6 ms·

> This is all within the 3.5-inch form factor. Data density is great and all, but this line got me wondering if we wouldn't be supporting more than just the 2.

> This is all within the 3.5-inch form factor. Data density is great and all, but this line got me wondering if we wouldn't be supporting more than just the 2.5" and 3.5" form factors. I'm suspect there would be a market for storage devices with twice the volume and 50% or more storage. Or does the 3.5" form factor have something to do with the limitations of spinning rust tech?

boomboomsubban 6y ago
>I'm suspect there would be a market for storage devices with twice the volume and 50% or more storage. With your extreme example, I doubt it. Already you can get two drives of 75% of the capacity for just a bit more money, and that would perform far better than one drive of twice the volume.
- sitkack 6y ago
  I wouldn't look at current drive economics to determine that a slight increase in platter size wouldn't enable more storage at a lower incremental cost. The current form factors are baked in, how would we know what the free variables are?
  - boomboomsubban 6y ago
    I think their example of 50% more storage at 2x the volume would need to be a fair amount cheaper than the comparison drive to be considered. I'm unsure what the economics would be, but even at the same price as the comparison drive there wouldn't be a ton of interest.
- pixl97 6y ago
  I think you're neglecting that the biggest issue in datacenter density is power usage.
  - boomboomsubban 6y ago
    I may be completely wrong, so feel free to correct me, but isn't that largely from cooling? I'm unclear how much larger hard drives would help, I'm not sure if it's obvious either.
timc3 6y ago
Changes in form factor/standardisation can have enormous others costs associated. For example shipping, enclosures, re-tooling etc..
sandworm101 6y ago
>>does the 3.5" form factor have something to do with the limitations of spinning rust tech? There are issues. As the platters get bigger they get wobbly, more probe to vibrations. The difference in platter speed between the inner and outer 'rings' of data also increases. There is no hard reason for 3.5 over 4.5 or 4.8 inches, but the advantages of going bigger do not outweigh the practicalities of trying to push a new form factor for a commodity product. >> spinning rust This is a common refrain, but seagate and the others are proving that HDD tech is not 'rusting'. With lower per-TB costs, and phenomenal reliability numbers (<1% chance of failure per year) HDDs are going to be around for a while. Outside of the datacenter, most users are more limited by more by network speeds than storage speeds.
- johnsoft 6y ago
  Other issues with larger platters: - more power usage - more heat - more noise Datacenters in particular are not interested in increasing heat or power usage. 5.25" hard drives used to be common, but were phased out sometime around Y2K.
- sennight 6y ago
  >> spinning rust > This is a common refrain, but seagate and the others are proving that HDD tech is not 'rusting'. It is a common refrain because disk platters were originally coated with iron oxide, and iron oxide is rust. It isn't a value judgement.
  - matmatmatmat 6y ago
    Isn't it a little hard to accept that it isn't a value judgement when we generally have a lower value of things that have rusted? I think the original comparison to 'spinning rust' was tongue-in-cheek and not meant to be taken so-very-seriously, but it's kind of taken on a life of its own and, personally, I feel that it's gone a bit far and undeservedly undervalues hard drives.
    - monocasa 6y ago
      It's a term from an era where HDDs were the best way to bulk store data (except maybe tape in some applications, which where also iron oxide). I think you're reading too much into it.
tablespoon 6y ago
> I'm suspect there would be a market for storage devices with twice the volume and 50% or more storage. Or does the 3.5" form factor have something to do with the limitations of spinning rust tech? There used to be larger form factors for HDs, for instance full-height 5.25 inch (think two DVD drives stacked on top of each other). This Wikipedia page has pictures: https://en.wikipedia.org/wiki/List_of_disk_drive_form_factors https://en.wikipedia.org/wiki/List_of_disk_drive_form_factor... This was apparently a product that took the strategy you described in the 90s: https://en.wikipedia.org/wiki/Quantum_Bigfoot https://en.wikipedia.org/wiki/Quantum_Bigfoot
- dehrmann 6y ago
  There might be issues with Bigfoot-like drives with platter stability at the edges, but I have no idea if this is actually the case. What I do know is you get better performance towards the edge of the platters, and for larger drives, the variability in throughput might be a downside for some buyers. Seeks times are also slower. There was a variant of the WD Raptor that took the opposite approach: it put a 2.5" 10k RPM drive in a 3.5" enclosure, but this drive played the role SSDs do today; they weren't for mass storage.
notacoward 6y ago
> storage devices with twice the volume and 50% or more storage More the opposite, really - same capacity in lower volume. The people who buy most disks are a lot less concerned with capacity per disk than capacity per rack. Sometimes capacity per kW or capacity per BTU, but mostly capacity per m^3 in my direct experience. I also suspect that a lot of consumers would prefer the same capacity in a smaller package, since they mostly boot off flash and their disk is basically a cache for much larger amounts of data elsewhere.

