Read more.Quote:
Researchers say new device can "combine the advantages of both without their drawbacks".
Printable View
Read more.Quote:
Researchers say new device can "combine the advantages of both without their drawbacks".
Great, will it come in ddr4 compatible sticks of 512GB or greater?
It could be the future or it could be a pipedream fantasy...
Surprised it didn't involve graphene sorcery or carbon nanotubes!
Intel's Xpoint made such bold claims right up until they tried shipping product. Good luck to them, would be lovely if they get it to work but I won't hold my breath.
So old school. Now it's nothing without being quantum in some way. This, of course, features "quantum-mechanical properties of an asymmetric triple resonant-tunnelling barrier". Now that's marketing progress.Quote:
Originally Posted by Tabbykatze
Looking forward to never hearing about this ever again.
Don't know why you wouldn't want to hear more about a potential technonogy advance, even if was just to learn that it didn't fulfil its promise when it came tp production.Quote:
Originally Posted by iggy
Probably because he's burnt out on lofty tech claims that are long on promise, and short on results.Quote:
Originally Posted by peterb
Just in case you're not aware, Professor, there are two issues with DRAM. If your memory uses 1% of the energy that DRAM uses but takes 100 times as long to change state, it's not a replacement. Unfortunately, until you inform us of the speed of your memory, you have told us almost nothing.Quote:
Professor Hayne said: “The ideal is to combine the advantages of both without their drawbacks, and this is what we have demonstrated. Our device has an intrinsic data storage time that is predicted to exceed the age of the Universe, yet it can record or delete data using 100 times less energy than DRAM.”
Nothing would go over his head. He is too fast.Quote:
Originally Posted by aidanjt
Well it seems it's a working piece of tech not just theoretical so could well have a large impact. The article linked in Nature is worth a read.
You mean, it's not going to be painfully expensive?Quote:
"without their drawbacks,"
For purely scientific reasons, this report is worthwhile, if only to stay current with the very latest reports about R&D in this field.
Nevertheless, what I find even more interesting, from a practical point of view, is the raw bandwidth that will be available from x16 PCIe 4.0 and PCIe 5.0 expansion slots e.g. using "4x4" add-in cards like the ASRock Ultra Quad M.2 card, particularly when multiples of the latter are installed and working.
A little simple arithmetic says that a single x16 PCIe 4.0 expansion slot has more raw bandwidth than most high-end DDR4 DRAM:
16,000 / 8.125 x 16 PCIe 4.0 lanes = 31,507 MB/second (raw bandwidth)
32,000 / 8.125 x 16 PCIe 5.0 lanes = 63,014 MB/second (raw bandwidth)
Doing the inverse now:
31,507 / 8 ~= DDR4-3,938
63,014 / 8 ~= DDR4-7,876
PCIe 4.0 has arrived, so key your eyes on the major storage players e.g. Samsung et al.
We are about to witness another "race to the top"!
Sounds too good to be true - there will be something that means it's not feasible for commercial production.
And am I the only one who is dubious of the idea of NVM replacing RAM? If all else fails, you turn if off, and on again - it clears the RAM. NVM, not so much.
In a world of erasing an EPROM with 20 minutes in the ultraviolet chamber, we once considered Flash to be too good to be true.Quote:
Originally Posted by Dashers
But I also see it unlikely that such storage will replace ram, though it introduces a new and interesting tier between dram and ssd. As for being stuck forever in a blue screen, we already have suspend to ram which is logically the same (and I don't think I ever use, though I'm not a big user of laptops).