Slashdot Mirror
Samsung Develops 16Gb Flash Memory
nofrance writes "As promised earlier this year, Samsung has unveiled the world's first 16-gigabit flash memory chip. These chips, when combined in a 16x16 configurations, will allow 32 GigaByte flash cards. Using 50-nanometer manufacturing technology, these chips will be in production by the second half of 2006, with Samsung promising that their 32Gb team will impress next year." From the article: "According to the company, the cell size of the fingernail-sized flash chip has been reduced about 25 percent from that of the 60 nm 8 Gbit NAND: The new 50 nm flash memory contains cells that measure 0.00625 square microns per bit. The 16 Gbit device holds 16.4 billion functional transistors, Samsung said. "
Although I would probably need at least 20..
With storage capacity becoming a big issue on the newer, more powerful cell phones, this flash memory comes in at just the right time.
For everyone who wants to buy Apple, this type of memory will be very nice for your Rokr. For everyone else, it'll be nice for your MP3 player phones.
Jesus saved me from my past. He can save you as well.
guess I should hold off on that Apple iPod nano, eh?
The revolution will NOT be televised.
FTA:
The 16Gbit device holds 16.4 billion functional transistors
Woah, that's a relief. I was afraid that I might be buying a device with billions of non-functional or even disfunctional transistors.
Now that Samsung has distinguished this for me, from now on, I'm going to make sure all the devices I purchase have fully functional transistors.
Falun Dafa is good!
Cue the "is that a 32GB pr0n flash card in your pocket, or are you just happy to see me" jokes.
Coming to an iPod near you....
Can this be put in an unpowered thumb drive? I feel it would be nice to have large, easily removable, USB storage that does not require external power. Right now, I store my accounting files on a 64MB stick that I can remove and take with me in an emergency much easier than taking my whole computer. The more room for backup, the better.
The previous comment is purposely vague and generalized, but all of the facts are completely true.
That outta be enough for anybody!
Please let me know when we no longer need hard drives, and we no longer need to "boot" our PCs every time we switch them on.
Also drop me a line when we can store the world's music on a small memory cube and download it at the speed of light, virtually killing the RIAA overnight.
Amazing, the tech just keeps getting better and better.
READY.
PRINT ""+-0
...at this point, is not really so much of making 32GB flash drives. Maybe I just don't see the application, but unless this is really competitive, costwise, with traditional hard drives, it seems impractical. As far as I know, most solid-state computing devices have no need for 32 GB of space. While interesting from a theoretical standpoint, storage media of this type seem to be inordinately expensive when compared with the average hard drive.
Going back to school for entry-level jobs?
When I look at local computer parts prices, DRAM has been stuck at the $100 / GB range for three years now. Flash passed its price point earlier this year and is not looking back. I used to marvel at how RAM prices used to drop. (Flash is slower and can only be written a limited number (1E5) of times.)
My RAZR has something pitiful like 3mb storage. It's the only thing that irks me about it.
Though gigs and gigs wouldn't convince me to buy a Samsung phone. They're like toy mobiles -- cute, simple, feature-light.
Yes, but individual memory chips often do not come rated in bytes but in bits and are configured in parallel to complete the byte. Hence "16x16 config" making 32GB.
Step 1: Pull out fingernails
Step 2: put fingernailsized flash memorychip on place of fingernail
Now just a way to power them up and use them. Any ideas
My wife's sketchblog Blob[p]: Gastrono-me
Sure stuff may be -smaller- now. But what about cheaper? Will it cost less than equivalent in smaller chips? As for memory density, 1 or 2GB SD cards are quite tiny already, stuffing the same technology in volume of a harddrive (well, CD-ROM maybe) would allow for a terabyte of solid state storage easily. But the price and speed are somewhat beyond reach... I really wish for CHEAP flash more than for BIG flash.
Anagram("United States of America") == "Dine out, taste a Mac, fries"
so should I wait to buy the iPod nano or not?
MORTAR COMBAT!
This must be a John Titor's quantic functional transistor desing.
I guess I should be amazed but to me its seems common sense. Imagne, an implant that you store data on and access any time you want. Oh yeah, I forgot to mention that it is inside your head. Now that will be the way to study. Download a few books, wait what am I saying... porn and away you go. Alot of people are scarried of the future but I say plug away; I'll be the first inline (no pun intended).
I eat Karma for breakfast, lunch, and dinner. That's why I don't have any.
Maxtor, Seagate, Hitachi, Western Digital and all the other storage companies better listen to what Samsung is doing here. Life is good when you are sitting in front of a really fast computer, but it's rather disturbing that the hard drives (and media players i.e. DVD) still operate at milliseconds instead of nanoseconds.
Has anyone thought about why hard drive development is so focused at increasing disk space by using similar technology and nothing beyond that? I mean, come on, this tech has been around for ages and you'd kind of want a solid replacement (read: no moving parts, nanosecond operation times).
Who knows what we'll see next in terms of hard drives AND WHEN?
Full Tilt
Uhm, the story is about 16Gb (Gigabit) chips. When talking about capacity of memory chips, bits are used instead of bytes.
"16 Gbit is eight times less than 16 Gbyte."
