from the computers-without-Alzheimers-department dept.
ULTRARAM™ is a novel type of memory with extraordinary properties. It combines the non-volatility of a data storage memory, like flash, with the speed, energy-efficiency and endurance of a working memory, like DRAM. To do this it utilises the unique properties of compound semiconductors, commonly used in photonic devices such as LEDS, laser diodes and infrared detectors, but not in digital electronics, which is the preserve of silicon.
[...] Now, in a collaboration between the Physics and Engineering Departments at Lancaster University and the Department of Physics at Warwick, ULTRARAM™ has been implemented on silicon wafers for the very first time.
Professor Manus Hayne of the Department of Physics at Lancaster, who leads the work said, "ULTRARAM™ on silicon is a huge advance for our research, overcoming very significant materials challenges of large crystalline lattice mismatch, the change from elemental to compound semiconductor and differences in thermal contraction."
[...] Remarkably, the ULTRARAM™ on silicon devices actually outperform previous incarnations of the technology on GaAs compound semiconductor wafers, demonstrating (extrapolated) data storage times of at least 1000 years, fast switching speed (for device size) and program-erase cycling endurance of at least 10 million, which is one hundred to one thousand times better than flash.
So... are we approaching the point where we get a plug-in RAM storage module that can be used like nonvolatile RAM -- because it is nonvolatile? And when you've built complex data structures on it with RAM efficiency, you can unplug it and put it, and of course the data, on a shelf for later use?
Or just plug it into a computer when you need an extra 24 gigabytes of RAM to formally verify a category-theoretical theorem?
How would *you* like to use this?
Journal Reference:
Peter D. Hodgson, Dominic Lane, Peter J. Carrington, et al. ULTRARAM: A Low‐Energy, High‐Endurance, Compound‐Semiconductor Memory on Silicon [open], Advanced Electronic Materials (DOI: 10.1002/aelm.202101103)
(Score: 4, Insightful) by Kell on Monday January 10 2022, @11:21PM (4 children)
If it's non-volatile, then it is effectively media. Like all media, its fate is to be used for porn.
Scientists ask questions. Engineers solve problems.
(Score: 0) by Anonymous Coward on Tuesday January 11 2022, @12:31AM (3 children)
A big part of thrill is in the hunt for new stuff. Who wants to keep looking at the same old stuff? Might as well be married. j/k
(Score: 0) by Anonymous Coward on Tuesday January 11 2022, @01:08AM (2 children)
Have you ever wondered why you can eat the same meal every day and enjoy it, but porn goes stale after about 3 times?
(Score: 2, Interesting) by Anonymous Coward on Tuesday January 11 2022, @01:46AM (1 child)
You can eat bland food every single day and not get tired of it. Consider bread and rice, for example, staples that traditionally were part of nearly every mealHowever, very spicy dishes will be extremely appetizing but you cannot eat them every day, lest you grow tired of them. It is because they make a very strong impression while you're enjoying them. Porn is like that. It is a spicy consumable and therefore you cannot enjoy the same exact stimulus every day. You become burned out on it.
This is why you can't whack it to the same picture of Nikki Dial for the rest of your life.
(Score: 2, Funny) by Anonymous Coward on Tuesday January 11 2022, @02:07AM
UltraRAM sounds like the perfect name for an "adult" toy! Porn it is, then!
(Score: 0) by Anonymous Coward on Monday January 10 2022, @11:30PM
Bubble Memory!
Then in early 90's... IBM was playing with RAM with battery backup (24 to 48 hrs). I have that multiple machines in the office, today. Helps revent data lost and curuption on Raid arrays. In-flight changes are in that memory and I can pull it for the card and replace the card and keep right on going.
(Score: 3, Insightful) by takyon on Monday January 10 2022, @11:36PM (4 children)
March 29, 2021: https://www.lancaster.ac.uk/news/-first-steps-towards-revolutionary-ultraram-memory-chips [lancaster.ac.uk]
If that's true, then the density and cost-per-bit should be more like NAND than DRAM. I didn't see that mentioned in the newer article. Merely having the non-volatility of flash is not that interesting.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by RS3 on Tuesday January 11 2022, @12:56AM (3 children)
Maybe it will have much longer life than FLASH, which "wears out".
(Score: 2) by takyon on Tuesday January 11 2022, @01:31AM (2 children)
You've already got that with DRAM.
So far I'm seeing a non-volatile DRAM, not the universal memory we crave. Guess I'll have to click the actual science paper.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by RS3 on Tuesday January 11 2022, @02:09AM
No, no! You know the rules, never do that! :)
(Score: 1) by fustakrakich on Tuesday January 11 2022, @04:24AM
1TB L1 cache
La politica e i criminali sono la stessa cosa..
(Score: 1, Funny) by Anonymous Coward on Tuesday January 11 2022, @12:18AM
I hope I can put some in my 2006 Thinkpad!
