Slashdot Mirror
Samsung Demos Future Memory Chips
Fletcher points to this story in CNET Asia, excerpting "The Korean electronics giant unveiled an 8-gigabit flash memory chip Monday based on the 60-nanometer process, as well as a 2-gigabit DDR DRAM chip based on the 80-nanometer process. Flash chips, which retain data after a host computer is turned off, are used in flash cards and cell phones, while DDR DRAM is used inside PCs."
Why aren't they using conventional storage standards, RAM, and disk space are all in megabytes (1024 vs 1000 debating aside) saying something is *bit (giga,mega,kilo) implies a rate connectivity doesn't it?
Error 407 - No creative sig found
People tend to get excited about new products like these; in a separate but equally relevant phenomenon, they tend not to RTFA.
From the article:
Both chips, however, are prototypes. Companies just began this year to make chips on the 90-nanometer process. (The nanometer measurement refers to average feature sizes on the chips). Eighty-nanometer chips may not come for at least another year, and 65-nanometer chips won't debut until at least the end of 2005.
In other words, 16GB flash MP3 players will not be available in time for Xmas.
I want to drag this out as long as possible. Bring me my protractor.
Comment removed based on user account deletion
"Flash chips, which retain data after a host computer is turned off, are used in flash cards and cell phones, while DDR DRAM is used inside PCs."
This being Slashdot and all, one wouldn't think that needed to be said. =)
That 16GBs of memory translates into storage of up to 16 hours of DVD-quality video or 4,000 MP3 audio files (at 5 minutes per song).
Can someone explain to me how 1GB/hr equates to DVD quality? Most DVD films I know of run at 2-4GB/hr...
Sure, low-bitrate DVD is 1GB/hr or less, but is that true "DVD Quality?"
How soon to get 8 gigabytes, so we can put the original DVD? Probably 3 years.
No I have not RTFA, but if the flash ram retains it's data when the PC is off, couldnt we use it as a hard drive substitute rather than a RAM substitute?
That would be pretty cool... Press button on. WHIZZ... Logon screen is there! Nice.. :)
Friends don't let Friends use Internet Explorer.
Few things - first of all the ram chip is GigaBITs. 8 bits = one byte, so that ram chip is actually 0.25 GB of space.
As well this is a ram chip, NOT a ram Stick. So you can have a number of this ram chip on one stick to make different sizes of ram sticks.
>I'm just wondering when we're going to start putting in ram for the
>sake of having more ram. Won't more ram eventually become unnecessary
>with all the bottlenecks computers have?
man top
man iostat
man vmstat
man netstat
8 gigabits = 1 gigabyte
2 gigabits = 256 megabytes
And this was quoted from the article, which isn't talking about speed, which would be gigabits-per-second (sometimes abbreviated gigabits), this is size, as in (quote) Both chips hold far more data than current chips in their respective markets and are smaller, which should make them cheaper and more powerful than existing chips.
Smaller, mabey. Higher capacity? No.
Technological neophyte journalists.
~Wx
sig?
RTFA...they're talking about using the chips to fit 16 gigabytes on one DIMM.
tasks(723) drafts(105) languages(484) examples(29106)
Comment removed based on user account deletion
Many people probably have more RAM than they need. However, certain operating systems tend to expand their needs to meet or exceed typical RAM configurations. Additionally, many of us in research really benefit from increased RAM, although you do need a 64-bit architecture before you can access more than 4 Gibibytes of it. For example, if you're doing large simulations, you benefit greatly from being able to keep everything in memory and not having to read/write to the hard drive.
Ben Hocking
Need a professional organizer?
Actually, increasing the RAM reduces the significance of other bottlenecks in a PC.
For example, you can buffer transfers that would otherwise go to or from the hard drive, so you spend less time waiting on I/O.
tasks(723) drafts(105) languages(484) examples(29106)
enough D's already!!
Shit, 640k should be enough for anybody!
I hope the land around you yields, a crop like all the other fields, and then your waiting might make sense...
He didn't mention 16 gigabytes anywhere. He said "2GB is a lot on one stick of ram."
One can already buy Compact Flash to ATA adapters to use CF cards as hard drives. There are two primary drawbacks to this approach at the moment:
CF cards tend to 'wear out' after a certain number of write cycles. Most estimates of lifespan range from 100K to 500K write cycles. (Working from memory - could be off a bit.)
CF cards cost more per MB than traditional magnetic/rotating media drives. The cost means that the largest currently available cards may not be practical for most applications. However, CF-to-ATA adapters are nifty options for single purpose systems and homebrew firewalls.
PS Go RTFA. Actually go read the Compact Flash FAQ and then go RTFA.
I want to drag this out as long as possible. Bring me my protractor.