Yes, that why you need 32 of these chips to get 32GB capacity.
But then 32GB appears to be fabricated by conventional means rather the new unobtanium substrates used by AtomChip.
When the people fear their government, there is tyranny; when the government fears the people, there is liberty.
So that makes it 1^5, which is still 1.
If you want to write 10,000, why not just write 10,000? (Geekdom is one thing, readability is another, and better, thing.)
No. Each chip is 16 gigaBITs. In a 16x16 grid of these flash chips, you'll have 32 gigaBYTEs of storage.
Wouldn't it be better to just glue it on or something?
Will the Nano be upgradable? that is, was the chip oldered in or is in in there in a stadard flash drive socket. If so did apple or the CPU maker, cripple the nano's address range? if not buy that nano now and upgrade it next year. On the other hand the Nano sells for about $30 bucks more than the retail price of the 4Gb NAND chip. Son unless you can buy it below wholesale like apple, you'll be better off buying a new Nano when the 32 GB ones roll out.
Some drink at the fountain of knowledge. Others just gargle.
Err, it is Gb. The individual chips are rated in Gigabits, and only the final 16 chip products in gigabytes.
Thus the 16 Gb chip is 2GB and when you have 16 of those you get, you guessed it, 32GB.
East Coast Brewers
So how many non functional transistors did they put on it?
Deleted
I shoot weddings. With my 10D I get approximately 540 images, RAW, written to the MD. I'll usually pound thru 4 batteries (2x2) in the course of a day event; I have 6 spares.
Assuming I win the lotto and/or can reinvest some of the wedding profit towards a camera instead of my leaking roof, I would move up to a 1Ds, selling for 3K, which writes out 11mb RAW files.
That means a 32gb CF card would store: 2400 images
Your typical wedding/reception lasts 7 hours. Add a couple of the bridesmaids getting dressed (You do NOT want to miss that, HAHAHA) and you're at a 10 hour day.
That means you're taking a frame about every 15 seconds, were you to fill that up.
Cost of film? Let's say you're shooting 35MM instead of medium format (arguably a 1DS is a little less in terms of quality than a Hassy at 16x20, but the customer would probably never see it) then thats 67 rolls of film. A propack of 400NC from BH Photo is 28.45 for 5 rolls, which translates 14 packs at a cost of 400$.
Plus processing, tack on about 10$ per roll and you're at $1000 worth of money.
Where am I going?
No one shoots 3K worth of photos. It's insane. It's insane by even MY standards. But on a trip it's definately worth it to have... and I'm not even adressing the transfer rate issues (my firewire transfer from CF is the fastest in the market at 7MB/sec that would take about 1.25hrs to transfer)
This is an incredible leap forward but the biggest advantage will be the price pressuer on lower sized cards.
After all, drop one of these babies and you're out a pretty penny.
16GB is not enough...speed up the curve please
He who knows best knows how little he knows. - Thomas Jefferson
...instead of being called a "thumb drive," the relative size increase will force USB drive manufacturers to start marketing the much heralded "stump drive," and the much-less popular "double-stump dongle" version.
Tim
Yes, but that's not what the news said before it was edited. :) Now it's okay, so my post is just useless now.
Full Tilt
Marvelling at the iPod nano, this makes me wonder about the usability of a nano-thin laptop. I mean, at some point, you need the actual keyboard depth for tactile reasons, don't you?
I'm all for a 2-pound 17" widescreen aluminum Powerbook that's under an inch thick (is it even thick then?) -- as long as I have a decent keyboard so I can launch all my attacks with confidence in World of Warcraft.
--- -a- "I'd love to change the world, but it'd be easier if the universe exposed its API."
Yep, and suddenly find out that your 64Kb iButton fits 8 kilobytes of data, then find out it actually is 64 kilobits...
Sorry, whoever modded the above "informative" has no clue. Chip memory capacites are given in bits, not bytes.
The chip is 2 gigabytes.
Anagram("United States of America") == "Dine out, taste a Mac, fries"
I was going to point this out, but luckiy I RTFA first.
Has anyone taken a bunch of the already available monster flash drives and built a PC on them?
I'm thinking 4x USB2 card readers (these are down to like $10 on eBay) each containing 8GB compactflash in a RAID-0 configuration = 32GB solid state storage that might not incur too bad a performance penalty.
With something like a 32GB compactflash, you could potentially create a 120GB RAID-0 with them.
Do CF cards have the reliability factor to act as primary storage? How about USB2 as the interface? I don't know enough about either set of specs to make a judgment.
STOP . AMERICA . NOW
Next, the 2GB has Toshiba Flash Memory Soldered to the board, whereas the 4GB has a daughterboard with 2x2GB Samsung chips. Therefore, it is possible that someone will reelase an upgrade to the 4GB Nano at some point in the future, but Apple may well have disabled support in the (closed) Nano sofware for flash support above 4GB in the current generation.
Get a free iPod Nano 4GB!
Of course it would have to be left on with a tiny amount of power, and automatic backups...
I suggest you read Slashdot
You left one out...
Step 3: Profit
Just to translate for the masses, that comes to roughly .0032 LOCs (assuming 1 Library of Congress = 10 TB). Sometimes terms like "Giga Bytes" can get confusing.
