Why Aren't SSD Prices Going Down?

Lucas123 writes "NAND flash memory makers took an economic beating from 2007 through the first quarter of 2009 due to supply outstripping demand. During that time, solid state drives dropped in price 60% year over year. But after the economic meltdown, fabricators pulled back on production and investment in new facilities and the price of SSDs have remained flat or increased over the past year, and that is not expected to change until 2011. Until that time, SSDs remain 10x more expensive than hard disk drives. SSD vendors, however, are using a few tricks to get sales up, including selling lower-capacity boot drives that hit a sweet spot in the techie/gamer market."

Comments

Price Fixing, Oligopoly, Collusion, Etc.

[eldavojohn]

It's going to be really really hard to convince me that Asian electronics manufacturers aren't engaged in price fixing en masse against the rest of the world whenever a technology cost remains unnaturally high. Hell, after realizing how many times I was the victim of it with LCDs I pretty much expect it.

I mean, really, I feel like a moron for ever knowing that they allowed price fixing -- even promoted it -- inside their borders and then believing that stopped at the rim of the continent. Right now the only question is how many markets is this happening in? They're obviously very good about it, little chance the regulators in other countries will catch it let alone the easily bribed authorities isntalled there.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[TubeSteak]

It's going to be really really hard to convince me that Asian electronics manufacturers aren't engaged in price fixing en masse against the rest of the world whenever a technology cost remains unnaturally high. Hell, after realizing how many times I was the victim of it with LCDs I pretty much expect it.

The /. title is "Why Aren't SSD Prices Going Down?" and the summary quickly provides and answer with "But after the economic meltdown, fabricators pulled back on production and investment in new facilities".

There's a difference between price collusion in a mature market like LCDs versus a lack of capacity in a new market like SSDs.

But congrats on your semi-paranoid stance.
Why look at facts when you can just say "I'll never trust again".

Re:Price Fixing, Oligopoly, Collusion, Etc.

[TheKidWho]

Intel's SSDs are made in the USA with Micron....

Re:Price Fixing, Oligopoly, Collusion, Etc.

[feepness]

I mean, really, I feel like a moron for ever knowing that they allowed price fixing -- even promoted it -- inside their borders and then believing that stopped at the rim of the continent. Right now the only question is how many markets is this happening in?

Yeah! We need them to stop artificially raising prices through fixing so we can artificially raise them through tariffs!

Re:Price Fixing, Oligopoly, Collusion, Etc.

[poetmatt]

your view is quote narrow here. How did you not realize they're doing it with non-SSD hard drives as well? They've been almost the same price for more than 2 years. It's 100% price fixing all around. Yes, densities are getting higher, yes its harder to make, but all these things actually drive the cost down. Meanwhile, 2 years ago? $70/TB. Today? $70/TB. That is no accident.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[d'fim]

Why look at facts when you can just say "I'll never trust again".

Why not both?
Historical facts tend to be more accurate than "gee I hope so" facts.
The past can be a useful tool, because that's where all of the experience is.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Em Emalb]

The old motto is "Trust but verify".

I use "Don't trust, verify, then trust."

Works pretty well for me.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[maxume]

Oh, it's even worse than that. If Newegg prices are a reasonable reflection of the market, Intel SSDs have fallen by quite a bit since October of last year:

http://camelegg.com/product/N82E16820167024

So apparently the real problem is that SSD prices aren't falling fast enough.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Smauler]

Or it could be that _everyone_ wants solid state devices now, and they're difficult to manufacture en mass quickly? There's no need for a global conspiracy theory here, it's just boring old supply and demand.

Note that supply and demand looks a lot like collusion in many cases - When the demand rises, all suppliers automatically will increase prices at about the same time to reflect the market. The best answer is to wait until the product gets cheap, which _will_ happen soon(ish).

Re:Price Fixing, Oligopoly, Collusion, Etc.

[gyrogeerloose]

But congrats on your semi-paranoid stance.

"In related new, tinfoil futures are trading up sharply today due to increased demand by hat makers."

Re:Price Fixing, Oligopoly, Collusion, Etc.

[pla]

There's a difference between price collusion in a mature market like LCDs versus a lack of capacity in a new market like SSDs.

Perhaps you misspoke, perhaps I misunderstood, but doesn't forcing prices down by conspiring to "pull[ed] back on production and investment in new facilities" look very, very similar to merely fixing the prices in a mature market?

They have the capacity, but would rather make more per unit. So they make fewer units to push the supply/demand curve back into more favorable territory. Collusion by any other name would smell as sweet.


