The problem is that there are only three manufacturers of HDD and they have decided not to compete with each other. Prices have been dropping steadily (price per MB/GB/TB) since they were invented decades ago until three years ago. Since 2020 the price for the exact same type of drive and capacity has been rock solid and newer/larger drives just have higher prices.
There are other explanations than them "deciding not to compete with each other". Have you read about HAMR drives? The reason they're doing that is conventional recording densities have basically hit a wall.
The way HDDs get cheaper is by increasing density, so they can either offer the same capacity using fewer platters or offer more capacity with the same number of platters. It's not by magic. There's no natural law of physics that says everything technological must get cheaper with time. Just look at automobiles: they've pretty consistently gotten more expensive, even though they're based on technology and there's been pretty vibrant competition in that industry.
So, if conventional platter densities stopped increasing and HAMR adds cost and is slow to ramp up, then
the natural expectation should be a plateauing of prices!
Data centers are already gearing up to make the move to NAND. HDD may still cost 50-60% less, but they take up more room, consume more power, and perhaps most importantly, they produce more heat.
Did you ever
look at the idle & active power specs for a datacenter SSDs, or are you just
assuming they match the consumer SSDs you're familiar with?
They are also an order of magnitude slower in performance.
There are lots of cases where performance doesn't matter so much. Backups and near-line storage, for example. Most of the data in "the cloud" is just sitting there, so you don't need to put it on the highest-performance devices.
The net costs are approaching parity.
You're focusing too much on the current market correction in SSDs as representing a trend. It certainly won't continue apace.
By keeping HDD prices higher instead of dropping to maintain a strong value position the industry is dooming itself. It is maintaining some profit margin now despite a huge drop in units sold.
So, if you divide these supposed profits by the volume of HDDs sold, how much do you estimate they could shave off the cost of each drive?
We are already seeing the endgame. 4+ TB NAND devices have been plummetting in price and entering the full consumer market. They have even gotten under $200 for 4TB. This makes sense. Adding more storage, especially in the 2.5" size format, just scales with the price of NAND chips. When manufacturing at scale it is cheaper to make a 20TB 2.5" SATA drive than 20 1TB NVME drives that sell for under $40 each.
Controllers have a limited number of channels, so now you need either more controllers or bigger ones. Adding NAND chips also increases power consumption. Finally, adding more components results in greater chance of failure, which means you need to add some redundancy. Yes, it does scale, but it's not linear.
The best way to scale is to get higher-density NAND chips. Those will come, but at a higher price (though still more cost-effective than today).
At some point, the cost of adding additional NAND cells via denser process nodes or stacking more 3D layers will increase so much that the cost per bit will no longer decrease. I don't know when it will happen, but it inevitably will. At that point, I guess you'll blame "lack of competition" for prices ceasing to fall.
Once higher capacity SSDs leave their specialty niche and go after consumers the HDD market is done.
The lowest cost per GB
is the consumer market. Try taking a look at datacenter SSD prices, some time.
The industry's solution to this is to keep a strong price advantage over NAND, but they just don't want to do it.
No, the industry's solution is to have a foot in both markets. All the big HDD makers also sell SSDs. They compete where there's money to be made. But, they're obviously not going to operate at zero profit or a loss, if they can possibly help it.
Right now, the state of the NAND industry is dire, which is why it's misleading to read too much into current pricing.
Phison remains optimistic, despite short-term NAND industry challenges.
www.tomshardware.com
Facebook
Twitter
Instagram