Get ready for the iPod Mano - Man-sized storage in a Nano package.
I might be missing something here. 16Gb is 2GB. There are 2GB flash chips already shipping in the iPod Nano. This is the first 2GB flash chip. Either this is very old news, or the important thing is the size of the chip rather than the fact it exists.
I am TheRaven on Soylent News
It's too slow.
It also has a finite number of writes that can be done before it quits working.
If you want your system to run faster, look at the gigabit ramdisk PCI cards that are coming out this month (?). Get four of those, a raid card, and hook them up together. Contents are kept even when the computer is switched off.
SearchIRC - Now with live chat directory!
Hum, I more or less thought step 0 or 1b would be missed:
0. Inject lidocaine in fingertips and wait until you can not feel anything with it anymore
or 1b
1b. Scream in excrutiating pain.
My wife's sketchblog Blob[p]: Gastrono-me
The person who wrote TFA said it's both 16 GByte and 16 Gbit. Read it, you'll see that both are used throughout the article. So we'll never know which one it is.
Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo!
I think the 2GB modules in the nano consist of multiple flash chips, like on a RAM stick; meaning that a similar module with these new chips would be several times larger. But I'm just guessing...
Am I the only one who thought the harddisk could be repaced with this? A minimal windows + office install easily fits 16gb. Ideally with no swap file given enough ram. Additional software may be run off a shared folder.
And if the windows (or linux) installation contains enough drivers, you could have a USB2.0 flash drive with 16 or 32GB space and carry the whole os around.
I know this is easier with knoppix on usb, but I'm thinking big, with the current windows install base. This can do wonders for the corporate maintenance until linux is ready for the desktop.
"Give orange me give eat orange me eat orange give me eat orange give me you." -Nim Chimpsky
And here I thought a 640k flash card ought to be enough for anybody.
The Russian Mafia will mod you down just to see if the Moderate button works.
Around the end of May, there were several sites
- on-Flash-disk-drives-2222.shtml
http://www.theinquirer.net/?article=23425
http://news.softpedia.com/news/Samsung-is-betting
reporting that Samsung would be having a 16GB flash hard disk (SDD) available around August 2005. Has anyone seen those? I know for a very good reason that I would be insterested in installing one of those in my Powerbook: the joy of silence.
Actually, nevermind that... Just re-read it and I misread part of it before.
Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo!
Now just a way to power them up and use them. Any ideas
Well you could use body heat, but you don't want to know where you have to put that finger to get the most efficient heating.
"I am the king of the Romans, and am superior to rules of grammar!"
-Sigismund, Holy Roman Emperor (1368-1437)
Correct me but: isn't 16 gigabit = 2 gigabyte. So 16 chips are needed to get 32GB capacity.
Also, a "16x16 configuration" is 256 chips, which gives 512GB, not 32BG.
If you haven't figured it out yet....let me explain. Hardware is much more advanced than software or for the consumer. This technology could have...probably should have been availble a few years ago. It is a smart move to stagger the release of technological advances for monetary gain.
For the pron conniseurs out there..."if you blow your load to quickly...," what do you have to gain?
With bottom being a combination of how cheap they can manufactorer the stuff with current processes and the profit margin they're willing to live with.
The person who wrote TFA said it's both 16 GByte and 16 Gbit. Read it, you'll see that both are used throughout the article. So we'll never know which one it is.
The only mention of anything being 16 gigabytes rather than 16 gigabits says the chips "if combined in 8x16 and 16x16 configurations - theoretically enable Flash memory cards with capacities of 16 and 32 GByte".
There are no other mentions of 16 gigabytes, but there are four mentions of 16 gigabits.
3x is typically the image increase that you see transitioning to digital, so divide my numbers by three and you'll get accurate film usage :)
Can you imagine that? Hard-drives without spinning parts!
:D
They will have to quadruple the throughput and we will have competitive hard-drives with seek rates to the order of nanoseconds.
You know, they could even replace CDs and DVDs:
- Data rate high enough for HD-DVD or BR quality
- Put them into a good plastic case (ala zip disks, but smaller)
- No scratches!
Sounds like the 21st century to me.
If Samsung plays its cards right, they can make some serious dough with that technology. We're almost there.
Giggidy-Giggidy
"All you have to do is be fragile and grateful. So stay the underdog." Chuck Palahniuk, Choke
In that language, 1E5 actually means 1*10^5.
C - the footgun of programming languages
Probably a slower boot than off of the HD, but runs much quieter.
Best Slashdot Co
This is long, and you may not read it all, but it offers a lot of insight into the New Orleans situation ..........
SPEAKING TRUTH TO HYSTERIA
The rains from Katrina's aftermath had barely begun to taper off before the utterly predictable, knee-jerk, blame-Bush for everything hysteria began to rage. The attacks are loud, strident and given top billing by the media, who have shamelessly and blatantly added their own negative, anti-Bush spin without investigating the facts or questioning the political motives of the critics.. It seems, those of us who look to the actual facts before we draw our own conclusions are forced to endure a hurricane of rhetoric, speculation, and just plain nonsense. So don your waders, as there are some actual facts amidst all the debris. Let's start with the tin-foil hat stuff.
