The Isostick, a Kickstarter project now in development, is meant to emulate an optical drive in the form of a USB stick. Although there is nothing new with putting an iso file (optical disk image) on a USB stick, what is unique about this drive is that it fully emulates a drive without actually having to worry about drivers or extra programs.
When plugged in, a computer sees a flash drive and a virtual optical drive. An excellent feature is that this disk can store multiple “iso” files and select them with a built in utility program. This could be invaluable for a technician or hacker since more than one CD or DVD is often needed to complete a task.
If fully developed, one could expect to attach this “drive” to your keychain and not have to worry about lugging an optical disk around anymore. Also, the activity light is programmable, which is a nice bonus.
93 thoughts on “The Isostick”
isn’t this alot like the S3 jumpdrives? from what i understand S3 was developed to create a virtual optical drive so it could autostart a program from the manufacturer. If this is just the hardware equivalent then i’m on board for one. Need one for a project i was working on.
$125 is an insane price point.
this one is bootable and requires no additional drivers.
All thumbdrives are bootable and require no additional drivers.
DO WANT. (Sucks reformatting same USB for a different ISO)
oh yeah, i forgot, this one is FAST. other ones use a CPU to serve up data (slooow) while this uses an FPGA.
i have inside info, so i know this for a fact. :P
Um, what? It comes with an FPGA?
I’m a little leery about these drives. I know that most of us are wise enough to disable CD autoplay, but this could make compromising computer systems way easier than it already was.
This sort of thing has been around for a while. There used to be a thing called an i-odd that did this, which has now become the Zalman ZM-VE200. It’s basically a USB HDD with optical drive emulation and much more besides.
So since windows machines can still autorun optical drives, this negates the fact that microsoft bought the s3 patent. Pretty sweet deal, except for the $125 “donation”. I know these guys are working hard on a product that will most definitely sell, but they have to come down on their selling point a bit before i buy one.
http://en.wikipedia.org/wiki/U3 claims that the U3 drives could be booted. And I’m pretty sure that they were reporting themselves as a USB drive, which is why they were used for a great deal of theoretical system attacks, because they could utilize windows autorun to propigate a payload without notification, where as regular pin drives won’t just autorun things. This combined with the tendancy for people to plug any usb drive they find into their computer proves to be dangerous. Plus those drives cost nothing. Of course, to access this partition yourself required a bit of software trickery, but for certain purposes (often nafarious, sometimes good) it works quite well. Or perhaps I’m missing something.
Its not quite the same as it isn’t a USB stick but Zalman do something similar called the ZM-VE200 (http://www.zalman.com/eng/product/Product_Read.asp?idx=431). I’ve got the iodd branded version of the same thing.
It is a USB 2.5″ HD enclosure with a built in LCD + controller that does this mount ISO as optical drive feature. It is significantly cheaper for the amount of storage you get but you lose on the size aspect.
Just ordered one of these. Does it work well?
I’d still use several usb sticks or microsd’s. It’s more flexible and if one would die I’d loose only 10 bucks or less. And no you don’t need to reformat a usb stick to put an iso onto it. Simply call ‘cat newimage.iso > /dev/sdX’
Meh, daemon tools lite (free) does this in software. However, I could see where it would be really handy if you are going between multiple computers and/or can’t install software. The price is way too high though.
Yes, but you can’t boot from deamon tools XD
Also, some software refuses to run if it detects Daemon Tools running.
E.g. Instead to having to use the original cd to play a game, or messing about with anti-anti-piracy, you can just use this and it thinks it’s a physical cd.
I just have five cruzer micro u3’s with my bootable iso’s on ’em ;)
A 32-Bit AVR With usb ong the go, a micro-sd card, and a USB jack total cost in ~100 quantity ~$25-30… $25,000 is a alot of money for that simple of hardware… and
if you look at the picture it’s not a FPGA, and switching to one wouldn’t make any sense the amount of dev work required would be way to high as well as increasing the parts cost significantly …
This isn’t something new, Its already been used in 3G usb dongles to store the necessary drivers and software. It is usually a 50mb partition in flash which is recognized by windows as a CD Drive. But yes if it can boot multiple iso’s it will come in handy.
