Some design and prototyping issues on a 20000 rpm HDD spindle motor with a ferro-flui

I was trying to find a prototype 20K RPM HDD - something like the Rumored WD Raptor 20K RPM SATA HDD that was supposedly going to compete with the first generation SSDs
 
Especially with 3d NAND sampling from major manufacturers now, I really doubt this will ever happen.
 
Olga-SAN said:
high capacity sata disks for backups did not kill tapes they just got better ))

the same about ssds they did not wipe off 15k rpm drives so yes we'll see 20k and 25k soon
Click to expand...
No we won't. You're delusional if you think we will. :rolleyes:
 
Olga-SAN said:
high capacity sata disks for backups did not kill tapes they just got better ))

the same about ssds they did not wipe off 15k rpm drives so yes we'll see 20k and 25k soon
Click to expand...

Whaaat? I think you're the only person on the planet that thinks this.
 
Even if you happened to be correct, it's not being wise to just wildly guess and then lucking out. It's called blind luck.

There are no indications that any manufacturer is making a 20k+ RPM drive, and with good reason. The cost in R&D alone is ridiculous, manufacturing it, probably just as ridiculous. And for what...? What's the purpose? Performance? Buy an SSD. Capacity? Buy some 10k drives or even 7.2k drives. Combination of capacity and performance? Cache those "slow" spinners with some SSDs.
 
I just don't see the benefit that 20k drives would bring. Fast storage is here already, and 7-10k drives have a solid hold on cost to size right now.
 
wdeydwondrer said:
I just don't see the benefit that 20k drives would bring. Fast storage is here already, and 7-10k drives have a solid hold on cost to size right now.
Click to expand...

I can only imagine a data center filled with 20k drives, that would be like walking inside the engine of a SR-71.
 
"blond luck" in my case ))

hamr deployment is actually slowing down the whole thing not power & heat issues (these can be depleted with smaller form factor sub 1.8")

spazoid said:
Even if you happened to be correct, it's not being wise to just wildly guess and then lucking out. It's called blind luck.

There are no indications that any manufacturer is making a 20k+ RPM drive, and with good reason. The cost in R&D alone is ridiculous, manufacturing it, probably just as ridiculous. And for what...? What's the purpose? Performance? Buy an SSD. Capacity? Buy some 10k drives or even 7.2k drives. Combination of capacity and performance? Cache those "slow" spinners with some SSDs.
Click to expand...
 
Olga-SAN said:
"blond luck" in my case ))

hamr deployment is actually slowing down the whole thing not power & heat issues (these can be depleted with smaller form factor sub 1.8")
Click to expand...

I don't mention heat or power at all, I mention R&D (basically developing a completely new type of harddrive) and manufacturing (something like this is going to require some mental tolerances), and if you want it to sell, it has to be WAY cheaper/GB than SSD.

You'd have to be one hell of a salesman to make that business case look good to your CEO.
 
Olga-SAN said:
high capacity sata disks for backups did not kill tapes they just got better ))
Click to expand...
Tapes attend to very different requirements. Trucks didn't kill freight trains, so what?

the same about ssds they did not wipe off 15k rpm
Click to expand...
... yet. Their significance is continuously dropping, because SSDs are taking their market share.

Fast rotating disks exist to optimize one metric: Access time / IOps.
And congrats, with 20krpm you'd be able to get about 333iops/s best case. Thats two orders of magnitude less than enthusiast class SDDs and about three orders slower than enterprise PCIe SSDs.
 
i completely agree

unless spinning disks are going to dramatically increase density and capacity to keep 5x-10x $/TB difference nobody would be interested

spazoid said:
I don't mention heat or power at all, I mention R&D (basically developing a completely new type of harddrive) and manufacturing (something like this is going to require some mental tolerances), and if you want it to sell, it has to be WAY cheaper/GB than SSD.

You'd have to be one hell of a salesman to make that business case look good to your CEO.
Click to expand...
 
hold on for a second i'm not talking desktops of servers i'm talking just one use case which is archive storage and long time data vaulting

guys like sepaton been claiming nobody would use tapes up to the point tape technology improved dramatically

sure but there are cases when you need more capacity **and** more iops **and** you're on budget

all-flash isp & hosting provider? :)

zrav said:
Tapes attend to very different requirements. Trucks didn't kill freight trains, so what?


... yet. Their significance is continuously dropping, because SSDs are taking their market share.