THE THEORIES OF THE LUNATIC FRINGE: THE DEMOCRATS AND THE MEDIA.
CLAIM: Global warming is Bush's fault and global warming caused Katrina.
First, of course, hurricanes in the US have not in fact been increasing in number or intensity since the supposed onslaught of global warming. As this schedule from the national weather service ( http://www.nhc.noaa.gov/pastdec.shtml) shows, both the number and intensity have actually decreased in recent decades, and were generally much more severe in the first half of the century.
More to this point, however, let's accept the global warming alarmists on their own terms, and assume global warming caused Katrina without inquiring too closely into what caused all those other hurricanes. Had Al Gore been elected President in 2000, and the day after his inauguration managed to get the US Senate (who had rejected the Kyoto Protocol
95-0 during the Clinton administration) to ratify and then fully implement its provisions in the US immediately, and had every other major country in the world also done so, the projected decrease in global warming after 20 years was projected, by its own proponents, to be only 0.7 degrees centigrade. In the first few years, that is, by now, even if it had been implemented in 2000, the supposed decrease would be essentially zero. So there is simply no conceivable scenario in which Bush's policies could possibly have had an impact on hurricanes. These claims, retailed widely in major US newspapers and by German and other European politicians, are nothing but despicable political posturing.
CLAIM: The Iraq war has dangerously depleted our National Guard Resources,and that's why help took so long to arrive.
Facts can be your friend. There are roughly
1,000,000 army personnel in the US, including active duty, National Guard and reserves. A bit over 100,000 or 10% are in Iraq (the rest of the forces over there are from the other branches of service). The Pentagon has agreed with the states that it will not mobilize more than 50% of the National Guard from any state, and only about a third of Louisiana's National Guard is on active duty. The National Guard units that have been mobilized for active duty in Iraq are for the most part heavily armored combat units, not the more lightly armed military police and search and rescue units that are the primary source for domestic disaster support.
You might never know this if you watched network news, but the Commander-in-Chief of each state's National Guard is the governor of the state, not the President, unless and until the National Guard units are called by the Department of Defense to active duty. It is also worth noting that it is against federal law, the long-standing Posse Comitatus law, for active duty troops to be used for law enforcement-the only National Guard troops that can be used this way are those commanded by state governors. There has been some talk since 9/11 of repealing or amending the Posse Comitatus law, but the changes were strongly opposed by groups from both the left and right.
Would it also be crass to point out the undeniable fact that by Sunday, September
The third comment on this page (if they know anything) says that the 2GB modules in the Nano are made of 4 500MB chips, so these new 16Gb (2GB) chips would allow for 8 and 16GB Nanos.
power by electricity conducted by the skin, then using m$ patented PAN (personal area network) you could sync your files while you had sex, just hope your not quick in bed with these 32Gb flash cards come out.
I was under the impression that storage is usually givin in *bytes ; anh network traffic more so in bits.
typo?
What?! I think if you are going around, shooting weddings, and making profit off of this grisly business, solid state drives are the least of your concerns. I would be more worried about things like: law enforcement, two families revenge, and Kaiser Sose.
In related news, you may want to do us all a favor and put down the violent video games. Thanks!
Side note, solid business model:
The IceMan, from HBO's Iceman confessions, would be proud.No one shoots 3K worth of photos. It's insane.
What makes you think these are intended for photo cameras? I'd want a 32GB card to stick into a special digital camcorder, since that's about 3 hours of DV-quality video footage in random-access format. Being able to put hi-res photos and video on the same card is just a bonus.
Well, you jest, but it's not far from something I happened across yesterday...
This article outlines an idea of how to store up to 5Mb in a fingernail, using lasers. I think maybe they're a little over-optimistic with the idea of storing biometric info there though.
But Zorak.. Where are we gonna get enough squirrels to build a kite and fly it to the moon?
heh?
16 gbit == 2 gbyte
you need 16 of these to get 32gbytes, not 32.
[SHOW SOME LENIENCY TOWARDS
Since I don't wear shoes, I wonder if I could use these in place of my toenails. I could probably fit two on each of my big toes alone.
I thought all the big flash devices where using some kind of multi-bit technology that makes it possible to save two or more bits per transistor/cell.
But "The 16 Gbit device holds 16.4 billion functional transistors" sounds like they got one transistor per bit+ some logic, drivers etc.
Jan
It sure is confusing for the most of us, but a flashcards exist from multiple chips. The biggest chip till now was 0,5GByte or 4Gbit. Samsung now has the new record with a 16Gbit(2Gbyte) chip while the actual size of the chip is the same.
You guys really should try at least to read the TFA. It says the chip is 16Gb, running in a configuration of 16x16 yielding 32GB (yeah, thats Bytes).
(No I'm not new around here, but comeon, lets start a trend and at least read some of the posted stuff before bashing)
I believe they ran that story a while ago on /. too. The point is that it grows out in several months (data loss due to cutting your nails (-: ).
You gave me an interface idea though: Combine it with a bluetooth chip, and use the just mentioned MS patent for creating electricity, and it could even run.
