Hacker News Reader

Comment by narrator | Hacker News Reader

It's weird how tape size just seems to increase steadilyin a smooth curve to roughly match whatever hard drives are doing. It makes me wonder sometimes if they had this tech all along, but it was better just to slow roll it out so they could sell multiple generations of hardware. I guess only the material scientists at these companies know what's really going on.

View on HN

Exactly.

IBM's had several generations of tape tech in the can, and has managed the rollouts with calm competence for at least back to '97 when I got started in the backup business.

Their research is a prominent part of the LTO platform, too. My observation of that landscape was that they gave toys to LTO at a rate calculated to extinguish the utility of tape research in other places. This means there's a plausible competition, which IBM wins all the time.

This may sound like I'm down on them for it; I'm not. Their hardware was rock solid, performed as advertised.

Huh that's pretty smart if true.

Look at Wikipedia [0], they have it planned for the next ten years like clockwork.

[0] https://en.wikipedia.org/wiki/Linear_Tape-Open#Generations

> Fujifilm has demonstrated a 580TB raw capacity tape using Strontium Ferrite particles so there is a lot of runway for future capacity increases.

I would guess so.

Well, yes, the article mentions the potential of the strontium ferrite, with existing prototypes having hundreds of terabytes of capacity.

Similarly, HAMR and MAMR will be milked for years in the world of magnetic hard drives.

In one way, this is better to have multiple generations because you have to fill the tapes and with new generations we have better transfer speeds. Also with the backward compatibility, old tapes are transferred to new tapes, so the shelf longevity of the tapes are never really tested.

Interesting video on this https://youtu.be/hlmIfhJkVxg?si=0mV8KO4E_7uBLYRR

Reply on news.ycombinator.com