(Score: 2) by Runaway1956 on Tuesday January 11 2022, @12:21AM
Not just quantum computing, but these guys are going to put two wells of quantums at each gate. Awesome!! With three wells, we could have EXTREMERAM!
“I have become friends with many school shooters” - Tampon Tim Walz
(Score: 0) by Anonymous Coward on Tuesday January 11 2022, @12:46AM (4 children)
I'll need about 2Tb of this stuff if they want me to stop using an SSD.
(Score: 2) by Runaway1956 on Tuesday January 11 2022, @01:19AM
Put it into NVME format, and put a 2 TB stick right beside your 4 TB NVME SSD. Something like your laptop can be closed up to "sleep", drive down the road a couple hundred miles, and open the laptop for use again. Virtually no battery drain along the way, but no memory loss either. Phones and tablets may have to wait a little longer, and/or settle for less memory in a smaller format.
“I have become friends with many school shooters” - Tampon Tim Walz
(Score: 2) by takyon on Tuesday January 11 2022, @01:54AM (2 children)
A lot of the cheap laptops are still shipping with as low as 32-128 GB of storage, with 16 GB being phased out. That's probably the same for phones. If they can put at least 32 GB in a small, cheap package, it's a start.
You could settle for as low as 256 GB in your desktop PC if it's cheaper than the same amount of DRAM with no significant impact on performance. Then just use a separate 8 TB SSD. Maybe penta-level-cell?
It all depends on the price-per-bit. There's a reason we haven't all switched to 3D XPoint/Optane.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by Freeman on Tuesday January 11 2022, @03:00PM
Anyone doing much into gaming, music, videos, etc. will not be happy with 256GB in the long run. My bare minimum for a new computer for the last decade, if not two decades has been 1TB of storage. I really should up that to 2TB nowadays, but 1TB is a good place to start. With 1TB NVMe boot drive and 1TB NVMe or standard SSD storage drive, one can get by easily enough. In the event that you really need more than that, maybe you should be looking into Network Attached Storage with RAID.
Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
(Score: 2) by hendrikboom on Wednesday January 12 2022, @02:51AM
I've seen phones for sale with 128GiB of storage (not RAM). I hesitate to buy them, because experience tells me they'll be packed with about 124GiB of preloaded software that I can't remove.
(Score: 0) by Anonymous Coward on Tuesday January 11 2022, @01:46AM
Announcements made to bamboozle investors do not translate into physical artifacts. No use discussing the latter when observing the former.
(Score: 2, Funny) by Anonymous Coward on Tuesday January 11 2022, @02:03AM
UltraRAM is used by the Science Patrol [fandom.com].
(Score: 2, Insightful) by Anonymous Coward on Tuesday January 11 2022, @02:17AM (2 children)
No. Programmers count on volatile memory being volatile. That's why you have to reboot all the time. Also, so that private data can be forgotten.
Anyway, ten million write cycles isn't anywhere close to enough to replace ordinary RAM. Its endurance is closer, in terms of orders of magnitude, to being a CD-R than to being RAM.
But if the cost works out, it could be a nice replacement for flash in SSDs. Or maybe it could be a cache on the SSD so that frequently written data hits the fast, high endurance memory and the slow, low endurance flash gets used for less frequently updated data.
An average PC writes about 10GB a day, but it's mostly the same 10GB getting rewritten over and over. If you leave Chrome open 24/7, it will write about 5GB per day just by itself, if not more, almost all of it state checkpoints. Current SSDs deal with this by wear leveling, but a dedicated high endurance area would do even better. Of course, they employ that strategy too, using part of the flash for high capacity QLC data storage and some for high endurance SLC cache. So perhaps the gains from this are not going to be that much.
Never underestimate the capability of refinements of existing technology in contrast to the next big thing.
(Score: 2) by dmbasso on Tuesday January 11 2022, @01:28PM
FTFY.
`echo $[0x853204FA81]|tr 0-9 ionbsdeaml`@gmail.com
(Score: 0) by Anonymous Coward on Tuesday January 11 2022, @06:32PM
And that's just the Windows telemetry data!
(Score: 3, Informative) by ChrisMaple on Tuesday January 11 2022, @03:12AM (2 children)
The write time is slow, 10 ms. Life is only 10 million cycles, also unacceptable for system RAM.
(Score: 0) by Anonymous Coward on Tuesday January 11 2022, @02:14PM (1 child)
I didn't find the write speed in the article. Where are you seeing 10ms? That's not fast at all. That's mechanical disk speed.
(Score: 0) by Anonymous Coward on Tuesday January 11 2022, @02:20PM
Ah, I see, it's in the paper, not the article.