(Stupid joke now taken to seriously)
My wife's sketchblog Blob[p]: Gastrono-me
Yeah, I noticed that after posting, hence my reply to myself saying that I'm an idiot. :)
Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo!
It has been said before: Flash can be fast, because you can have many drives in parallel. Chips survive lots of rewrites, because they remap logical blocks when phisical wear out.
I have a webserver and I could EASILY fit the whole drive on one of these.... Talk about an EASY backup... I could have a cron job "dd" the drive every night at 2am off to a USB drive...
(and of course, ssh transfered to a remote server not on the same power grid)...
td
hard core geek-ware
Well DUH, Mr. Wal-Mart shopper! Today's 2 GByte cards are made from multiple chips. While this new chip is pricey--given it's from a new process--it's only one chip.
So yes, it's already a fraction of the overall cost.
The same process, if applied to smaller density chips, would make smaller size cards even cheaper than they are now.
Every advance in chip density means lower cost for any older design made in said new process. Hint: it's the same reason today's consoles now cost half what they did originally.
Now go back to K-Mart and find your blue-light-special.
On hearing a heckler in the front row question his sanity, George Carlin replied... "Nice...I see you've been given the gift of a functional brain - please let us know when you unwrap it and take it out of the box."
...for something to be a carburetor is for it to mix fuel and air in an internal combustion engine--carburetor is a functional concept. In the case of the kidney, the scientific concept is functional--defined in terms of a role in filtering the blood and maintaining certain chemical balances.
Ok, I'll spot you this one, but next time, do yourself a favor and pay attention during class...
Functionalism has three distinct sources. First, Putnam and Fodor saw mental states in terms of an empirical computational theory of the mind. Second, Smart's "topic neutral" analyses led Armstrong and Lewis to a functionalist analysis of mental concepts. Third, Wittgenstein's idea of meaning as use led to a version of functionalism as a theory of meaning, further developed by Sellars and later Harman.
In the world of transistors, during manufacturing, we have functional and non-functional dies, where the non-functional are discarded, and the rest are further tested and assigned a 'functional level'.
This is where we end up with ram (or processors, etc.) being 'speed' rated, such as 80ns, 70ns, 60ns... These different speed components can all surface as part of a yield from the same 'batch', when some refuse to lock at one speed, and then pass inspection at another. Samsung doesn't run production of 60's on one day, and then 70's on the next. It's all about their 'functional' status as they come out of the oven...
A 'functional' transistor can go thru as many as 300 steps before it earns that title.
Good point. Not to mention the insane file sizes of the latest hi-res cameras, the rise of photojournalism has really made storage an issue, especially for those that shoot RAW. It's not uncommon for me to hear of PJ-style photogs shooting 2000+ images over the course of a 6-8hr wedding day.
Developments like this also impact the world of broadcast, video, and film production. Camcorders such as the Grass Valley Infinity series utilize CF memory for storage, and Panasonic has based it's newest camcorders on the P2 format (which is just multiple SD cards arranged in RAID-0, wrapped in a PC-card container). As increases such as these continue, it will further solidify solid-state memory in the video production world.
I've got a summary of the subject at FresHDV, along with common video transfer rates (permalink here)
Matthew Jeppsen
www.FresHDV.com
You're going to cheap out on the CF readers, when the CF cards themselves will cost you a good $500+?
You're talking about $2000 minimum just to get off the ground here. How does the poor man afford this?
I can't imagine how the write performance would do anything but stink. In theory enough flash in parallel should have decent write performance, but I doubt this setup will manage to extract it.
http://lkml.org/lkml/2005/8/20/95
imagine combining a flash-based harddisk with a big RAM buffer for speeding up reads and a separate one for writes, the second one protected by a small rechargable battery or super-capacitor. The disk only commits to flash if it feels it is running out of power. This should easily outlive the 4-year lifespan of an average corporate laptop.
10 ?"Hello World" life was simple then
You could wire them to your central nervous system if only your body could generate the proper electrical impulses.
WTF?
What are you talking about? Increased demand raises prices, decreased demand lowers prices. Likewise, weak demand in a market causes an increase in competition with its attendant reduction in prices. (Think about airfares after the 9/11 attacks: no demand = rock bottom prices.)
Prices are high due to a lack of excess supply.
A much better answer to the OP's question is that DRAM has reached an equilibrium point where the suppliers are able to charge a price low enough to discourage new entrants, beyond which they have no incentive to lower prices. Once (marginal cost = marginal revenue) you stop looking for increased volume through lower prices because your profit is already maximized. Lowering your price further just decreases your profits.
Make no mistake about it, if demand increased tomorrow for DRAM, prices would rise. If that demand stayed steady, the prices would stay high. They would come down if the demand collapsed because new entrants would be trying to recoup the fixed costs they had incurred to enter the market. That's not a function of a healthy market though, that's a function of poor demand forecasting.
Saying that prices are too high because there's no demand for the product (assuming the production is into efficient economies of scale, which DRAM production is) is like saying all of those people in Africa starved to death because they weren't hungry enough to eat.
The article, had you read it, clearly says GBytes. Maybe Samsung is making up a new industry term or something; but I'll assume they mean GigaByte.
put the what in the where?
Simply cache writes and parallelize them.