Agreed. Seems to me that this is something anyone could build with an ATTiny and an sd-card. I’m not sure what the hype is all about, or why it costs $125…
@Decius YUMI http://www.pendrivelinux.com/yumi-multiboot-usb-creator/
Does this an alot more.
i got 16 gig of livecds and installers (including *nix bsd and windows) got about 15 diffrent things now.
Dosen’t emulate a cdrom but able to chainload iso using grub2.
Do you really don’t know how much it cost to do hardware development one a serious level as this. You need to include the 4 layer pcb cost, the man hours,… The usb vendor ID and maybe the FAT license.
Most poeple doesn’t know how handy it is. You just drag and drop your iso on to the stick. And you can boot that particular iso. You don’t need to do any other handling!
Again, this is something anyone could do with an AVR and an sd-card. I’ve seen the vfat source in the generic android code under the linux kennel somewhere, so I assume you can just use that from Linux. Shouldn’t be any licensing issues there.
I’m not following the large benefit overall, so maybe I’m missing something. Aside from the fact it’s hardware-based of course.
I can’t recall the name of the two tools right now, they’re on another computer than what I’m on, but there’s two utilities (among several) I use which format any USB stick, regardless of brand or size, and turn it into a bootable drive.
Not just recognized as an optical drive, but a bootable HD. This way you don’t have to enable bootable optical in the BIOS. Older BIOS require you to enable bootable USB, but newer ones you don’t. Tried on multiple systems, multiple USB sticks.
I currently have 14 ISO’s on there, as well as three different Windows installation packages streamlined.
Pop the stick in, a utility comes up and lets me select from the ISOs or enter a command mode (OS of my choice, several to choose from) as well as mount partitions present on the HDD’s it can see.
In addition, there’s no size limit and no particular format limit (such as having to have FAT32 on the first XYZ cylinders). There’s FAT32, NTFS and Extended partitions on it right now.
Aside from several days of gathering all of the software and putting it together into one quick 5 step system.. I spent no money.
Why would I spend $125 (I could see $30 or less..) on a hardware version of what I can do with my software-based system?
What am I missing?
“but there’s two utilities (among several) I use which format any USB stick, regardless of brand or size, and turn it into a bootable drive.” — ummm, you mean like diskpart?
I have had this guy’s website bookmarked for months I know its not the same group but they seem to have a more open policy about there project
@Spiralbrain and @that1guy daemon tools and S3 drives only work once windows boots. isostick makes any bootable iso a bootable cdrom on a flash drive, you can boot any official XP installer off a flash drive with this, no hacking or messing around injecting drivers or fiddling with bootloaders, just copy iso to stick, done.
@cirictech your math is missing some numbers, and a bigger picture. $30 is on the cheap side, don’t forget custom case and board and mass assembly, and his time to create, test, support, and improve this. Then, practical retail pricing practices say to set price at least double what he, as the creator, needs to get from it so that he can afford to buffer DOA’s and holding stock and managing shipping and so resellers can take their cut. $100+ is very reasonable for this in quantities < 1 million. assuming $125 is 50% markup, then the kickstarter is going to give him around $10-12K extra to order some to ship off to retailers, that's only about 200 devices over what's promised to the backers. $25K isn't a lot of money for launching a new product.
I’m sure $125 is very reasonable as far as what the developer would expect for the amount of work he put in. The problem is, he’s developing something useless. We can already boot from thumb drives (no need for useless .iso files.)
And there are free software based solutions to mount .iso file for every major OS in as many virtual drives as we like.
Using this would be a pain in the ass. Once your flash drive got full, you would have to delete files on it and copy new ones on to make room. Which is just stupid since you already have everything you want on your PC and can just mount it.
