XPoint capacity to surpass DRAM by 2030
We have just learned about a report by Coughlin Associates and Objective Analysis called Emerging Memories Take Off, courtesy of Tom Coughlin. The report looks at 3D XPoint, MRAM, ReRAM and other emerging memory technologies and says their revenues could grow to $44 billion by 2031. That's because they will displace some server DRAM, and also NOR flash and SRAM — either as standalone chips or as embedded memory within ASICs and microcontrollers.
The emerging memory market is set to grow substantially with 3D XPoint revenues reaching $20 billion-plus by 2031, and standalone MRAM and STT-RAM reaching $1.7 billion in revenues by then. The report predicts that the bulk of embedded NOR and SRAM in SoCs will be replaced by embedded ReRAM and MRAM.
A chart shows XPoint capacity ships crossing the 100,000PB level in 2028 and so surpassing DRAM, whose capacity growth is slowing slightly. The chart shows XPoint capacity shipped being 1000PB this year. That number will grow 100x to 100,000PB in 2028.
Related: Micron Abandons 3D XPoint, Puts Fab Up for Sale
Micron Sells 3D XPoint Fab to Texas Instruments, Not Intel
(Score: 4, Interesting) by FatPhil on Thursday September 09 2021, @10:31AM
Exponential growth being extrapolated way into the future is always a problem. IT-related things have managed to keep the pretence it's possible up for way longer than any other field, but even that has to slow down eventually. Predicting a *faster* exponential in the future seems even braver (stupider? marketting bullshittier?) than just predicting that it can continue at the same rate.
I'm not being Debbie Downer about the tech - RAM's seen very few really inovative developments for many decades, and this not-just-a-grid-of-transisters development is a very interesting one. I hope they're right, as either costs or power consumption (and thus cost) would naturally drop if they are.
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 2) by progo on Thursday September 09 2021, @02:14PM (4 children)
What is 3D XPoint? Did I miss a memo?
(Score: 4, Informative) by takyon on Thursday September 09 2021, @02:25PM (2 children)
It is a non-volatile memory technology intended to exist in a tier between DRAM and NAND, or replace/reduce one or the other in some cases.
Compared to DRAM, 3D XPoint is denser, cheaper, but slower.
Compared to NAND, 3D XPoint is faster, but less dense and more expensive.
https://en.wikipedia.org/wiki/3D_XPoint [wikipedia.org]
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by Booga1 on Saturday September 11 2021, @01:25AM (1 child)
If it had originally launched in larger capacities, I think it would have had a serious chance of displacing the database SSD market entirely. Unfortunately, the tiny sizes were neither big enough to displace the SSD nor fast enough to displace DRAM.
I think it would have been a hit if Intel had just been a bit bolder on making a push for a specific market. Instead, they half-assed it and set a trend. Second generation Optane was really good, but by then the SSD market was on par for capacity and "good enough" to start taking over bulk storage in the VM/VPS market.
Optane(3d Xpoint) is great, but their timing was off. I think these predictions are seriously miscalculating the current SSD market and trends. Optane is so dead that even Intel doesn't seem to believe in it any more.
(Score: 0) by Anonymous Coward on Saturday September 11 2021, @07:39AM
Maybe you are right about the botched launch strategy (similar to how SMR has been botched), but we love it in certain roles. Used correctly, we've seen increases in excess of 10x on some workloads and the minimum we see is in the neighborhood of doubling. For locality of data especially, it really is hard to beat.
(Score: 5, Informative) by rigrig on Thursday September 09 2021, @03:50PM
I didn't remember it either, but apparently we covered it [soylentnews.org] when it came out. Guess this is one of those rare cases where new tech actually did take off.
Besides an explanation, this "ancient" article [theregister.com] also contains a handy chart [regmedia.co.uk] showing how it fits between SSD and DRAM.
No one remembers the singer.
(Score: 3, Insightful) by Rosco P. Coltrane on Thursday September 09 2021, @05:45PM
Even back in the 60s when technology moved slowly by today's standard, he would could predict which technologies would win out 10 years later was a visionary man indeed. So when someone in 2021 tells me what the industry will be like in 2030, I really don't pay much attention because something entirely different will almost certainly happen.