back to article Nanodot memory smashes RAM, sets new speed record

Boffins in Taiwan and the University of California predict that nanoscale CMOS memory could soon be on its way after research showed nanodot memory operating 10 to 100 times faster than current RAM. The electro-optics researchers also emphasised that they had used materials that are compatible with mainstream integrated circuit …

COMMENTS

This topic is closed for new posts.
Boffin

How are they steering the laser?

Surely using a parallel beam and a DLP chip, or similar, would bring it back to conventional speeds. If they're using a perpendicular beam, well, that sounds complex, bulky, and expensive.

1
0

maybe i'm stupid

but i didnt think RAM was the bottleneck at present to require a 10-100 times improvement.

0
4
Silver badge
Boffin

Re: maybe i'm stupid

why do you think we have fast caches for chips? Imagine the entire memory working at the speed of the CPU. That would be awesome.

At the moment, I have to think hard about cache friendly processing orders. Getting it wrong can incur a 10-fold speed penalty, easily. If have a set of for-loops to traverse an image, having the x-coordinate loop outside the y-coordinate loop is tens times slower than the reverse, because of the way images are stored (row by row). A step in x moves to the next element in memory (= cache hit with standard read-ahead), whereas a step in y steps a whole row of data further, yielding a cache miss.

Such simple cases are easily sorted out, but some image processing has data-driven processing orders, very frequently requiring odd memory jumps. In these cases getting rid of latency is a godsend. Also, think of multi-core: ensuring cache coherence is a pain. Older Cray machine had no cache, and the memory worked at the speed of the CPU. This is much simpler and yields much better parallelization.

8
1
Anonymous Coward

Re: maybe i'm stupid

CPU/RAM communication speed is indeed a big bottleneck, which is why there is so much attention paid to CPU caching issues both on the hardware and software sides.

Now, faster RAM is clearly important but it's pretty much useless if the bus speed isn't up to par. Unfortunately I don't have the faintest idea whether bus speeds can easily be improved, maybe some competent commentard can enlighten us on this matter?

1
0

See... i knew i was stupid.

Cheers for that. Today I have learned something that, granted, I will probably never use, but none the less, its good to know :)

13
0
Headmaster

Re: See... i knew i was stupid.

Nobody is stupid because they don't know something. Being stupid is being incapable or unwilling to learn.

24
0
Silver badge

Re: See... i knew i was stupid.

Conceding you are stupid (or that you do not know something) is a clear sign of intelligence.

Politicians never make such concessions

10
0

Re: See... i knew i was stupid.

A great philosopher once said "stupid is as stupid does".

0
0
Bronze badge
Happy

Re: See... i knew i was stupid.

Ignorance is curable, alas stupidy is not.

0
0

This post has been deleted by its author

Bronze badge
WTF?

I dont quite see the progress here

Latency of DRAM is normally around 70ns. They claim to have achieved 1us which equals to 1000ns. Where is the progress?

2
0

Re: I dont quite see the progress here

I was wondering the same thing, and as i've not yet read the paper and don't have an AIP subscription (or want to pay just for the curiosity) I hope we might have someone who has who can enlighten us.

I was equally mystified by the description of 7V for a write/erase cycle as 'low'. Both the write cycle timing and voltage are ok in relation to NVRAM technologies such as EEPROMs, but make no sense in the DRAM usage context.

1
0
Paris Hilton

Re: I dont quite see the progress here

I am similarly confused. Did they mean they have a write cycle time of 1 nanosecond, perhaps?

0
0

Re: I dont quite see the progress here

My guess is that since the only mention of RAM is outside of the quote (which says non-volatile memory,) maybe El Reg got it wrong? It certainly does look more like a replacement for NAND (or possibly NOR) Flash rather than DRAM. A 1us p/e cycle would be a hell of an improvement over current NAND.

Also, +/- 7V doesn't seem remotely 'low' compared to any sort of modern volatile or non-volatile solid state memory. Maybe they mean it's low compared to alternatives that are currently in development? Or low compared to their last prototype? It certainly doesn't scream 'efficient operation on battery power,' which is kind of necessary for mobile use. Then again, DC-DC converters aren't all that bad these days.

2
0
Anonymous Coward

Re: I dont quite see the progress here

"positioned on the part of the metal gate layer above a dot and firing a sub-millisecond burst of hot light"

I hope it's **really** sub-millisecond since RAM can be access/changed in a matter of nanoseconds.

0
0
GTX
Devil

Re: I dont quite see the progress here

Ya! When I saw mention of the dreaded "millisecond" (even if it IS "sub-"), my mind was dragged, kicking and screaming, back to the millisecond delay lines, from which our programs loaded... OMFG!! - Almost as slow as a Windows boot!

0
1

7v for _burning_ metal oxide is very low

when you consider how much power it takes to make a laser burn a spot that melts metal rather than just lighting it up. look at it in that context, then 7v is very low voltage for the power needed to melt the metal spot in the nanodot area.

1
0

And don't forget

that the burning part that takes a longer time (up to 1ms) is the writing part, not the reading part of the cycle, so the relationship to cmos is much closer than that of regular ram.

0
0
Gold badge
Unhappy

non volatile sub-*microsecond* memory

welcome to the 1970's.

I'll get my flares and cheese cloth shirt out.

1
1
Anonymous Coward

when we can buy it then we will care

as usual, yet another faster RAM story that is going nowhere any time soon ( the next two years) until the general public can actually buy these faster ram products in shops no one really cares do they!

there's been far to many stories proclaiming faster RAM that we still cant actually buy in retail i don't care what they do in the lab any more that's for sure.

0
0
Bronze badge

at least they picked a paper title for the stupid and the knowing

0
0
Coat

No, I'm being stupid....

Data write (storing charge) and data erase (removing charge)

So... this RAM can only store 1's?

Does that mean it can only store data that uses "zero-compression"? - you know, the compression where they fit a movie onto a 5.25" floppy disk by removing all the 'useless' data - the 0's.

Yes, thats my jacket with the highly illegal terrorist weapon (green laser pointer) in the pocket. Back off man, I'm an architect.

0
0
This topic is closed for new posts.

Forums