Fast rotating disks exist to optimize one metric: Access time / IOps.
And congrats, with 20krpm you'd be able to get about 333iops/s best case. Thats two orders of magnitude less than enthusiast class SDDs and about three orders slower than enterprise PCIe SSDs.
Click to expand...
 
Olga-SAN said:
hold on for a second i'm not talking desktops of servers i'm talking just one use case which is archive storage and long time data vaulting

guys like sepaton been claiming nobody would use tapes up to the point tape technology improved dramatically

sure but there are cases when you need more capacity **and** more iops **and** you're on budget

all-flash isp & hosting provider? :)
Click to expand...

They still wouldn't use 20k's if you are only looking for single-write drives.
 
Olga-SAN said:
sure but there are cases when you need more capacity **and** more iops **and** you're on budget
Click to expand...

Yes, and in these situations you use spinners (typically 10k) and SSD or flash caches. You would never, ever, use 20k drives.

The fact that you think otherwise - and insist on it - tells me that you haven't set foot in a DC in the last 5 years. Even cheapish servers from the major players like IBM/Lenovo come with SSD boot drives as the entry level option. The price of upgrading from 15k to SSD is so little already that it's a non-issue for most (not talking SAN here, just local disk).

20k drives compared to SSD/flash backed 10k drives would give you:
Less usable space
Fewer IOPS
Higher power usage/IOPS
Higher power usage/usable space

So, you have a higher CAPEX, higher OPEX which results in much higher TCO and you get less performance and space in return.

Seriously, why is it that you think 20k and 25k drives would EVER make sense?
 
To add: OEMs are sinking tons of money into flash based storage and their production facilities as it is cheaper to produce, less models of failure and require a lot less space.

I believe they are quicker to produce as well but I don't know much about that.
 
The Enterprise industry is moving AWAY from 15K and moving back towards 10K/7.2K augmented with flash. This 20K nonsense is laughable. Sounds like a 10yr old conversation. Oh wait.
 
Trimlock said:
To add: OEMs are sinking tons of money into flash based storage and their production facilities as it is cheaper to produce, less models of failure and require a lot less space.

I believe they are quicker to produce as well but I don't know much about that.
Click to expand...

It takes about a month to produce a silicon wafer. However, these are usually made in 24 or 25 wafer lots and on a 300mm wafer you could be looking at 300-1000 chips per wafer (more depending on the exact size of the die/chip). And there have been talks of 450mm wafers, though I do not know how that's actually progressing. But it will be at least several years before we see any signs of that.

And you would have a string of these lots in the pipeline. So it might take a month to get them, but at the end of the month you could be looking at 100,000 chips ready to go into product.
 
There are ZERO good reasons to bring 20K+ RPM spindle drives to the market.
- High manufacturing cost
- High end-user price
- High power consumption
- High heat output
- High field failure rate

4TB SSDs are already out. Solid State Drives are only going to get smaller physical sized, larger capacity, cheaper, faster, and even more reliable as time goes on...and rapidly, at that. Speculate about what will realistically be out in 1 year from now. 2 years from now. 5 years from now. It sure as hell isn't going to be data storage devices reliant upon electro-mechanical driven platters.
 
westrock2000 said:
It takes about a month to produce a silicon wafer. However, these are usually made in 24 or 25 wafer lots and on a 300mm wafer you could be looking at 300-1000 chips per wafer (more depending on the exact size of the die/chip). And there have been talks of 450mm wafers, though I do not know how that's actually progressing. But it will be at least several years before we see any signs of that.

And you would have a string of these lots in the pipeline. So it might take a month to get them, but at the end of the month you could be looking at 100,000 chips ready to go into product.
Click to expand...

Right, these figures aren't foreign to me, what is foreign to me is how long it would take to produce a mechanical hard drive. I'm assuming but I believe it takes quite a while longer, more materials and a more complicated process.

Another topic this brings up: I don't think spinners are going away soon (although the market could kill them off in a year). I think they will keep producing them as they have factories for them but I'm sure they are close to maxing out as in they won't dump anymore money into making them bigger. They will use their current factories until they are all converted to fab's for flash memory.
 
Someone should make an SSD that makes whirring and clicking noises because retro.
 
xorbe said:
Someone should make an SSD that makes whirring and clicking noises because retro.
Click to expand...

I bet there's a program for that. *some searching later* Seems like not. You have found an unfulfilled niche, my friend. Take advantage quick.
 
You must log in or register to reply here.
Back
Top