I notice that DRAM prices, for the same technology, have stayed at their 2001 price level at $100 to $150 per gigabyte. During the same period flash memory has fallen from $300 per gigabyte to $80. I like to look for "odometer threshholds" when prices drop the next factor of ten (about every every five years). For example, hard disk fell below $1 / GB in 2003 and flash $100 / GB in 2004.
I did read recently there was some price fixing in the DRAM market.
I did RTFA - the parent didn't. Thats what I was trying to say in my post, you could fit MORE than 2 GB on the RAM stick since its not a 2 Gb RAM stick, its a 2 Gb ram CHIP.
Comment removed based on user account deletion
How about a system that runs the OS off a flash chip, but not just like a handheld. Embed that little flash chip on a Mother Board for the OS and use that to boot the system and keep the system state even after power off.
Mmmm... instant on computers maybe?
"Don't sweat the petty stuff and don't pet the sweaty stuff." -- by an Unknown Wise man.
Perhaps, but not yet by a long way.
For instance, I presently want to find linear dependencies in a sparse bit matrix which has something over 8.1 million rows and columns. Even with a sparse representation, the computation requires over 2 gigabytes of active memory. In my field of interest (integer factorization) having access to machines with several gigabytes is almost an entry requirement. There are a number of other fields which need much more RAM than this if the programs aren't to be paging themselves into torpidity. Numerical simulation and large database access are a couple of examples.
Paul
Lasciate ogne speranza, voi ch'intrate
As others have mentioned, memory parts are alomst always described in terms of bits, and always have been. I believe there are a few reasons for this.
Not all computers use (or used) eight bit bytes. The PDP-10 (?) and a few IBM machines used nine bit bytes organized as 36 byte words. I also think there were a few machines with sub-eight bit bytes, but I can't think of any right now.
I seem to recall memory some memory parts being available with data bus widths other than eight. Caffine hasn't kicked in, though, or I would be more sure about this.
Also, because of error detecting and/or correcting memories (eg, parity, ECC, etc), the physical memory width may we wider that the logical width.
(S(SKK)(SKK))(S(SKK)(SKK))
Won't more ram eventually become unnecessary with all the bottlenecks computers have?
No... err... rather, something will fundamentally change.
Instead of having a hierarchy of memory (hard drive, ram, cache, etc), you'll see RAM and flash merge into a "universal memory". Everything will come on a single chip - processor and storage. RAM won't be required since the on board storage will be both quick and nonvolatile.
Currently, as much as 75 percent of a processor's area is used for cache memory. This is a number that is increasing, too. This is because RAM is too electrically "distant" from the main processor to be of any high-performance use. The near-term solution has been to pile on lots of cache memory in order to make up for it.
Recently, Ovonyx licensed their phase-change technology to Nanochip. Now, the phase-change technology is the same thing that is currently used in CD/DVD-RWs. With this implementation, they'll be programming and reading the material electronically instead of optically. Since they'll be doing it with MEMS and atomic probes, the density will reach levels of 1 terabit/square inch (125 gigabytes) and will do so very quickly. For more information, see HP's probe storage page. As a side note, HP and Nanochip are just a couple miles apart so it is rumored that Nanochip is hiding the HP plan at this point. Commercialization in 2006 isn't too far off. Also note that Microsoft is an investor in Nanochip as well. Bill Gates mentioned at Cebit that terabit chips will be here "very soon". Something to think about.
Life is the leading cause of death in America.
(groan) Sorry...replied to the wrong comment. :(
tasks(723) drafts(105) languages(484) examples(29106)
80 nm? 60 nm? Surely they mean 65nm and 90nm...
8 Gigabyte
Spot on.
I think it'd be a real benefit for massive amounts of RAM to be commonplace, even for the home user. It'd free up system designers to do things a little differently.
(please don't flame me for using the following as an example -- it's simply one system with which I'm familiar, and works in a way that would benefit from 'excessive' RAM)
OS X's document-centric approach to applications means that you rarely need to close programs. The only on-screen overhead a running app has over a closed app is a small black triangle below/beside the app's Dock icon. After working like this for a while, you forget what 'application startup-time' is. Apps become just another widget to click - a service of the system rather than a mental context-switch (if you catch my drift).
So, with oodles of RAM, your common apps and data are always a nanosecond or two away.
(incidentally, this is why my puny 500MHz G3 iMac is still usable. It's stacked up with RAM to the point that my apps rarely get closed and are available with only a smige of lag. Certainly not as quick as new machines, but with a perceived speed that belies the machine's actual power).
What's the frequency, Kenneth?
Can someone explain why the FLASH memory is sooooo much larger than the DRAM chip? DRAM is just one cap and one transistor per bit, while the flash uses a MUCH more complicated structure for each bit. It involves at least two transistors per bit, one with a floating gate.
