Corporate BS debunked, this time RED

[blueskull]

Not the best voice I've heard, but definitely good contents.

How leader companies can lie so blatantly.

[bdunham7]

I'm not the least surprised that a premium-priced company just repackages ordinary stuff and passes it off as special or custom--that's been going on forever.  I'm surprised that they were so inept at the coverup.  I'd expect labels to be removed, devices potted and enclosures bonded together and maybe some pins swapped on the connectors.  Here you can just unscrew it and see what's what.  I think that they are depending on their customers being idiots.

[Mr. Scram]

That clip is all over the place. I've never seen such a high effort attempt result in such a low quality clip. They really need to work on how they bring their information.

This guy obviously doesn't know SSD manufacturers reserve part of the drive for reallocated sectors down the line. The card likely has the capacity they indicate, they just provide less of an invisible buffer. No software tricks are involved, which changes the story a bit. The guy also runs all the important tests on his own, aftermarket SSD. No shit that's the same as a standard SSD. He doesn't prove the firmware on the Red SSD is actually standard Micron firmware, he just shows that standard firmware is compatible with the camera though no filming is shown. In theory different firmware could even change a drive from MLC to SLC although the capacity shows that's not the case here. He also completely ignores the possibility the manufacturer is binning and cherry picking drives, which isn't uncommon for hardware and is actually a value adding process. He does mention the binaries being the same but glosses over that point the entire story hinges on so quickly it's hard to judge whether it has any merit at all. Although it's entirely possible that Red uses stock firmware and doesn't cherry pick a thing, if you crucify a company you should at least do your homework. This is just a sloppy half-ass job which doesn't provide the key evidence.

Looking into it more they're apparently selling "Jinnimags" and the guy posting this video is Jinni.Tech and apparently part of that Jinnimag endeavour. It'd be stupid to completely fabricate the story, but they obviously have a financial stake in the story themselves. Apparently both companies have been taking each other to court before the release of this video, which seems to indicate there's more to the story than this guy is letting on.

http://www.reduser.net/forum/showthread.php?155462-Jinnimag
https://ymcinema.com/2019/07/10/red-mini-mags-why-are-they-so-expensive/

[wraper]

He also completely ignores the possibility the manufacturer is binning and cherry picking drives

How you could even possibly bin off the shelve drive with off the shelf part number and off the shelf firmware. It's not a RAM stick which you could actually test for overclocking and bin. If they cared about quality a little bit, especially at this price, they would not use cheap consumer grade SSD. What I would really like to see, is what happens when different model SSD is used.

The card likely has the capacity they indicate, they just provide less of an invisible buffer.

Extra FLASH capacity used for internal operation of the drive, does not make it 512GB.

[wraper]

BTW, I immediately noticed that it's illegal to call it "Made in USA" and that guy actually made a second video where he talks just abut that as well. You cannot simply put foreign made PCB into case and call it "Made in USA", only "Assembled in USA" is allowed and in this case and it's still under question if it qualifies even for that.

[Mr. Scram]

How you could even possibly bin off the shelve drive with off the shelf part number and off the shelf firmware. It's not a RAM stick which you could actually test for overclocking and bin. If they cared about quality a little bit, especially at this price, they would not use cheap consumer grade SSD. What I would really like to see, is what happens when different model SSD is used.

Extra FLASH capacity used for internal operation of the drive, does not make it 512GB.

Note that no proof was provided that off the shelf firmware is used in Red drives. That's claimed in the video, but they neglect to provide anything tangible other than that regular firmware is superficially compatible. You could bin anything, so you could bin SSDs too. Especially if you have your own firmware the real capacity of the drive could be used, but if that's not the case that amount of NAND is still present on the drive. Red may not do any of these things and judging by their statements I gather they don't, but as I said this guy will need to provide something more solid if you're going to crucify a company. Especially if you've been in a few bitter court battles with them and have a financial stake in the matter.

[mariush]

Here's part 2 :

[Mr. Scram]

BTW, I immediately noticed that it's illegal to call it "Made in USA" and that guy actually made a second video where he talks just abut that as well. You cannot simply put foreign made PCB into case and call it "Made in USA", only "Assembled in USA" is allowed and in this case and it's still under question if it qualifies even for that.

It does seem to be a disingenuous claim, but I don't pretend to know enough about the US laws and Crucial manufacturing involved to say anything definitive about that. It seems Crucial has an assembly plant in Mexico, but I'm unaware of US manufacturing or assembly being done.

[wraper]

How you could even possibly bin off the shelve drive with off the shelf part number and off the shelf firmware. It's not a RAM stick which you could actually test for overclocking and bin. If they cared about quality a little bit, especially at this price, they would not use cheap consumer grade SSD. What I would really like to see, is what happens when different model SSD is used.