boomboomsubban 6y ago

>I'm suspect there would be a market for storage devices with twice the volume and 50% or more storage. With your extreme example, I doubt it. Already you can get two drives of 75% of the capacity for just a bit more money, and that would perform far better than one drive of twice the volume.

sitkack 6y ago
I wouldn't look at current drive economics to determine that a slight increase in platter size wouldn't enable more storage at a lower incremental cost. The current form factors are baked in, how would we know what the free variables are?
- boomboomsubban 6y ago
  I think their example of 50% more storage at 2x the volume would need to be a fair amount cheaper than the comparison drive to be considered. I'm unsure what the economics would be, but even at the same price as the comparison drive there wouldn't be a ton of interest.
pixl97 6y ago
I think you're neglecting that the biggest issue in datacenter density is power usage.
- boomboomsubban 6y ago
  I may be completely wrong, so feel free to correct me, but isn't that largely from cooling? I'm unclear how much larger hard drives would help, I'm not sure if it's obvious either.

timc3 6y ago

Changes in form factor/standardisation can have enormous others costs associated. For example shipping, enclosures, re-tooling etc..

sandworm101 6y ago

>>does the 3.5" form factor have something to do with the limitations of spinning rust tech? There are issues. As the platters get bigger they get wobbly, more probe to vibrations. The difference in platter speed between the inner and outer 'rings' of data also increases. There is no hard reason for 3.5 over 4.5 or 4.8 inches, but the advantages of going bigger do not outweigh the practicalities of trying to push a new form factor for a commodity product. >> spinning rust This is a common refrain, but seagate and the others are proving that HDD tech is not 'rusting'. With lower per-TB costs, and phenomenal reliability numbers (<1% chance of failure per year) HDDs are going to be around for a while. Outside of the datacenter, most users are more limited by more by network speeds than storage speeds.

johnsoft 6y ago
Other issues with larger platters: - more power usage - more heat - more noise Datacenters in particular are not interested in increasing heat or power usage. 5.25" hard drives used to be common, but were phased out sometime around Y2K.
sennight 6y ago
>> spinning rust > This is a common refrain, but seagate and the others are proving that HDD tech is not 'rusting'. It is a common refrain because disk platters were originally coated with iron oxide, and iron oxide is rust. It isn't a value judgement.
- matmatmatmat 6y ago
  Isn't it a little hard to accept that it isn't a value judgement when we generally have a lower value of things that have rusted? I think the original comparison to 'spinning rust' was tongue-in-cheek and not meant to be taken so-very-seriously, but it's kind of taken on a life of its own and, personally, I feel that it's gone a bit far and undeservedly undervalues hard drives.
  - monocasa 6y ago
    It's a term from an era where HDDs were the best way to bulk store data (except maybe tape in some applications, which where also iron oxide). I think you're reading too much into it.

tablespoon 6y ago

> I'm suspect there would be a market for storage devices with twice the volume and 50% or more storage. Or does the 3.5" form factor have something to do with the limitations of spinning rust tech? There used to be larger form factors for HDs, for instance full-height 5.25 inch (think two DVD drives stacked on top of each other). This Wikipedia page has pictures: https://en.wikipedia.org/wiki/List_of_disk_drive_form_factors https://en.wikipedia.org/wiki/List_of_disk_drive_form_factor... This was apparently a product that took the strategy you described in the 90s: https://en.wikipedia.org/wiki/Quantum_Bigfoot https://en.wikipedia.org/wiki/Quantum_Bigfoot

dehrmann 6y ago
There might be issues with Bigfoot-like drives with platter stability at the edges, but I have no idea if this is actually the case. What I do know is you get better performance towards the edge of the platters, and for larger drives, the variability in throughput might be a downside for some buyers. Seeks times are also slower. There was a variant of the WD Raptor that took the opposite approach: it put a 2.5" 10k RPM drive in a 3.5" enclosure, but this drive played the role SSDs do today; they weren't for mass storage.

notacoward 6y ago

> storage devices with twice the volume and 50% or more storage More the opposite, really - same capacity in lower volume. The people who buy most disks are a lot less concerned with capacity per disk than capacity per rack. Sometimes capacity per kW or capacity per BTU, but mostly capacity per m^3 in my direct experience. I also suspect that a lot of consumers would prefer the same capacity in a smaller package, since they mostly boot off flash and their disk is basically a cache for much larger amounts of data elsewhere.