Always nice to see you. You’re welcome. we’ll see you next year.