Why look at facts when you can just say "I'll never trust again".

You quoted the single most damning part of TFA! The facts in this case explicitly say that they could make more and bring prices crashing down, but have agreed not to solely to keep prices high. I trust, alright - Trust that the semiconductor industry rivals the RIAA in dirty underhanded tricks designed to maximally screw their customers (and even each other, when any one player momentarily gains enough of an upper hand to get away with it) in the name of profit. occasion) for the sake of profit

Re:Price Fixing, Oligopoly, Collusion, Etc.

[hairyfeet]

While I'm sure the geek here at /. will probably have their jaws drop at me saying this, I think there is an even simpler explanation, although price fixing may be adding to the equation. The simple reason is this....most folks don't want them, period.

While the geeks and gamers are probably having heart attacks at saying this, the simple fact is most PCs have gone waaaaay past good enough several years back, and for most have reached ludicrous speed. I offer SSDs as an option on my new builds, and even after explaining the speed benefits don't have any takers, why? Because folks want bigger more than they want faster, that's why. And frankly with 2 Windows 7 PCs sitting here the difference in wake from sleep between SSD and HDDs isn't enough to worry about. My own Windows 7 PC at home wakes from sleep in about 8 seconds from cold to desktop, how much faster do you want?

So I would say it is simply the fact that machines are crazy fast now, and with adequate RAM there simply isn't a need for SSDs unless your a gamer wanting the biggest ePeen. The smallest build I sell ATM is an AMD dual with 3Gb of RAM and Windows 7, and my customers just rave about how fast it is. For the same price as a 32Gb SSD they can get over a Tb of HDD space, and my customers would simply rather have bigger than faster. Plus with Windows 7 all you need is a fast 4-8Gb flash drive for Readyboost and you gain a lot of the SSD speed benefits without the SSD prices, at least in my experience. With games easily coming in at 5-7Gb a piece installed you really need at least 64Gb to see the benefits of SSD anyway.

So I'd say the simple fact is SSD simply isn't needed on the desktop. Mobile is another story, with the non volatile nature of SSDs making them a good choice, but since most of my customers are simply doing the basics on their laptops (word processing, surfing) they really don't need anything bigger than the basic bottom of the line SSDs, which means there isn't the demand driving prices down. Even my hardcore gamer customer decided to go RAID 0 with a couple of Raptors rather than give up space for an SSD. Most folks would just rather have more than faster at this point IMHO.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Killall -9 Bash]

More like everyone wants a 2nd gen SSD, and distributors and retailers want to unload their stock of garbage-ass 1st gen drives. I know what MSRP is on some of these drives, and I can only come to the conclusion that retailers are trying to get consumers to pay for 1st gen shit that will never, EVER sell. Not me. I already paid the early adopter tax ($400 for 16GB), and got burned (drive controller chip has no cache, causing stuttering). A lot of people paid the early adopter tax, and got fucking burned. I'll be waiting for 2nd(or higher) gen drives with 80+GB for $100.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Killall -9 Bash]

I offer SSDs as an option on my new builds, and even after explaining the speed benefits don't have any takers, why? Because folks want bigger more than they want faster, that's why.

I do the same, and 90% of the time, they want SSD. You're not explaining it right.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[ZFox]

Possibly, ya'll are selling to different markets.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[TheKidWho]

That's not how you're currently supposed to use an SSD for gaming...

The optimal way is to load the OS onto a small 16-32gb SSD and then keep the games on a hard drive or RAID Array. Keeping the games on an SSD is silly since the only thing it will effect is loading times.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[ceswiedler]

You're right, most people have a hard time saturating the CPUs on their new computers. But they have a very easy time saturating a physical hard disk. I think that 90% of the time a person is waiting on a computer to do something, it's because the processor is waiting on IO (either reading from the filesystem or from the page file). Making IO faster is by far the best way to make a PC feel faster and more responsive.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[pla]

How did you not realize they're doing it with non-SSD hard drives as well? They've been almost the same price for more than 2 years. It's 100% price fixing all around.

I like a good conspiracy as much as the next guy, but I seriously have to wonder if mere (lack of) market pressure has kept HDD prices from falling over the past year or two.