For the same size die, I would expect that the DRAM would hold a little more than the FLASH. Either the FLASH die is huge compared to the DRAM die in this case, or I am missing something.
Can anybody clue me in?
"-1 Troll" is the apparently the same as "-1 I disagree with you."
Because your CF card has more than one chip inside?
)9TSS
....in a more/bigger/faster and more throw away culture. I'm impressed! But it's also depressing! How is that? Well, tell ya, this is like the horsepower wars out of Detroit in the 60's, more cubes and higher revs with higher compression. Ok as far as it went, but........it meant throw away cars, too.
Tell ya whut I would be more impressed with technologically, if some RAM company wants to make a splash and show off some real branez. A smart and adaptive memory chip reader that you could stick in a ram slot like a daughtercard that you could then insert any mix or match multiple RAM sticks into and it would read and access and use them all.
We are awash in so called "obsolete" RAM that is still functional. It used to be just a coupla decades ago we threw away stuff when it was broken. Now we throw away perfectly fine stuff, things that aren't broken, they are just "obsolete" although they might only be a few years old.
Anyone see anything potentially wrong there? Same thing with CPUs. We have SMP mobos (and kernels), how about NON-SMP MO mobos, any braniacs got any examples of that, were you can mix and match older processors and keep using them? I know you can make a cluster whatsis with older boxes, I am talking a single machine that you could add tons of older oddball ram sticks to and plug in a variety of CPUs.
To me, RAM and CPUs should be treated like drives and other peripherals, you should be able to daisy chain them better (different kinds, sizes, functions, etc) on a single machine.
Last I checked, most memory technologies required at least 1-T per bit. I don't know if that's true for flash technology, but still. 8Gb would be 8+ billion transistors not including decode logic and amplifiers. Wouldn't this make these flash chips have the highest transistor counts on the planet?
Last I checked, the highest transistor counts we had were around 400-500 million. That's like 1/16 of what would be needed to do this. What am I missing?
Comment removed based on user account deletion
I'm beginning to think I'd like to see machines where we have dual storage setups - use the hard disk for write-many-read-many data (general data files), and use the flash memory for write-limited-read-many data. For example, imagine installing your OS and programs to flash memory - booting times and program loading times would be nearly instantaneous. (/me drools) You'd just have the flash memory mounted like any other storage device, and maybe some "wlrm" flag available to applications so they could automatically prefer that storage for installations and steer away from it for write-many data files.
This is what personal devices like ipods, etc... usually do. The operating system is stored in flash ram (so it can be upgraded in case something goes wrong), and data (i.e. your music) is stored on the hard drive.
QuakeIII is for... is for... oh who am I kidding, I need this like Bush needs credibility.
"A witty saying proves nothing." ~Voltaire
"d'Oh!" ~Homer
Memory gets used up in all OS. It allows to to cooler things faster. It allows us to worry about doing cool interfaces instead of writing cool small code. It allows us to write protable API with memory and CPU wasting layers.
One can see the importance of memory with the way that some of the cheaper MOBO become useless. Many years ago I bought two machines with several months of each other. The first had a hard 384MB limit. The second has a 1GB limit. Both were towers. The first is now useless. The second is still in use. Go figure.
"She's a scientist and a lesbian. She's not going to let it slide." Orphan Black
A Mac waking from sleep takes around 4 seconds to be fully functional (waiting for network etc) - is that not quick enough for you?
I've had problems with hibernation on PCs, but never had a problem after waking from sleep on my iBook.
What ever happened to rom memory? I thought that rom memory was easier to make than ram and was a generation ahead of it. So they should be making 8 giga bit rom chip than. Someone should be able to make a 8 giga byte rom chip than. With that much storage they should be able to put a full enclycopedia, dictionary and thesaurus and a operating system able to access both the internet and one's drives all on a rom chip. All programs than should be able to access these and I should be able to spell check this without any problems. I should be able to look up any word in both the encylopedia and dictionary without leaving this program and have the result instantly. I hope that with voice recognition I could do it without typing.
True, a RAM cell is just one capacitor and one transistor per bit. However, the capacitor needs a much larger area than the other components. Its capacitance must be much larger than any stray capacitances in the circuit, to hold the charge that represents the bit without any errors for the length of one refresh cycle.
You forget one thing: ROM is easier to do than RAM, BUT you need to know what to store on it during the creation process. you cant just store "black roms" and write them just in time.
So EVERY update, every revision, every EVERYTHING would need a change in the produciton cycle, shipping times, ect)
Its just not very practica, even more considering that 8Gb is just one GB, which is about 50c worth of hd space. Even the socket for the rom chip is more expensive...
HI O WISE PRINCE. WHT TOOK U SO DAM LONG?
If you want to learn more about future memory technologies which could replace today's SDRAM and flash-memory try searching for "MRAM", "FeRAM" and "PCRAM". The latter is probably the most promising one due to the (expected) ease of design and manufacturing ("Samsung says that it will enter full-scale production with the technology in 2006").