Say you write a 10Mb file. Have 8MB of cache on the device. The write goes to DRAM first. Then, data in RAM is written in parallel to multiple banks of flash memory (the disk layout can be interleaved on the actual chips).
This will speed writes greatly and also allow for medium failure (if the write fails, the sector can be written elsewhere since it's still in the RAM cache)
-Z
To clarify: Samsung has developed a process to create 16Gigabit(2GB) chips that when set in a 8(chips)X16(Gigabit) configuration yields a 16GB flash drive. The 16(chip)X16(Gigabit) configuration yields 32GB flash drives.
put the what in the where?
am i missing something? a 16GigaBit chip can hold 2GigaByte, and a 16x16 array contains 256 chips, so, this array should be able to hold 512GBytes.... i assume the article should say 4x4 array (32GBytes) (sorry about my english)
* Imagine a Beowulf cluster of those!
* In Soviet Russia, card flashes YOU!
* Sounds great, but does it run Linux?
* 1. Create huge flash memory card
2. ???
3. Profit
* I for one welcome our new 32GB overlords
* Rumor has it these cards will be shipped with Duke Nukem Forever.
* 32 GB of pr0n in your pocket
* ALL YOUR MEMORY ARE BELONG TO SAMSUNG!!
* CowboyNeal posted this two months ago.
* Free iPod with 32GB flash filled with naked and petrified Natalie Portman!
* Stephen king is dead/you must be new here/I have no memory you insensitive clod/no carrier/^H^H^H^H^H^H/BSD is dead/Google will do this better/Hot grit down my pants/etc.
There - no need to post any joke posts. And PLEASE don't reply with the ones i missed.
SIG: TAKE OFF EVERY 'CAPTAIN'!!
Unfortunately, other than my work with producing DVDs. 3CCD Camcorders are still out of my price range and an extra 'cost' feature that most of my clients can't afford...
I'm seriously contemplating getting one myself, for backup and sneakernet. Leaving one PC on all the time is much less appealing these days.
... but none with their fingers upon what I would consider the answer: A $199 digital video camera would have a pretty* large market. * warning: understatement
Suck my balls you donkey-raping shit eater.
Replacing disk with flash RAM is not feasible: flash isn't fast enough, and doesn't survive enough re-writes to the same blocks.
It's not only feasible, it's been done. It's horrifically expensive, but it works. A "wear leveling" algorithm is used to ensure the same flash cells aren't erased and re-written continuously. Heck, even the flash keychain drives and digital camera cards do that. No, it probably won't hold up to as many write cycles as a magnetic disk will, but writes are much less common than reads, especially in some database and web applications. The drive doesn't need to last forever anyway, since the computer it's part of won't either. I've heard that these guys have had one of their flash drives on a continuous rewrite cycle for a few years now - no errors yet.
Where do you get the notion that flash is slow? It's slow compared to RAM, but it's way faster than a hard disk. That's one of the selling points of these things.
Yeah ! and to power it put a bi metallic strip on it and stick in in your mouth to generate power. Course youd have to keep it in your mouth all the time which would be in keeping with some geeks I know.
"It's so convenient to have a system where everyone is a criminal" - A. Hitler
When will people learn? It's not gigabit, but gigabyte. GB, not Gb.
Did you RTFA? It clearly states "gigabit", which makes sense if a 16x16 array of them yields 16 GB, as also stated in TFA.
Downmodding is the refuge of the weak. Don't downmod, make a better argument!
Step 3: Learn to type with one hand plugged into the usb port. (Cue the one handed broswing jokes...)
Woah, that's a relief. I was afraid that I might be buying a device with billions of non-functional or even disfunctional transistors.
Actually, they can be hugely useful.
There was an old experiment I read about where they used a gentic algorithm to build an amplifier. I forget the exact numbers but it was something like "there are 27 essential components to an optimum simplicity design - we wanted to see how many it could get down to."
It promptly got down to something like 16. Despite the optimum simplicity designed by a human requiring 27 and there be "no way to simplify beyond that."
Except, it turns out that, sure: discrete elements, functioning entirely discretely, behave one way. Put in to a system though, they're not discrete. Be it the lag from one, the RF noise from another, whatever... they do affect each other.
Just because a human isn't smart enough to see a way they can interact and do something totally outside what's supposed to happen doesn't mean there isn't an even better way.
For me, I'll quite happily buy a chip with a huge number of faulty transistors if those faulty ones happen to do something radically useful that couldn't be done without spending many times more.
please explain why this is modded down? was it because he ask people to RTFA?
The chips themselves (minus packaging) are pretty small. They fit comfortably into a dinky SDIO card. What about stacking them into a 3.5" drive? Now we're talking about maybe 128 * 16Gb chips, 256GB, stuffed in 4 rows of 32 chips along the inside, along with busses and controllers/interfaces. If they did it right, we'd have a "RAIM", (Redundant Array of Independent Memory), offering "permanent" online storage as fast as the data being sent to it. Which would, in turn, create a demand for faster, even multiple, busses to the RAIM, a solution that finally found its problem.