Also, clearly the developer is doing *something* wrong, because emulating a cd-drive and a thumb drive over USB is something that any AVR/USB hacker should be able to throw together in a free weekend with an sd-card, and make it open source.
the point of this is???
there emu software GNU free out there that does the same thing.
plus if you HAVE to boot from the CD this wouldn’t work because it’s USB.
There’s no point, the dev is solving non-existent problems.
Though you are incorrect about booting, as many PCs can boot from USB Optical drives, which this identifies itself to the PC as, so it would boot just fine. But still pointless as those PCs can usually boot from jump drives as well.
Sure but I have yet to read about speeds of them. I know for a fact the group I already linked to in my post above could only accomplish 2x CD drive which is very slow in today’s standards about 100k a sec if that. Not sure if this will be faster but without specs on it. I will continue to carry my disk around until 16x or so is reached because it still faster than something like this
@Gravis looking at the photo of the prototypes the main chip is an Atmel 32UC3A3128 which supports USB 2.0 with embedded host capability.
@cirictech I fully agree with you that there is no way that a FPGA would be used.
I do like the idea, very cool, presenting a fake SCSI CD-ROM at a hardware level. I suspect though, that the maximum CD size will be limited to FAT32 filesize limit of 4GB-1Byte. So it will not be able to support DVD and BluRay sized devices. Although maybe this could be added as a feature for future 64/128GB versions.
Good point, there aren’t really any good open source NTFS drivers, though, I suppose Fat64 is open, and actually designed for solid state drives, so the 4GB limit is surmountable.
Where can I find this software?
You don’t understand what this is it emulates a CD drive so when you are installing software Windows, Linux. You don’t need a CD driver or even a CD in the computer pop this in and it just boot from a the drive.
Removing All the problems with USB boot support
Removing all the problems with USB *thumbdrive* boot support. It’s still a USB Optical Drive.
Unfortunately for your argument, most of the machines which can boot from USB *Optical* drives can also boot from any USB drive.
I agree, pricing seems a bit excessive. The single most expensive thing is the sd card. $30 will buy you a 32GB SD qty 1. I realize there are sunk development costs, but a 32 bit PIC, crystal, vreg, sd slot, PCB and some software are all that is really needed to add to the mix. Probably less than $15 + SD card. $225 is too much for the 32GB version. I’d pay $100 for 32GB. Lower your prices and you’ll get more supporters for the project.
renosite.com’s project is very similar, but if you read his forum posts it still has a lot of bugs, and the on-offs cost $120 to make, not including labor. $125 for a clean finished product with support is sounding less high.
@willow, how do you get multiple ISOs and installed windows? What utility did you use to do that? I was thinking about getting one of the Zalman VE200 cases here: http://www.newegg.com/Product/Product.aspx?Item=N82E16817379021 But I could use some good stuff on my 16gb and 8gb flash drives.
Some people seems to missing the point here, the thing with this gadget is that it for the BIOS (or whatever) looks lie a cd drive and allows –>booting<– a single iso without messing with cds. Not really what 3G dongles or those S3s does… Yes, I'm aware you can flash some isos to a jumpdive and 'hey presto' but that never works easily when you really need it.. say windows for example.
A real pity that you can't pledge/pay trough paypal.. amazon is just wrong.. :(
No one is missing that. Those dongles present themselves as USB cd-drives as well, it just so happens that the iso they have mounted isn’t bootable (or particularly useful.)
This is nothing any dedicated hacker couldn’t put together in a weekend with their MCU of choice and an sd-card.
Even with that functionality, it’s still pretty useless. It’s cool. But no one needs it. The computers that can boot from USB optical drives can boot from thumb drives too.
I guess this would be good for testing linux live distros, or maybe a portable XBMC stick?
I could pay over $100 for this if it had 64, maybe 32gb+ of storage. Very handy tool that, although isn’t irreplaceable, makes things much handier.
The technology to have a BOOTABLE virtual CD/DVD on a USB stick is not new. Several USB manufacturers are already offering it. The only novelty I can see is that they have custom boot-loader and presumably software that’s not written in Chinese. But otherwise it’s an existing product available from many (most?) custom USB drive manufacturers/distributors.