As you point out, you can get a 1TB drive for $70. A terabyte. At the risk of pulling a "640k", very few typical users would even know how to begin filling 1TB. You have a few academic and corporate users who need that much (or much, much more), but they tend to have far larger budgets and buy more based on superstition than reality; case in point, SCSI still exists and commands a premium - Meanwhile, my home fileserver has 6x the (raw) capacity, gives better performance, and has more redundancy (fully mirrored with live daily snapshots, rather than RAID5 with "backups" in the form of crappy ol' tapes that took all night to create and literally failed about a third of the times we tried to recover something from them), than my last employer's NAS - For about a quarter of the total price. On the other end of the demand curve, you have home users, who have exactly one use for such high-capacity drives - Storing media files (I won't get into the issue of a legit home media server vs piracy). Personally, I own a LOT of CDs (in the thousands), rip them losslessly, and they still weigh in at under half a terabyte. Even ripping entire DVDs, you can fit 100-200 per terabyte.

Compare that with SSDs... Yes, SSDs have some serious advantages (and a few disadvantages) over HDDs. But the average user really does feel the pinch of the affordable sizes (32MB for under $100), and can't afford to blow two grand on a terabyte.

So, personally I don't see so much demand for larger cheap HDDs as to put pressure on manufacturers to lower the prices. At the same time, everyone would love to see larger, cheaper SSDs, so seeing the prices go up as the manufacturers deliberately cut capacity... that I'll call unkosher.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[nabsltd]

My own Windows 7 PC at home wakes from sleep in about 8 seconds from cold to desktop, how much faster do you want?

Since waking from sleep doesn't involve the hard drive very much, it's not really a good indicator of hard drive speed. Waking from hibernation, OTOH, uses the hard drive a lot.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[dekemoose]

Decreasing production in response to decreased demand is a fairly typical business practice, it's just good operations. Likewise, most organizations are going to decline to make major investments on new operational capacity during a down economy (there is a school of thought that says that's the exact right thing to do, taking advantage of lower costs during a recession and preparing yourself for the upshot out of recession but we'll leave that argument for another time). This is not collusion it's intelligent business operations. I know that we as consumers would like all businesses to spew out as much product as they can at the lowest price possible and margins be damned but that simply isn't realistic to always expect in all circumstances. It's a luxury that the tech consumers have largely enjoyed but that doesn't mean that it has to be that way.

Now, if all the makers of SSDs established an agreement between themselves that they would constrain production to a certain level (and I'm not saying this isn't happening) then it's collusion. There's a decent chance it's happening here, just don't automatically equate a business trying to maintain a decent margin on a product to unfair business tactics.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[fizzding]

You don't buy tinfoil hats, man. That's how they get you.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[camperdave]

I'll never understand why people want tinfoil hats. They're just perfect parabolic reflectors for the underground mind ray transmitters. Think about it: where are government strongholds? Underground.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[chronosan]

Forgetting: ", then verify again periodically."

Re:Price Fixing, Oligopoly, Collusion, Etc.

[petermgreen]

hmm, the info on that page only seems to go back to january of this year and it shows a lot of fluctuation but little in the way of overall downward trend.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[gman003]

Is he being paranoid, or not being paranoid enough?

Re:Price Fixing, Oligopoly, Collusion, Etc.

[haruchai]

Yeah, right. It's my non-techie friends who are most impressed by SSDs, mostly because it comes close to the instant-on
that they get from their other electronic devices ( the Blackberry being a notable exception ).
Also, the resistance to shock ( more important in mobiles, true ) is also a wow-factor - most of my friends have kids
and accidents happen.

Your customers are raving about that build of yours because they've been stuck with Windows XP for years and, if they've
not upgraded their machine recently, they probably have a 512 or 1GB box with a dying disk.

You can argue that something may not be needed, and you might be right for a large number of people if you build it
and make it affordable, they will come and find a use for it.
I've heard the "that ain't needed" argument about every piece of tech since 1970 and someone probably said the same thing
about fire and the wheel.

Here's how I see it - faster random access leads to better multitasking which leads to greater productivity.
Also, the best way to speed up a system is to significantly enhance the slowest of the heavily used components / interfaces.
The SSD represent the most significant narrowing of the gap between CPU / Memory / Storage performance in 2 decades.
I say it's way overdue and the pricing on end-user drives can't get to $1/GB fast enough to please me.

And, RAID 0? Sounds good in principle but I wouldn't trust this anymore on my boot drive. After several months of flawless performance on Nvidia FakeRAID, the boot partition up and disappeared without any warnings.
It worked one morning and didn't that evening.
And before you start telling me about getting a real RAID controller, all the ones that have been recommended have been so expensive, I'd be better off spending it on an SSD.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[maxume]

hmm, look at the giant red box on the left hand side of the screen (the one that lists the date with the highest recorded price!).

Re:Price Fixing, Oligopoly, Collusion, Etc.

