Seagate To Double HDD Speed With Multi-Actuator Technology

Handruin

Administrator
Joined
Jan 13, 2002
Messages
13,737
Location
USA
I'll be curious what the noise level will be like. It's a neat idea but I really have no faith in Seagate hard drive reliability. I would have liked to see multiple actuators in opposite ends on the same platters to improve random seek times.
 

Tannin

Storage? I am Storage!
Joined
Jan 15, 2002
Messages
4,448
Location
Huon Valley, Tasmania
Website
www.redhill.net.au
Three comments:

(a) This is the return of the undead. Seagate were building 2HP drives back in ... oh ... must have been the mid 1990s.
(b) But what is the point? It's not as if you don't use SSDs when performance matters.
(c) The Tom's article looked quite interesting, but the website was so appallingly designed that I stopped reading and gave up. I mean, that shitehole is seriously bad.
 

Handruin

Administrator
Joined
Jan 13, 2002
Messages
13,737
Location
USA
I could see this implementation as a way to manage/mitigate increasing seek times as the number of platters increases thereby causing the size of the actuator to increase in size. We didn't have helium drives back in the day so now with that as an option, a stupidly-large platter count could be more feasible.
 

sechs

Storage? I am Storage!
Joined
Feb 1, 2003
Messages
4,709
Location
Left Coast
I feel like platter counts already are stupidly large.

Putting the two actuators is the same place is what makes this economically feasible. Having two completely independent ones would really be two drives sharing a common spindle.
 

LunarMist

I can't believe I'm a Fixture
Joined
Feb 1, 2003
Messages
16,624
Location
USA
It's not clear if the two drives can be accessed without a special controller or drivers.
 

LunarMist

I can't believe I'm a Fixture
Joined
Feb 1, 2003
Messages
16,624
Location
USA
I have an old 10K WD drive with a clear poly cover. The actuators are quite noticeable when running the random access tests. It is interesting for a few minutes.

I suppose it will be years before there are any NAS that support these drives. This new style effectively adds nearly another spindle per drive, so presumably the reduction of spindles at a given high capacity doesn't kill the IOPops. It would be nice if there were a firmware setting for an internal RAID that would not require HBA support.
 
Top