Extra FLASH capacity used for internal operation of the drive, does not make it 512GB.

Note that no proof was provided that off the shelf firmware is used in Red drives. That's claimed in the video, but they neglect to provide anything tangible other than that regular firmware is superficially compatible. You could bin anything, so you could bin SSDs too. Especially if you have your own firmware the real capacity of the drive could be used, but if that's not the case that amount of NAND is still present on the drive. Red may not do any of these things and judging by their statements I gather they don't, but as I said this guy will need to provide something more solid if you're going to crucify a company. Especially if you've been in a few bitter court battles with them and have a financial stake in the matter.

They don't have any custom firmware. Nor they possibly could as it turns out they use micron, kingston, Innodisk and other off the shelf consumer SSDs.

[wraper]

BTW, I immediately noticed that it's illegal to call it "Made in USA" and that guy actually made a second video where he talks just abut that as well. You cannot simply put foreign made PCB into case and call it "Made in USA", only "Assembled in USA" is allowed and in this case and it's still under question if it qualifies even for that.

It does seem to be a disingenuous claim, but I don't pretend to know enough about the US laws and Crucial manufacturing involved to say anything definitive about that. It seems Crucial has an assembly plant in Mexico, but I'm unaware of US manufacturing or assembly being done.

To claim "Made in USA" virtually all of production must happen in USA. If any significant part is not produced in US, claiming so is illegal.

[Dubbie]

Hrm, I've got a bunch of these cartridges.
Time for a cheap upgrade I think!

[m98]

It just wouldn't make much sense to bin those SSDs a second time. They already use a lower-grade model.
There also doesn't seem to be a need for a custom firmware, as the competition just uses CFast cards or 2.5" SSDs without any bigger problems.

[Mr. Scram]

They don't have any custom firmware. Nor they possibly could as it turns out they use micron, kingston, Innodisk and other off the shelf consumer SSDs.

Likely, but this guy didn't demonstrate this. When there's copious amounts of mudslinging going on we really need to have some solid proof. It should have been fairly trivial to show stock firmware being used, but Jinnimag decided to make an annoying video instead.

[Mr. Scram]

To claim "Made in USA" virtually all of production must happen in USA. If any significant part is not produced in US, claiming so is illegal.

That's indeed what I can find, but I won't pretend I'm a lawyer. The claim the drive is made in the US could be a flat out lie, or they have some kind of angle that permits them to make that specific claim.

[Mr. Scram]

Hrm, I've got a bunch of these cartridges.
Time for a cheap upgrade I think!

Time to open at least one up and check the firmware too. It'd be good to do some independent verification.

[Mr. Scram]

It just wouldn't make much sense to bin those SSDs a second time. They already use a lower-grade model.
There also doesn't seem to be a need for a custom firmware, as the competition just uses CFast cards or 2.5" SSDs without any bigger problems.

It doesn't make much sense and likely didn't happen, but that's not the point. The point is not to buy into the story of a party with conflicting commercial interests too readily. I suspect most of the story is true, but the video does a terrible job providing decent proof. It's silly to run tests on your own store bought stock SSD and showing all kinds of fairly irrelevant things while omitting the important bits and not showing the stock drives are as good for filming. This story begs some proper, independent verification.

[bdunham7]

Hrm, I've got a bunch of these cartridges.
Time for a cheap upgrade I think!

Time to open at least one up and check the firmware too. It'd be good to do some independent verification.

https://www.ftc.gov/public-statements/1997/12/enforcement-policy-statement-us-origin-claims

Seems a pretty clearly deceptive practice.  The angle that they have is that they have expensive lawyers willing to tell an egotistical CEO that he can get away with anything as long as he pays their outrageous bills.

Seriously--you think RED wrote their own firmware for Micron's SSD and didn't bother to use a different revision code for it?  You can check if you like, but I wouldn't even bother.  However, even if they did, they still have no cause to complain that Jinnimag's product works in their cameras because they haven't even alleged that Jinnimag used anything other than stock Micron firmware.  So where is the IP issue?  And the complaint that the RED camera reads the Jinnimag cartridge as a genuine RED product has nothing to do with Jinnimag and everything to do with the RED camera lying to the user.

I agree the video is annoying, but IMO the conduct alleged--and pretty well demonstrated--is outrageous. 

[mariush]

They claim they use special algorithms as to how they write data to the SSDs ... that's their proprietary stuff and all that crap.
It's just SATA connection, so not much they can do.
The only thing I could think of is that they're buffering writes in such a way that they're trying to always write blocks of 512 KB or something like that, basically an exact multiple of the size of a block on a SSD.
Basically, you can write pages of 512 bytes or 4096 bytes at a time (or some other value) on a SSD, but you have multiple such pages arranged in a block.  You can't erase a single page, in order to overwrite data in a page you must erase a full block, so if a block must be erased, the SSD has to read the pages with useful content, write those pages in other random blocks all over the SSD to minimize wear on individual blocks, then erase the block.  This process adds latency so it could slow down transfers in theory.