[petermgreen]

so seeing the prices go up as the manufacturers deliberately cut capacity... that I'll call unkosher.
Not nessacerally when you consider delayed reactions.

Consider each company in the market for widget X (lets assume for the moment that different brands of widget X are interchangable) has a minimum price at which they consider it worthwhile to make widget X. Each customer also has a maximum price they will pay for widget X. This gives us graphs of production VS cost and demand VS cost. Where these graphs cross is the natural price of the product.

However it takes time to react to things. When demand at a given price point drops there are still half-made products, products sitting in warehouses and so on that the manufacturers want to shift. So it takes time for production to adjust downwards. It also takes time for production to adjust upwards when the market price rises. In other words the rate of production is a function not just a function of the current price but also a function of previous prices.

Overshoot is hardly surprising in such a system.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Archangel Michael]

Heck, Maybe this time fire isn't so hot. OUCH! Yup, still hot, maybe next time it won't be so hot.

I'm simply amused by people who are eternally optimistic even in the face of all reason.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[kaputtfurleben]

Or maybe you're not explaining it right? Impossible to tell, really.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[TheKidWho]

That's not true at all, a quick check on newegg:

Intel X25-M 80gb: $225 or $2.8125/gb
Corsair Extreme 64gb: $239 $3.73/gb
Mushkin 64gb: $214 $3.34/gb
Kingston SSDNow V+ 64gb: $205 $3.2/gb
Kingston SSDNow V 64gb: $145 $2.25/gb

Considering the performance and reliability edge you get with Intel compared to the other drives, it's well worth it.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Apocryphos]

mod parent up: insightful

Re:Price Fixing, Oligopoly, Collusion, Etc.

[sunspot55]

I work in the semiconductor industry. I don't deal with memory, I deal with ASIC chips but I can absolutely assure you that the current situation in the semiconductor business is exactly as described. Picture this, you are a manager in a semiconductor company approximately one year ago. You are facing an unprecidented once-in-a-lifetime global economic downturn. You can feel it. The press feels it and reports on it non-stop. Nobody knows when the world will pull out of it but it is obvious to everyone that it hasn't ever, in anyone's lifetime, "been this bad". You know, as a semi company manager, that you have *tremendous* capitol costs (a new factory costs in the billions). What are you going to to? Continue to invest billions in capitol for factory expansion and improvement in the face of unprecedented plummeting demand? Your company's billions are already evaporating in the economy. If you keep spending as you were you would quickly exhaust your company's finances and in addition cause a glut in the global supply of chips (where demand is shrinking) driving down prices like a rock. No, you do what every other company in every other productive sector of the economy does. Cut back production to match demand. The thing is, you can't quickly turn the huge ship of semiconductor production. I'm sure you'll find that semiconductor capacity *always* lags demand. I'm sure in the downturn there was a glut as they realized and as quickly as they could (albeit relatively slowly) responded to market conditions. I can assure you that right now we have the problem that we can not produce enough parts. We are leaving money on the table in the form of unfulfilled demand at the moment. Is that ideal? No, better than the alternative but still a problem. Is there collusion in the industry? Hell if I know but in sum, take it from an industry insider that the factors mentioned in the article are absolutely real and more pronounced than they have ever been in my career.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[jareth-0205]

Well if you find their prices unfair then perhaps as a western country you could manufacture your own SSDs or LCDs? Oh... no... you can't because it would still be much more expensive.

Perhaps we might not like price-fixing, but since manufacturing in the east means the west can take advantage of their far lower cost (and standard) of living so we can have our toys cheaper, I'm willing to cut them a little slack over some price-fixing to (only slightly) redress the balance.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[billcopc]

If the demand is low, then there is no point in producing a large excess of inventory, particularly with something like SSDs where not only is the manufacturing cost relatively high, but the technology itself is still in heavy flux and thus products are quickly rendered obsolete by later revisions.

If OCZ were to stockpile a gazillion Vertex drives today, who's to say one of their competitors won't release a twice-faster SSD tomorrow that leaves all that Vertex stock to rot on the shelves ? We're talking about niche products marketed to high-end users. Generations are measured in months, not years.