@Doc As I said I’m not on the computer which has the tools. When I can get to the computer and look it all up I will.
All of my stuff (mostly utils) are on an 8gb stick.
As to how, if memory serves me, the stick boots up to a pre-ISO utility screen. I want to say it’s GRUB-based. You can make the menus using a simple text configuration file.
The selections simply point to folders with the ISO image, unless it’s a bootable folder itself then you tell it that it has to boot that.
It’s that simple off the top of my head, which is why I didn’t understand the whizz bang about this project (aside from the hardware aspect).
I use the tools I have to prepare the stick, once prepared I can format.. partition..etc the stick however I want or need. If I want to add an ISO or bootable folder I simply copy it over and add the item to the menu list. Done.
Would love to know! Sounds way cooler than having 12 thumb sticks for each thing I want to boot…
Maybe I’m missing something, but why not just burn a cd? Cost = 30 cents.
Fuck yeah, this would be great to have. Anyone know if beta testers are allowed?
I like the idea. What I’m not seeing here is an ability to store multiple ISOs and choose which to boot. If I’m gonna use this to boot live cds and I have to only put one live cd on at a time, this is barely different from making a thumb drive bootable with some software utilities. What would be nice is a small screen and a way to scroll through ISOs and choose which one to load.
@Mike: You want to boot from a CD but also have a mutable USB drive for persistent storage. Or you have a system without an optical drive (netbook) and you’re too lazy to figure out how to extract the ISO and make a bootable USB. Or you want to do Autorun attacks on old Windows systems.
not even a neat idea, why not just use tools already available to make a bootable usb drive. actually this seems pretty worthless to me. and is it me or are all these kickstart/other *cough* hacks *cough* pretty much e-begging for insane amounts of money to do something that is already done. Personally i have yet to see one of these projects that isn’t either a completely vaporware pipe-dream or an already developed product.
Classic biz proverb: “You have no product when the novel part isn’t marketable and the marketable part isn’t novel.”
This is just some hackers trying to get money for mixing baking soda and vinegar — a great experiment for the budding enthusiast, but not something they will fare well trying to sell to others…
I’ll stick with alcohol or daemontools for now since this doesn’t mention anything about safedisc or secureROM. If I want to compromise a computer system by rebooting onto my own OS, then a live USB with some form of Linux would be the best choice.
This is a good idea, but for non technical people, the rest of us have had this figured out for years.
This is nothing new at all. I have a few U3 Sandisk Cruzers which do the same thing. They show as a CD-ROM and a USB Drive under operating systems and are driver independent. They are also found in modern BIOSes as USB CD-ROMs so I use them to install operating systems.
Under window$, I use the excellent ‘Universal Customizer’ program and this will format, resize and ‘inject’ an ISO into the CD partition. It works nicely under XP with no problems but it works fine on Vi$ta and 7 using XP SP3 compatibility modes.
If you’re interested in this kind of thing, give this method a try. It’s much cheaper.
I’ve been doing this for years and my primary machines don’t even have optical drives anymore. ;)
Are you talking about XBOOT?
I just played around a bit with it and it seems to do what I want (build a bootable USB stick by simply dragging a couple isos on the xboot window).
the stick wouldn’t boot using qemu as supplied by xboot, but it did boot in reality.
I accidently left a movie on the particular stick, and that file didn’t get deleted…
xboot doesn’t seem to remember the settings of the usb stick you created. apparently it does not delete previous boot configurations created on a stick, but add any new isos to the already existing setup (i ended up having memtest86 several times on the boot menu).
and it did boot the iso of my old win2k disc.
now i need to get another 16gb stick, to create the ultimate boot stick, he he
Emulating bank buffering over a USB bus isn’t that hard really..it can be done in pretty much any FPGA using a OEM HWID..
Thanks for the post! I’d like to address some questions and concerns here. I realize I failed to emphasize a few important points, I will update the Kickstarter to reflect them.