If they code their stuff to always write block size worth of bytes, they could more or less guarantee each time a write is done, a full block erase happens and the block is filled with new data.
However, this would require also modifying the firmware of a SSD, because right now even if a SSD gets a block worth of data in its write cache, it will spread the writes to random pages all over the drive, across all the channels (to increase write speed)
It would simplify the firmware of a SSD drive by a lot, to just always erase blocks and write new data in, as the controller no longer has to track pages all over the place, but the camera would have to keep track of erases better and eventually refuse to write on a SSD that has too high of an erase counter.

For example they have those 120 GB mags which have the Kingston 120 GB SSD with TLC memory that cost 20$ ... those basically have around 80-100 TB worth of lifetime writes.... basically fill the ssd 1000 times and you're done.
Considering the RED camera fills could fill 120 GB in around 5-10 minutes of recording... someone could easily kill such a  SSD in less than a year.

[Dubbie]

I opened up the smallest oldest one I have.
It's a 64GB one. Looks like an earlier revision than the one in the video, but same basic thing. Even less components on the adapter board, just a couple of 0 ohm resistors.
Looks like I'll update it to a 512GB

I'll let you know how I get on once the replacement SSD arrives.

I'm 99.9% certain that this is just a generic firmware. There is absolutely zero to suggest otherwise.

[Mr. Scram]

https://www.ftc.gov/public-statements/1997/12/enforcement-policy-statement-us-origin-claims

Seems a pretty clearly deceptive practice.  The angle that they have is that they have expensive lawyers willing to tell an egotistical CEO that he can get away with anything as long as he pays their outrageous bills.

Seriously--you think RED wrote their own firmware for Micron's SSD and didn't bother to use a different revision code for it?  You can check if you like, but I wouldn't even bother.  However, even if they did, they still have no cause to complain that Jinnimag's product works in their cameras because they haven't even alleged that Jinnimag used anything other than stock Micron firmware.  So where is the IP issue?  And the complaint that the RED camera reads the Jinnimag cartridge as a genuine RED product has nothing to do with Jinnimag and everything to do with the RED camera lying to the user.

I agree the video is annoying, but IMO the conduct alleged--and pretty well demonstrated--is outrageous.

I don't think Red wrote their own firmware. They didn't demonstrate that though and we should be unwilling to be recruited to whatever Jinnimags cause is until someone does so in a reasonable manner. These companies have been going at each other in court, which means we should be very wary of anything not demonstrated. If you can't manage to put the relevant bits in a video, you don't get to make these claims.

[Mr. Scram]

I opened up the smallest oldest one I have.
It's a 64GB one. Looks like an earlier revision than the one in the video, but same basic thing. Even less components on the adapter board, just a couple of 0 ohm resistors.
Looks like I'll update it to a 512GB

I'll let you know how I get on once the replacement SSD arrives.

I'm 99.9% certain that this is just a generic firmware. There is absolutely zero to suggest otherwise.

Excellent! It seems likely it's generic firmware but confirming that is definitely a good idea. I'm not sure whether it's possible to retrieve the firmware from a drive though. We may be stuck with performance comparisons.

[Dubbie]

according to the datasheet > https://download.siliconexpert.com/pdfs/2017/12/31/3/7/51/131/toshs_/manual/hg3_series_rev11.pdf

There is a command to "download microcode" not sure if that means to download firmware TO the drive for FROM it.

[BrianHG]

Ahhh, Red's equivilant to printer paint-jet cartridges.

[BrianHG]

according to the datasheet > https://download.siliconexpert.com/pdfs/2017/12/31/3/7/51/131/toshs_/manual/hg3_series_rev11.pdf

There is a command to "download microcode" not sure if that means to download firmware TO the drive for FROM it.

You can always check if there is a verify firmware command.  Just verify the SSD's listed firmware with the same version available from the manufacturer's website.

[Dubbie]

As far as I can see there is no verify command or anything like it.
I checked the ATA command docs, and there is nothing at all to do with firmware besides the DOWNLOAD MICROCODE command which is indeed for updating firmware. You just send the new firmware, and upon the receipt of the last byte, it starts executing the new firmware.

No facility for verification. Not sure what happens if there is an error in the transfer. I guess you brick the drive.
The only other solution is probably JTAG. There seems to be an 8 pin header on the bottom. I'm afraid that might be me out of my depth.