[crazy][spam] was Re: This 1.5TB microSD is surely witchcraft
On 6/21/22, jim bell
I visited this link. I ended up spazzing out for a bit around other things, and then commented on the page at https://www.newsbreak.com/news/2643344097912/this-1-5tb-microsd-is-surely-wi... . When I commented, they asked me to register an account, which I did under junksuper@gmail.com , since they seem to market pretty aggressively. Here is my comment. Linebreaks were lost in the posting, unfortunately. I value my linebreaks. junksuper now hello newsbreak i am a cluster of vulnerable neurons who have visited your website. how do you make money? what can i do to help you make money? or would that not be a good idea? would it be better not to help you make money?it seems silly it has been so long without a microsd storage size upgrade. i understand that moore's law was a business agreement, anyway, with regard to how to plan r&d. guess something changed.there are a lot of solutions, like you can wire a huge harddrive to a microcontroller and proxy it into a microsd card slot by speaking the protocol. takes your own r&d. Reply
to state this more clearly: this advertisement jim bell posted is an entire article that is entirely an ad. it is not at all impressive that a 1.5TB microsd card is _in the works_. it seems clear that a cypherpunk named jim bell did not read the article, and likely barely comprehended the headline. could be wrong.
the deeper link is https://www.theregister.com/2022/06/21/micron_i400_microsd/ . the one newsbreak wrote an article about. here's a paste: Micron aims 1.5TB microSD card at video surveillance market Ideal for corporate fleet dash cameras, smart home security, police bodycams, VSaaS and more, says chip giant Dan Robinson Tue 21 Jun 2022 // 12:30 UTC 17 comment bubble on white EMBEDDED WORLD Chipmaker Micron is offering a microSD Card for embedded applications with an impressive 1.5TB capacity, enough to hold four months of continuously recorded security camera footage, according to the company. Announced at the Embedded World 2022 conference in Nuremberg, Germany, Micron's new i400 [PDF] is claimed to be the highest-capacity microSD card yet and was designed with a focus on industrial-grade video security applications. The device is sampling with potential customers now. Micron says the video security market is estimated to be worth $83 billion by 2030. In this category applications include corporate fleet dash cameras, smart home security, police body cameras, AI-enabled cameras in factories, and cloud-based video security as a service, or VSaaS. With its 1.5TB capacity, the i400 microSD card should be a good fit for video storage at the edge and hybrid VSaaS deployments, Micron claims, as it can store up to 120 days of video security footage locally, allowing users to prioritize which data gets uploaded and stored in the cloud. Micron i400 Micron i400 Beyond security video there are numerous other edge and IoT applications where large volumes of data need to be stored, managed, and analyzed, rather than simply transmitting it all back to a centralized data repository for processing. Micron claims that treating the i400 microSD card as primary storage for edge applications will give users real-time AI analytics and faster decision-making within smart cameras and other devices. The i400 has a mean time to failure rating of two million hours, and can concurrently handle 4K video recording with "up to eight AI events per second," according to Micron, which presumably means simultaneous access to the data for applications such as object detection and classification for license plate or facial recognition. DRAM prices to drop 3-8% due to Ukraine war, inflation Chipmakers to spend record $109b on fab machines this year Micron dangles predictable memory price agreements in front of vendors Epyc move: Micron shifts high-demand chip design apps to AMD The microSD card is manufactured using Micron's 176-layer 3D NAND process, which the company began using for volume shipments of products towards the end of 2020. Since then, Micron has revealed it has 232-layer 3D NAND under development with a roadmap to delivering chips with 500-plus layers in future, as reported by our sister site Blocks & Files. Micron also disclosed that its LPDDR5 DRAM embedded chips have received ISO 26262 ASIL D (Automotive Safety Integrity Level) certification for in-vehicle applications such as advanced driver-assistance systems. The certification reduces the burden on automakers by minimizing their need to build in additional mechanisms to mitigate risk, Micron claims. ® Get our TECH RESOURCES