Right now, 4GB FlashROM retail for $113, $28.25:GB. 320GB EIDEs retail for $111, $0.35:GB. Include EIDE drive packaging for a RAIM, and HDs cost 100x Flash. These new Samsung parts will cost thousands of dollars to the early adopters paying down some of Samsung's sunk investments, but will drop, sending the lower density chip price down even further. HD prices are dropping, too, but not as fast. And have a shorter lifecycle, as well as power/manufacturing/use costs much higher than the chips. How long before all those differential equations resolve to $Flash < $HD?
--
make install -not war
I think the point you're trying to make is that GP can't unstand that standard flash memory _cards_ allow for up to 16 individual _chips_. each individual chip is 16 Gbit, or 2 GB. and 16Gx16 is how a memory modual capacity is written and calculated, but c'mon this is slashdot do you really expect somone whos modded to +5 insightful to Understand Computers Enough To Know how to do basic multiplication, or that the RAM/solid state indusry has used the same convention since you no longer had to soldier the chips directly to the board?
https://www.gnu.org/philosophy/free-sw.html
besides the obvious advantage of having no moving parts does FLASH memory have a noticable speed advantage over SATA/PATA? I know that SATA is 150mbps max and that USB can take up to 480 however because I have only ever used flash memory to transfer files from a HDD to a Flash drive it is rather difficult to compare.
Basically I've been hoping for a larger flash memory for some time. I would like to eliminate the use of a standard HDD which tends to be a bottleneck on most computers for loading programs. Would the speed match USB2.0? or would the integrated cache of standard HDDs outweigh that speed increase?
Building bigger flash chips is not as exciting as building a flash chip as fast as volatile RAM.
That would be a real technology breakthrough!
Can you imagine a PC that doesn't need a real switch off and that can turn back on in a second or two?
Well, if you run that popular operating system you would still need complete reboots for a number of tasks. But that would be another story!
Maybe Computers will never be as intelligent as Humans.
For sure they won't ever become so stupid. [VR-1988]
hmm, imagine a beowulf cluster of 9/11s!
In Soviet Russia, New Orleans floods hurricane Katrina!
Wrong thread, asshole. This one's about the fancy new 16Gb FLASH memory.
...if combined in 8x16 and 16x16 configurations - theoretically enable Flash memory cards with capacities of 16 and 32 GByte.
Also, a "16x16 configuration" is 256 chips, which gives 512GB, not 32BG.
16 chips x 16 Gb/chip == 32GB
640 thousand functional transistors ought to be enough for anybody.
Science : Proprietary , Knowledge : Open Source
Actually, it wasn't even necessary to RTFA to see that, it was in the summary too. Some people need to get their vision checked.
All your base are belong to Wii.
In TFA, they don't say what the seek time of this device is. I mean, if I got one of these 2 Gb things (16 giga BITS, right?), and put it in my compu with one of these and boot off of it, will that be faster than the fastest possible magnetic hard drive (those 30000 rpm Seagate SCSI's)?
l'Homme n'est Rien l'Oeuvre Tout: Gustave Flaubert to George Sand
Is anyone here aware that Samsung is a Korean company?
Take the DV format for example, DV tape holds roughly 5 minutes of NTSC video per gig if you look at it from a storage point of view. A 60 minute DV tape =~ 12 gigs.
Moving to a higher chip density will allow these cameras and the new generation of HD pro-sumer cameras to record directly to "disk", saving huge amounts of production and post-production time in terms of media management.
The Panasonic AG-HVX200 http://www.dvxuser.com/articles/HVX200/ records directly to their "P2" cards which are currently 2GB per card. Recording in HD100 mode eats up 1GB per minute. Only has slots for 4 P2 cards. Move up to 8GB per card or 16GB per card drastically expands on-camera capacity. This is a HUGE market.
Filmo The Klown
Where do you get the notion that flash is slow?
Right here:
Timing cached reads: 1584 MB in 2.00 seconds = 791.72 MB/sec
Timing buffered disk reads: 20 MB in 3.20 seconds = 6.26 MB/sec
Timing cached reads: 1568 MB in 2.00 seconds = 784.12 MB/sec
Timing buffered disk reads: 118 MB in 3.03 seconds = 38.92 MB/sec
Step 2: put fingernailsized flash memorychip on place of fingernail
Now just a way to power them up and use them. Any ideas?
Hmmmmmmm...
With this kind of idea I do not expect my central nervous system to generate anything important anymore, lets stand enough energy (-:
My wife's sketchblog Blob[p]: Gastrono-me
I am struck by the utter lack of facts in the second article. It is highly emotionally charged and relies on self-evident truths that are less than self-evident.
.Without the improvements - a flood gate in the Harvey Canal and raised levees along the Intercostals Waterway - a tidal surge produced by a hurricane 'could result in the catastrophic loss of life and property damage,' corps officials reported." Not that the Bush administration stepped in to do more since, but given the time these projects take - one item of ACOE funding that the Bush Administration cut was for a study about upgrading the levees, and the study wouldn't have been completed until, I believe, 2008. (As a side note, the Internet also lets us discover that the New York Times, now on its Olympian high horse condemning the irresponsibility of the Bush administration's funding cuts for the ACOE, in fact, last April bitterly condemned the very same legislation as an expensive boondoggle. The paper thundered that "the bill would shovel $17 billion at the Army Corps of Engineers for flood control and other water-related projects - this at a time when President Bush is asking for major cuts in Medicaid and other important domestic programs." Ah well, no one ever expects to hold them accountable for anything.)
