You have to smile when the financial guy rather than the techie is the one giving the talk on what a vendor's product is going to look like next year. But this is not uncommon in the disk drive and SSD industry, especially among its Big Three: Seagate, Toshiba and WD. Their respective chief financial officers want investors to …
Is HDD market a cartel ?
The Big2 or Big3 manufacturers keep buying smaller competitors. No inovations, just buy outs.
HDD prices remain high and are priced similarly. Consumer always loses.
Then there are those rumours (based on tests though) that WD and Seagate are playing games...
WD restricts hardware functionality in the firmware (to protect other hdd lines) and sets the firmware to do unessecery tasks that reduce the lifetime of the drive. Seagate on the other hand, understates the capabilities of their hdd so people that want NAS or enterprise specs, don't go and buy cheaper desktop hdds.
Re: Is HDD market a cartel ?
> Seagate on the other hand, understates the capabilities of their hdd so people that want NAS or enterprise specs, don't go and buy cheaper desktop hdds.
I was about to reply to this saying that I picked up 2x 3TB Seagate NAS drives yesterday from ebuyer for the same price as their standard 3TB Barracuda drives (£89.99 each). Which I did. However I note that today the NAS drives are £112.72 from ebuyer, so it appears that I got lucky and your point about excess pricing of NAS drives remains valid...
Re: Is HDD market a cartel ?
Gulp!! I work for Seagate. These misconceptions in this post are disappointing and are sometimes repeated. I'd just like to clarify some points.
Each company spends more than $1Bn per year on research and development. HAMR Heat assisted magnetic recording, with a roadmap to 20TB by 2020. The heads needed for this technology use 3 separate quantum effects, some of which quantum physicist weren't even aware 50 years ago.
A 1TB drive today has >100km of track on it, and the head flies at 128kph at a height of about 1/60th the wavelength of light (i.e. about 10 air molecules). The head sucks up or writes bits at >1 Billion bits per seconds. Oh and it all costs less than a pair of hair straighteners!
"HDD prices remain high - consumer always loses"
The cost of a Gigabyte of HDD storage has fallen in 20 years by a factor of about 500,000.
In 1981 our drives cost $340,000 per gigabyte - now it costs $0.07 per gigabyte.
Why the poster says a HDD company would deliberately increase failure rate, I don't understand. There's interplay between capacity and reliability. It's really easy to double capacity - by doubling the bit density or the track density - but then the reliability tumbles. It's really hard and requires significant R&D to get reliability back to where it was. If a HDD company had reliability headroom, would they not use it to drive more capacity???
As for the point on reliability of desktop versus NAS/nearline versus enterprise drives, it is difficult to demonstrate on a single sample (of course) or even with experience of a few tens or hundreds of drives. BUT IT IS REAL. We make 60 million HDDs a quarter. We have the data. There are as many as 15 points in the manufacturing process where the output is a normal distribution (roughly) of reliability. Consider balancing the platter alone. Consider the 'roundness' of bearings. Consider the flatness of the media surface or the distribution of thickness. All of these contribute to reliability and MTTF. We can yield drives from the reliable to the very reliable on this basis alone. In addition, there are totally separate designs for different environments that affect MTTF. Some drives can tolerate constant access, 24 hours a day, >100 IOs per second, for 5 years with an AFR of <1%. Others a few hours a day, light loads with a failure rate of >1% If you use the latter in the former environment you will save yourself a few dollars on purchase cost, but the AFR goes through the roof.
The HDD guys are good guys and gals. Seagate is >57,000 people, who turn up to work everyday to do a great job. They amke about 450 drives every minute of every day, that store about 40% of the world’s data. Personally I think that's awesome and is a testament to man's ability to organise himself. Trust them!
Seagate -- Well, seems it's business as usual.
Seagate’s average capacity/drive is higher than both Toshiba and Western Digital
...But going on my huge box of dead Seagate drives, 'tis far less reliable!
In recent years, trading reliability for density has been a Seagate hallmark (seems it's still business as usual).
45-450EB by 2020
Only a factor of 10 increase, whilst demand is increasing ata much higher rate than that.
Of course, neither SG or WD own NAND foundries, so they're likely to end up being devoured in the same way they killed off their competitors, long before 2020
Re: flash supplies
So SSD projections are a x10 increase and demand is about the same order for HDD at about x13. This is a capacity projection, not units. Since SSD units are much smaller, the units share will be much higher than 450EB/6ZB
The reason HDD is growing slightly faster is that the demand for capacity as opposed to a demand for performance. We will have 20TB drives by then to satisfy the demands of cloud based storage providers, where performance is not critical.
Joe - from Seagate
- Comment Renewable energy 'simply WON'T WORK': Top Google engineers
- Useless 'computer engineer' Barbie FIRED in three-way fsck row
- Game Theory Dragon Age Inquisition: Our chief weapons are...
- 'How a censorious and moralistic blogger ruined my evening'
- Amazon warming up 'cheapo web video' cannon to SINK Netflix