Just the facts, STT-MRAM: Your DRAM replacement's on its way

Spin Transfer-Torque Magnetic RAM (STT-MRAM) is a future DRAM replacement candidate. It uses one of two different spin directions of electrons to signal a binary one or zero. IBM and Samsung scientists have published an IEEE paper* demonstrating switching MRAM cells for 655 devices with diameters ranging from 50 down to 11 …

[Novex]

Call me jaded...

...but after the non-news of 3D Xpoint (to date, at least) any news of 'new' memory technologies feels like vapour. Can we actually see a working module of manufacturable parts before we start talking about how such tech is going to revolutionize memory?

[2460 Something]

Re: Call me jaded...

Agreed. Massive hype over xpoint ... then it's just fizzled to nothing.

The cynic in me wonders how much someone in the intel advertising department made on share sales.....

[Anonymous Coward]

Re: Call me jaded...

The difference with 3dxpoint is that it actually exists, and it exists in commercial quantities. STT-MRAM is still in the lab.

Right now there are hundreds of Intel and Micron partners who are working with real engineering samples of xPoint. Production of xPoint is proceeding and stocks are being built. Unfortunately, that isn't enough to get xpoint working.

If you really want xpoint to do anything you need to have chips and chipsets that work with it. You need those to enter the market. And you can't push those into the market until the current stuff is out of the channel.

3dxpoint isn't held up because the technology is pants. It's held up because the chips and chipsets that will ultimately use it have been delayed! Some of that delay is are the problems around the 10nm shrink, but a lot of the delays are due to the x86 market just being so ass that existing stock can't be cleared fast enough.

But xpoint does exist. It's out there in the hands of partners. The question is simply "when do plebians get it"? That answer is -------+++++++CARRIER LOST

[Brandon 2]

Re: Call me jaded...

Sounds a lot like all the new battery technologies that fuel dissertations...

[Kubla Cant]

Re: Call me jaded...

Sounds a lot like all the new battery technologies that fuel dissertations...

Isn't fuel dissertation one of the new battery technologies?

[Bronek Kozicki]

Re: Call me jaded...

Well you do have the point, but personally I approach such news differently. I rather see it as a news from the front of a prolonged and very expensive war. We are far from winning the war, or even a battle, but at least we have not given up yet.

[Charles 9]

Re: Call me jaded...

Trouble is that TOO MUCH news is creating a "crying wolf" problem. We've been looking for a fusion of fast non-volatile DRAM for so long that ANY news draws excessive attention. IMO, it's better to be pleasantly surprised (assume the worst until good news comes along) than to be disappointed (assume the best until bad news comes along). Given this context, headliner articles should be reserved for the news one of these "promising" techs has begun mass production. Until then, drop the priority of these kinds of articles. Or at the least, plaster the caveat "still in early development" or "not due in the mainstream anytime soon".

[e_is_real_i_isnt]

Re: Call me jaded...

I felt that after bubble memory vanished in a puff of smoke. Rotating rust has a long, if increasingly unappreciated life ahead of it.

[JeffyPoooh]

Re: Call me jaded...

Hype is the new normal. Everything is Pure-B.S. press releases these days.

Once we're all perfectly jaded, the actual cure for cancer and free energy will be ignored for 15 years.

[Anonymous Coward]

Re: Call me jaded...

And that's why I've stopped buying newspapers. And perhaps why they're feeling the pinch, too. Because people get SO jaded the headlines don't draw them anymore. Unless it's a REAL miracle (or the end of the world), no one feels motivated anymore.

[allthecoolshortnamesweretaken]

Thumbs up for the "Dragnet" reference, Chris!

[Pascal Monett]

"(STT-MRAM) is a future DRAM replacement candidate"

Oooh, yet another fantastic achievement with enormous promise and mouth-watering performance !

I am so happy for it.

Call me when the future has arrived, I'll see about getting excited about it then.

[Anonymous Coward]

It will only replace DRAM if it is cheaper

If it is faster it might see use in servers and a few gamers who like to tune their systems to an inch of its life, but it won't see the mass market if it costs even if it is twice as fast if it costs 20% more.

If it is lower power it might see use in premium phones, and servers, but in standard laptops DRAM power is a rounding error and in PCs connected to the wall it is irrelevant in the cutthroat zero margin PC market.

[Charles 9]

Re: It will only replace DRAM if it is cheaper

So you're saying its nonvolatile nature won't help it any?

[Anonymous Coward]

Re: It will only replace DRAM if it is cheaper

To the extent that it saves power, sure. But most people don't turn their computers (PCs, laptops, phones) "off" very often anyway and unplanned power loss is really only a problem for PCs without battery backup, and even then only for poorly written apps that don't write their state to stable storage often enough.

[Ru']

Re: It will only replace DRAM if it is cheaper

"It will only replace DRAM if it is cheaper" - wouldn't it be like SSDs though, which were smaller and (way) more expensive than standard HDs when introduced but are now quite popular? As long as there is a definite advantage (and prices can be reduced with quantities and ongoing development) I can't see why it couldn't necessarily take DRAM share. If it ever becomes viable, of course.

[Anonymous Coward]

"IBM and Samsung scientists have published an IEEE paper..."

"IBM and Samsung scientists have published a paper, in an IEEE rag (an organization who will happily publish just about any old crap, so long as it has colourful pictures)..."

There. Fixed it for you...

[PNGuinn]

I can't be bothered to think of one silly enough

OOOH, clever - but it'll never catch on ...

... Not with that acronym.

Answers on a postcard ....