U3 drives: yes, they do also present a CD drive with the contents of an ISO that you put on the drive. The major difference between U3 and isostick is the bootloader isostick gives you to switch between ISOs on the fly, and the fact that you don’t need to copy the ISO contents onto a special area, you simply keep the ISO file on the flash drive as you would any other file. There is no restriction on where the ISO file must be.
Existing software solutions: Very good point, the major difference here is there are quite a few utilities we have tested which do not play well with those utils, whereas they work fine with isostick because neither they nor the computer can tell there’s a difference between the isostick’s optical drive and a *real* optical drive.
The project at renosite: Yes, it is similar and more open. I wish them the best of luck, as competition breeds innovation! After all, that’s what keeps business interesting and fun.
My plans are to opensource the isostick at or before the time I stop supporting it. I would like to opensource it before then, but because the hardware is so simple I am wary of doing so, as you may understand.
Why $25,000: As mentioned by myself in the Kickstarter and by several people in the comments here, the costs to starting a small business are typically much higher than $25k. Test jigs must be built, setup fees must be paid for the PCBs, assembly, plastic enclosure, and numerous other costs must be covered when doing a small batch of devices. MicroSD cards that are both fast and reliable are notoriously difficult to obtain in quantities below a few thousand without resorting to buying them all through retail channels (not fun at all!).
As to the per unit cost, because I cannot anticipate the volume of demand I have to assume the worst-case sales for my $25k goal, and I have priced the units so that I can still afford everything at that quantity of sales. I realize it is pricey, and I hope to lower it later, if I can achieve higher volumes in production.
Lastly, I’d like to say thanks for all the comments even those which are critical, that’s what keeps people like me on our toes!
@Gravis thanks for the support, though yes, it does use a 32-bit AVR rather than an FPGA, oops. FPGAs were discussed, but AVR32 was more cost-effective. No worries, anyway :)
Best of luck. Hope you come up with something novel soon. And in the meanwhile, I hope you will be supporting Fat64 for images bigger than 4gb, and formats other than .iso…
Since this is basically just an AVR and a flashrom, the software better be good.
It *WOULD* be novel if it had some buttons (maybe protected by a sliding shield; 3 at a minimum) that could hot-swap between .iso files on the drive…
Almost forgot to mention speed, the write performance depends on the MicroSD card but we will not choose cards less than Class 4, so minimum speed being 4MByte/s. I will try very hard to source cards Class 6 and above.
Read speeds are on average 12.5MByte/s, sometimes they dip slightly above, but rarely below.
Thanks for posting up Eric.
OH! And for file size limitations: The software we’re writing to manage your ISOs (you don’t need to use the software, but we thought it would be handy) will auto-split ISOs larger than whatever the limits of your filesystem may be. You can also do this by yourself if you want.
The isostick will recognize files ending with a .1, .2, … and will read from the correct file when you read from the drive.
@Eric Agan: Will you have management software for Mac and/or Linux as well, or only Windows?
You know this sounds like an awesome device but the price point is about double what it should be.
“My plans are to opensource the isostick at or before the time I stop supporting it. I would like to opensource it before then, but because the hardware is so simple I am wary of doing so, as you may understand.”
Cool idea, but yeah, I’ll pass.
By your own admission there are already a few proprietary solutions on the market, and open projects trying to compete with them. Why would I donate money just to start another proprietary project? I would rather see the money go to an open project that benefits everyone.
@Sitwon Yep! It will work on mac/win/nix no problems. The bootloader may have issues on Mac, we’ll be testing that in the coming weeks.
Just updated the Kickstarter FAQ to reflect this, thanks!
“The bootloader may have issues on Mac, we’ll be testing that in the coming weeks.”
Most USB flash drives do not work with macs (they don’t even show up on the EFI boot menu). Only the ones that come with osx (for the air etc) show up (I think those show up as a disc).
So if your solution works in the EFI boot selector you’re at least halfway there.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)