The comments show the article is indeed interesting: "a mean time to failure rating of two million hours" It'd be nice to know the lifespan in write cycles per cell as well. Given that it uses 4 bit cells, that could be less than one might think as we're no longer talking binary at the cell level - it's back to analogue. 40Reply Tuesday 21st June 2022 14:27 GMTYet Another Anonymous cowardSilver badge Reply Icon Re: "a mean time to failure rating of two million hours" Except for police body cams where it will fail after every shot 60Reply Tuesday 21st June 2022 15:22 GMTeldel Reply Icon Re: "a mean time to failure rating of two million hours" Probably fail to turn on. Safer that way. You never know when a "convenient failure" inadvertently fails to happen and leaves awkward evidence. 00Reply Tuesday 21st June 2022 12:56 GMTVoiceOfTruth I joked about it years ago, maybe it will come true We will have ZFS on our phones to safely store the amount of data. 20Reply Tuesday 21st June 2022 14:27 GMTYet Another Anonymous cowardSilver badge Reply Icon Re: I joked about it years ago, maybe it will come true A raid array of SD cards? Actually given their small size, low write speed and unreliability - that's not a bad idea. Quick invent a silly name with no vowels and call the VCs 60Reply Tuesday 21st June 2022 16:10 GMTRAM Raider Reply Icon Re: I joked about it years ago, maybe it will come true Card Redundancy Access Protocol? 20Reply Tuesday 21st June 2022 18:23 GMTVoiceOfTruth Reply Icon Re: I joked about it years ago, maybe it will come true ZFS - ZFS Fone Storage. 10Reply Tuesday 21st June 2022 12:59 GMTTrevorH Yes, but how much? Enquiring minds want to know
I don't have a bunch of 1.5TB microsd cards to turn into a tiny raid array just to use raid arrays. But the idea of turning something nonstandard into a raid array sounds cool. I don't know, maybe a whole bunch of tiny microsd cards? maybe some flash chip? something you could connect a lot in parallel, and get a device view of. maybe usb keys? you could make a raid array of USB flash drives ! what kind of raid array would be good? You could put some on a network, some on different USB hubs, for different redundancy factors. With RAID 1, speed would be _blazingly fast_. with RAID 0, you'd have _tons of space_. is there a way to do both? is that what raid 2 is?
the speed increases are the nice areas of RAID with raid 1, you can read really fast. i wonder if that's economical for mining read-speed-limited cryptocurrencies. with raid 0, you can write really fast because it's in parallel. _and_ you get tons of space. _but_, you don't get to read really fast. and you don't get redundancy, i think if your brain is working, there's a raid with a higher number that does both raid 1 and raid 0, so you can read fast, an duse a ton of disks to write fast, and you get failure redundancy. but_ you still don't get backups with raid, which means if a political enforcement group brainwashed you never to engage in forensics again, you'll have trouble recovering your data when you hit the writelimit of your flash.
i guess it's not a write limit thing, more a raid corruption thing y'know, i tried raid for a bit, it is fun, but i noticed the linux driver code seems very unmaintained / unfinished people have been dealing with issues with raid for years, that are just bugs or missing features sticking around that whole time it's also pretty silly we still differentiate between monolithic and microkernels
unless you're a normie king! the wealthy, comfortable, esteemed people who don't understand their technology or how their culture was built by a bunch of marketing firms then you can just upload your files to a factory with a picture of a cloud on it and when the factory breaks it doesn't matter because support doesn't have to answer your calls
long ago there were various fads like somebody made a filesystem that stored files in gmail accounts that's stuff you could put a raid on, right there.
no, i'm not set up to mount a cloud device at this time but i recently learned how easy it is to set up and use rclone. all you have to do is give root access to your system to the internet: curl https://rclone.org/install.sh | sudo bash then configure and mount something: rclone config create newsbreak http url https://newsbreak.com/ mkdir newsbreak rclone mount newsbreak: newsbreak this mounts the newsbreak website as a folder inside your filesystem. $ ls -l newsbreak total 689 -rw-rw-r--. 1 user user 82542 Dec 31 1969 following -rw-rw-r--. 1 user user 200758 Dec 31 1969 headlines -rw-rw-r--. 1 user user 179481 Dec 31 1969 local -rw-rw-r--. 1 user user 23070 Dec 31 1969 newsletter -rw-rw-r--. 1 user user 93914 Dec 31 1969 privacy -rw-rw-r--. 1 user user 123061 Dec 31 1969 terms
my comment is not on the newsbreak article! this started happening to me around 2014. i can infer from the massive news on it, that it has happened to many other people. i wonder how it works. i should probably make another. luckily i saved it as a flurry of overwhelming spam to a community mailing list.