The claim is that since they are working in a 20 micrometer process (that's micro, not nano) that it's faster than flash would be at that process. That's a ways away from it being faster than current mass produced flash, which generally has write times of less than 0.5ms (sometimes much less).
(Score: 0) by Anonymous Coward on Tuesday January 11 2022, @02:37PM (1 child)
It's hard to keep track of all these different RAM technologies. Is there any good resource with all the relevant information tabled and comparable?
(Score: 2) by Freeman on Tuesday January 11 2022, @03:17PM
You don't need to, they are mostly for specific use cases, or just to stupidly expensive.
Use regular RAM. In this case that means DDR4 or DDR5 RAM from a reputable reseller, like https://www.newegg.com/ [newegg.com] . In the event that you're having trouble beyond that use a site like https://pcpartpicker.com/ [pcpartpicker.com] for guidance.
Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
(Score: 2) by DannyB on Tuesday January 11 2022, @03:15PM (1 child)
I've wondered for a couple decades how OS design might change when something like this eventually becomes real.
The fact that we have separate volatile working memory and file storage is simply an artifact of the realities of past hardware. Because of that, the whole idea of the files (storage) and memory being separate is very deeply ingrained into our thinking.
What if the processor could execute out of any page of "storage" (eg, memory)? Think of it as if the processor could access any part of storage (eg memory). There would be no more "boot" process to "load" an OS. The system could simply immediately begin execution at the OS start point and rapidly initialize. That is for a cold start. For warm start, the system could simply go into a "sleep" mode, and then wake up later. All running processes still intact.
Imagine the architecture of how an executable might run. We typically think of memory as having, from low address zero, the program code, read only initialized constants / variables, then read/write variables, then a heap, and from the other end of address space, a stack. Now what if, by having a vast address space, one could easily have multiple stacks, heaps, or maybe things we cannot easily think of due to our ingrained historical thinking. There would be no virtual memory. Programs could simply malloc storage. There would need to be some way to limit and manage this. And one possibility is GC which has definitely already come of age. Just rambling about various ideas here.
My point is that OS design might be able to radically change due to a development like this.
If a lazy person with no education can cross the border and take your job, we need to upgrade your job skills.
(Score: 1, Interesting) by Anonymous Coward on Tuesday January 11 2022, @04:09PM
Just because it flows naturally from the attributes of the hardware doesn't mean it's a bad idea.
Suppose something crashes. Now you can't just restart it, you have to reinstall.
Most of what computers do is ephemeral. Hardware devices are used by one program, then another. No program can assume it will be in the same state it was in before. Network connections expire if not used, or need to be reconnected when the device changes networks. Data gets stale and must be updated.
You couldn't dual boot operating systems.
Most of the things you cite as advantages are already available. If you want multiple stacks, that's what threads are. If you want multiple heaps, you can do that, although the main use for that is for generational garbage collectors, which are also a thing. Presenting files as memory is already done by mmap().
The answer to "it would be nice to not have to use virtual memory" is not actually "non-volatile memory" but rather "more RAM."
There are really just no advantages to this. Working memory and storage should be separate.
(Score: 2) by VLM on Wednesday January 12 2022, @01:05PM
The user experience has been around a long time. Historically the tech was either disposable or slow but that wouldn't change the experience and procedures of use.
In the old days you could buy a sram with a lithium backup battery that would sometimes last as long as five years or so. Somewhat more recently you can buy "fram" chips that are like very small slow core memory.
The main problems I recall from trying to mess with this stuff revolve around ROCK solid power supply supervisor chip type problems and crash recovery and failure modes.
With a 50 MHz clock and a 32 bit address space, if the CPU goes "nuts" as the power shuts off normally no one cares, but with nvram the address bus can cycle thru the entire space sooner or later. Worse with older tech, imagine a 2 mhz z80 and 16 bit address space, that thing can wipe thru its entire address space 30 times per second as power is lost. So all you need is write enable active and the data bus full of noise as the CPU address bus tries to fetch every address in sequence to "execute" it as it powers down and the memory is wiped.
Also see crash recovery, nvram means the nvram is always consistent so write your code assuming it'll always be consistent, but its really more like in practice that memory access is like reading a disk that needs to be fsck'd.
As for failure modes I vaguely recall a shortwave radio an Icom R-71 or something like that where the CPU went bonkers when the backup battery lost its mind. Was recoverable but was a pain. I would imagine most of those radios hit the trash pile the instant the battery died. "If you can write it you can read it" well not really and not all the time. I assume the marketing claims this device will last forever, just like flash lasts forever and my cellphone battery lasts for a week between charges. IRL when the chip fails the software will have to tolerate it. Or not tolerate it, to drive more sales, greenwashing style "think of all the little watch batteries we saved the planet from, too bad you have to buy an entire new car every five years because the engine computer will self destruct when the battery fails" I'm sure failure will continue to be marketed as greenwashing.