More information about phase-change (->PCRAM) materials can be found at E*PCOS website.
we can get a meg^H^H^HGiga^H^H^H^HTerraByte of ram!
640 PetaBytes should be enough for anyone....well everyone, really
The Kruger Dunning explains most post on
This isn't always strictly true.. For eg, in Linux, after 896 megs, you need to enable large memory support, which slows things down.. Personally, I've got 1 gig, and I never reach past 500megs usage..
..
Even with 10 gigs of ram, you'll still need stuff like swap anyway, for reasons described on http://kerneltrap.org/node/view/3202
Also, think about how you can access memory in a linear way with 4gigs of ram (which is a very large amount of memory pages), some of the operating systems internal data structures would be getting pretty hefty. Fortunately, one benefit from this will be that computer products will require more optimised access methods if the product is to succeed..
I'm so glad to get useful computer tips like this. This is why I love this forum. You never know when some obscure technology will suddenly become crystal clear by means of a helpful, knowledgeable story poster. My spelling has improved because of Slashdot, too. Thanks, guys! You're real swell.
Sincerely,
A grateful IT professional.
Any sufficiently simple magic can be passed off as mere advanced technology.
I got one of those 1GB SD chips for $110 a couple of weeks ago, and now they're $94. That's a 10% drop in 10 days, and they're still dropping. And there's now a miniSD form, so 1GB the size of my fingernail, or 10GB the size of a stamp, is looking good. Samsung is driving down these prices. I'd love someone to get the power requirements down to the point where something like RFID could let me read them wherever they were stored in my apartment. I'd keep the data encrypted, and sticking them into every product I buy is another story. But these little memory chips are a miracle. If they did sew them into my garments with sales info, I'd erase them with my own device and wear my storage around, with encrypted backups of all my media literally at my elbow, wherever I go.
--
make install -not war
In communications, there are more protocol issues to worry about. For modems, there's baud--the number of changes in sound per second (handwaving), the bitrate--each change in sound can represent more than one bit, compression--for example, e-mail and HTML don't often need more than 7 bits to represent its highly non-random content and can be compressed, bits for error-detection/correction--because there is always line noise, etc. so any given byte in a communication may take 6-12 bits (handwaving) to transmit.
s ysadm-326.html and http://www.protocols.com/pbook/tcpip2.htm
Ethernet, token ring, and almost all other networks similarly add overhead to data in the form of headers describing the contents of packets which encapsulate user or application data.
Plus, your data is probably encapsulated in some kind of protocol (SLIP, PPP, IP, TCP, UDP, etc...) which add their own headers and other structure to your data. Individual applications usually need their own headers (content-type/HTTP negotiation for your web browser, SSL, etc), adding more overhead.
And of course, at each part of the communication stack, headers can be added or removed, packets repackaged and possibly reordered, etc. depending on what you are connecting to (the TCP/IP stack for an uplink is very special compared to the one in your Linksys). A poorly written TCP/IP stack in the OS does worse than a good one on the same network and network card (let's ignore PCI bus latency and software modems in the client machine).
In short, pulsing a piece of wire or a fiber x times a second can have very little to the actual data transfer rate as experienced by the user, hence data-link bitrate doesn't translate into experienced bitrate in a particularly quantitative sense. On a wired connection, it is generally safe (handwaving) to assume that it takes 10-15 bits of physical layer to transmit a byte of user data for many applications.
See: http://wks.uts.ohio-state.edu/sysadm_course/html/
There are 1.1... kinds of people.
"To me, RAM and CPUs should be treated like drives and other peripherals, you should be able to daisy chain them better (different kinds, sizes, functions, etc) on a single machine."
--ya, I'd like to have a mobo that had different sockets on it that fit different processors that are already out there by the billions. If there was a way to assign apps as you opened them to the appropriate processor, it would be nifty. Some apps don't need much at all and will run spiffy, others can take every cycle you can throw at them. I know they don't make anything like that, but I'd still like to see it somehow, maybe even a universal generic board that the entire CPU and socket was removable, like PCI slots, just plug and play different ones in, along with a similar deal with the RAM.
I know not likely to happen, just would be neat, to me anyway.
NAND is very much like disk in that you read and write it in pages. NOR is random access. NAND writes/erases faster than NOR which makes it useful for file storage. NOR is most used for things like BIOS etc. Most Linux mobile devices use either JFFS2 or YAFFS file systems (goole will find) with flash because they are tuned for flash and are far more robust against power failure corruption etc than conventional file systems.
NAND flash isn't as fast as rotating media for desktop applications (and can't do swap very well) so does not necessarily improve boot time.
Engineering is the art of compromise.
Nice post, thanks.