junksuper now delete newsbreak, if your reply is delete-your-comment, then i think ihave received it, not sure. i am pasting my comment back in, because we need to preserve public stories. hello newsbreak i am a cluster of vulnerable neurons who have visitedyour website. how do you make money? what can i do to help you makemoney? or would that not be a good idea? would it be better not tohelp you make money?it seems silly it has been so long without amicrosd storage size upgrade. i understand that moore's law was abusiness agreement, anyway, with regard to how to plan r&d. guesssomething changed.there are a lot of solutions, like you can wire ahuge harddrive to a microcontroller and proxy it into a microsd cardslot by speaking the protocol. takes your own r&d. Reply 0 junksuper 1h ago delete hello newsbreak i am a cluster of vulnerable neurons who have visited your website. how do you make money? what can i do to help you make money? or would that not be a good idea? would it be better not to help you make money?it seems silly it has been so long without a microsd storage size upgrade. i understand that moore's law was a business agreement, anyway, with regard to how to plan r&d. guess something changed.there are a lot of solutions, like you can wire a huge harddrive to a microcontroller and proxy it into a microsd card slot by speaking the protocol. takes your own r&d. Reply 0
Comments / 1 What are your thoughts? Post Community Policy default-avatar junksuper now delete i am commenting because my interface appears to be misbehaving, so naturally i want to discern its patterns of misbehavior so as to use it in the future in a way i can rely on Reply 0 junksuper 1h ago delete hello newsbreak i am a cluster of vulnerable neurons who have visited your website. how do you make money? what can i do to help you make money? or would that not be a good idea? would it be better not to help you make money?it seems silly it has been so long without a microsd storage size upgrade. i understand that moore's law was a business agreement, anyway, with regard to how to plan r&d. guess something changed.there are a lot of solutions, like you can wire a huge harddrive to a microcontroller and proxy it into a microsd card slot by speaking the protocol. takes your own r&d. Reply 0
it looks like: - on the main site it shows no comments - when i deleted my comment, it showed no comments in the comments list - when i make a comment, it shows a retained list of comments that respect the behaviors of adding and deleting single comments
i visited wikipedia. a 1TB microsd card cost $449 in 2019. This is 2022, 3 years later. a little scary! is this an information bubble i'm in?
i guess i'd rather build my own microsd card than plot a chart of economic factors. i'm thinking it'll probably take me 30 years or so to build a microsd card, so i better get started.
the SD specification is huge. there are "simplified" specifications available at https://www.sdcard.org/downloads/pls/ . they ask for registration for the full things. The first simplified specification is called Part 1: Physical Layer. https://dweb.link/ipfs/bafybeid6u6p3tkr5a6fiyfqxjsvy2whdwlz2uvinqvcqvt6jq7oq...
The basic SD Memory Card communication is based on 9-pin interface (Clock, Command, 4xData and 3xPower lines) designed to operate in at maximum operating frequency of 208 MHz and low voltage range. Additional communication methods, based on differential signaling interface (UHS-II and PCIe/NVMe) are specified as optional. The communication protocol is defined as a part of this specification unless specified otherwise (i.e. PCIe and NVMe cases).
It lists specifications: - audio specification - file system specification - security specification - physical layer specification (includes basic command protocol) the mechanical section describes a "mini" SD between micro and SD. didn't know about that one. - mc-ex interface specification mobile commerce - sdio specification sdio sd cards are sd cards that have IO peripherals, and can only be engaged with sdio-capable interfaces
-> SDXC is up to 2TB -> SDUC is up to 128TB so we could crack the market with existing SDXC technology. by crack i mean stimulate.
suddenly this is very hard for me to continue! it was really nice to move the spamming towards project-like behavior though
y'know it's hard to tell what's real without having things you know are real i probably dissed a real article about a real new technology posted by a real person i imagine coronavirus and decreasing use cases have slowed the storage curve
3.6 SD Memory Card--Pins and Registers The pin layouts differ between SD cards (9 pins), UHS-II cards (second row of UHS-II pins), and SD Express cards (second row PCIe/NVMe pins). Looking at a standard SD card, contacts at top, the diagonal inset is in the upper left. The first pin, inset below it, on the left, is #9. Then they proceed #1 -> #8 from left to right, with the write protect switch if present on the very right edge. The card has two modes, SD or SPI. The pins have different functions in these modes. Each pin can be any of power Supply, Input, Output, or i/o using Push-Pull drivers. Ground: Pins 3 & 6 V+: Pin 4 CLK: Pin 5, input only SD mode CMD: Pin 2 Data IO: Pins 7, 8, 9, 1 are Bits 0, 1, 2, 3 respectively Notes: Data Bit 3 is also Card Detect Data Bits 1 and 2 are also used for signals in SDIO cards. SPI Mode Chip Select CS: Pin 1 Data In DI: Pin 2 Data Out DO: Pin 7 Notes: Pins 8 & 9 are reserved in SPI mode. CS has a 50KOhm pull up in the card at power up. The host can drive the line high or let it be pulled high to select SD mode. If the host wants SPI mode, it should drive Pin 1 low. To detect the card, the host detects that the line is pulled high. The Pin 1 Pull-up should be disconnected by the user, during regular dta transfer, with SET_CLR_CARD_EDETECT(ACMD42) That's page pdf page 32, physical page 12. The next page shows the internals of a card and describes the information registers. A description of the command protocol was in an earlier chapter; I skipped it since voltage information is more satisfying to start with when prototyping.
participants (1)
-
Undiscussed Horrific Abuse, One Victim of Many