Pretenders such as these cannot extricate us from a debilitating war, nor can they rebuild the nation they destroyed; they have no idea how to allocate resources against terrorism, nor how to prepare for the disasters that will surely come. What the Republicans in power can do is set up photo ops, repeat spin points, concoct hollow slogans about "compassionate conservatism," and sidestep responsibility by whining about "the blame game."
Is there any evidence to support this? One would suppose so, but is left wanting after reading this article.
In his pathetic insufficiency, Brown evidently was not alone at FEMA. The deputies and acting deputies and various other high-ranking pork-choppers -- many of whom had landed at the agency from positions with the Bush-Cheney campaign -- showed up with no experience in the hard work of saving lives and restoring communities.
This is an adhominem attack, again with no supporting facts. It would be nice to see a few examples of these "pork-choppers" in action. It would be nice to get some factual evidence, but there is no such requirement for the type of arguments that frequently appear on the Salon. Compare these blurbs to a similar one from the original post:
It is also perhaps inconvenient to point out that major flood control projects that could have been completed by now were in fact cut-by the Clinton administration. From the New Orleans Times-Picayune in 1995 (isn't the Internet fun?): "The Clinton administration is holding back a Corps of Engineers report recommending that the $120 million project proceed.
Note how different the tone of this blurb--and indeed the entire article--is from the second article. It may lack the energy and shrillness of the second article, but it seems to go much further in actually proving a point. The second article is the typical chest beating, blame Bush attitude that has paralyzed the "liberals" and the democratic party ever since Clinton left office (and arguably before he came to office as well). This sort of lack of focus on facts or reality is exactly what has driven democrats from the white house, the house of representatives, the senate, the state governerships, and now from the courts.
This isn't a serious way to get any indication of the performance of a flash-based HD replacement.
The SD equipment you're testing is very different from a solid-state disk system, even if the basic flash technology that actually keeps the bits is the same. The SD interface was designed to be tiny, not to be blazing fast. And the SD card probably contains only a single flash part, whereas a solid-state disk unit is likely able to read from or write to many flash parts in parallel. It wouldn't be fair to condemn all magnetic systems as slow based on the performance of a Microdrive unit, and it isn't fair to condemn all flash systems as slow based on the performance of this SD card.
Also important is that you're testing cached reads and buffered writes, which might be useful system metrics but tell you nothing about the mass-storage hardware in isolation. The whole purpose of caching/buffering is to hide the timing characteristics of the underlying storage. If the system software and device drivers supply a large system RAM buffer for HD writes but not for USB mass storage devices, then one would expect slower USB writing, even if both interfaces were actually ultimately connected to the same mass storage hardware.
In many real-world applications, it's seeks, and not the reads or writes themselves, that consume the most time. Flash obviously has it all over mechanical systems when it comes to seek times, and flash-based HD replacements are generally specified very competitively with real HDs when it comes to sustained transfer speeds. This one claims 44MB/s sustained writing - considerably more than whatever your buffer-obscured HD is actually capable of, I'll wager.
Well for some reason I thought you were testing cached reads and buffered writes, not cached reads and cached writes. My goof. In any case, my objections to this "test" stand.
Okay, I'll grant that the SD card interface may be a poor one, but the same holds true for every other solid-state interface I've used--CF, SmartMedia, Sony's little Memory Stick thing, and USB2 (High-speed) flash devices. (FWIW, I recall similarly poor results when I was doing R&D with a solid-state disk on 2.5" IDE, but that was a few years ago so things may well have changed in the meantime, and speed wasn't important for the particular application so I never took down numbers.) I am of course aware of the seek issue, and I'll agree that for a server that accesses lots of different files randomly seek time can become a factor; when I wrote my reply, dumps of large, relatively contiguous data streams (e.g. MP3 files) were on my mind, and even a 20ms seek time is only a small fraction of the time it takes to actually get the data off the device, so sustained transfer rates really do make a difference. But I suppose I'll keep an eye out for how they progress.
Incidentally, the speed tests in my original reply were done with hdparm, which tries hard to keep system overhead from affecting the measured transfer rate. As it turns out, I was misrunning it, since it seems to support O_DIRECT now (previous versions used the cached read timings to adjust for buffering), but since the data read from the disk wasn't in memory in the first place anyway, buffering can't have sped it up--the data has to come off the disk one way or another. For the record, I just re-ran the tests with --direct, and got the same results +/- 1%.
This one claims 44MB/s sustained writing - considerably more than whatever your buffer-obscured HD is actually capable of, I'll wager.
Let's see:
# hdparm --direct -t /dev/hda (ATA100 HD)
/dev/hda:
Timing O_DIRECT disk reads: 134 MB in 3.04 seconds = 44.14 MB/sec
Whaddya know?
Whoops, I totally cocked up & didn't realise until it was too late. 'e' is the natural logarithm, exponential is '^', i.e. 3^2 = 9.
Yar.