Once SSD adoption really takes off, once everyone and their mother has a flash-based boot drive, then we'll see much more flexibility in manufacturing and pricing, but for now it's like complaining that Ferraris are too expensive. The only people complaining are the jealous types who want one but are too cheap to pony up the cash. Sure, the average SSD costs more than most people's entire PC, but these things aren't marketed to Joe Random and his $249 computer. Really, when you look at the kind of people who use SSDs, it is often the least expensive part of the entire system (excluding the ATX chassis). How about I lay it out for you:

High-end motherboard (GA-EX58-UD5) $350
High-end processor (i7 950) $600
12gb DDR3-12800 Ram $350
Two high-end GPUs $1000
1200w power supply $300
60gb SSD $200

So it's really not that big of a pain, price-fixing or not, IMHO.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Yvanhoe]

And the past says that trusting a not-so-friendly country to have all your manufacturing capacities is not a good thing to do...

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Anonymous Coward]

Intel's SSDs are made in the USA with Micron....

And...?

FOR IMMEDIATE RELEASE WEDNESDAY, DECEMBER 17, 2003 WWW.USDOJ.GOV WASHINGTON, D.C. -- An executive for Micron Technology Inc. (Micron) has agreed to plead guilty to obstructing the grand jury investigation of a suspected conspiracy to fix the price of dynamic random access memory (DRAM) products sold in the United States, the Department of Justice today announced.

http://www.fbi.gov/dojpressrel/pressrel03/micron121703.htm

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Anonymous Coward]

You are doing something wrong (or you are a paid actor for some hdd company). The SSD should absolutely be noticeable for everything you do on the machine outside of surfing the web and watching dvds. This page:
http://www.anandtech.com/show/2829/20
pretty much describes the experience of everything I do on my machine[1] compared to the one my wife has in her lab (~$2500 dell). The school wouldn't approve the ssd for her work (don't ask me why, the it people in the chem dept there seem pretty backwards to me). I'd swear my machine is precognitive (if I didn't know any better). Apps start before I hear the click of my mouse button.

You did that gamer a huge disservice by letting him pass on the ssd. An Intel 160GB G2 is absolutely worth every penny when compared with any of today's high end video cards. If you are even thinking about considering a $400 video card compared to a $200 one, you should be investing that extra $200 into the hard drive. Almost every single game currently on the market or advertised in the near future can be played at near maxxed out settings using the video cards that came out 18 months ago. Your ATI 4870 (Q3 2008) or equivalent geforce card (not to familiar with them) can play everything currently on the market. For most games the difference between a $100 card and a $400 card is maxxed out settings vs maxxed out settings. The difference between a Velociraptor and a G2 is almost 10 seconds sitting there waiting for your game to load.

Note that very few people have any use for over 120 GB hard drive space. If they did they would be much better served by a good external backup system anyway (I am certain every one of your customers is concerned about not losing their iTunes music and their photos and videos, all of which should be backed up to external media [2]).

Btw, mine takes about 15 seconds to come out of hibernation or to go into it (12 GB ram). It cold boots in 8. It shuts down in under 5. It is faster for me to just turn it off and then turn it back on than to hibernate it.

1: my current system:
i7 920 @3.2 ghz
12GB ddr3 1600 6x2
Intel G2 160GB ssd
ati 4890x2
win7 ultimate x64

2: my system backup plans
3x 500GB usb hard drives;
1 plugged into my desktop used for general system storage
1 plugged into my print server as backup storage
1 in my fire safe
automatically back up desktop to print server every night
every 6 months (coinciding with my hardware upgrades) I move the one at the print server to the safe, the one in the safe to my desktop and the one at my desktop to the server

I expect to begin replacing the drives at the 3 year cycle (coming up in 2011). I'll probably get 1.5TB drives then.

Is it overkill? Probably. Will I ever allow myself to lose several years worth of photos/music/documents/... again? Absolutely not.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[TheKidWho]

Which is nice, but they are patching BFBC2 and adding a 15 second timer before the game starts this week.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[unother]

Oy. Warmed-over Libertarianism, anyone?

Anyway: I'd like to point out something from the article I think is silly, sort of in the same vein:

"The problem today is prices are prohibitively high for the average consumer," says Gartner analyst Joseph Unsworth. "When you consider a hard drive, you can get a terabyte for about $90. If you look at an SSD -- the Intel one I had with 160GB was $400. The point here is SSDs will never, ever be able to match hard disk drives on price per gigabyte."

Let's leave aside any regards we may have for the opinion of an "analyst". However I find his statement to be especially pernicious because he has completely ignored the scale-of-production factor with cost, i.e. the same thing which was pushing down prices before...

To illustrate my point? Look up the price of any DIMMs of DDR2 or lower speed, and note that prices are higher than when they were at peak production. Why? Well duh--good ol' supply and demand--and where supply exceeds demand, prices drop; and where they do not, prices rise.

My point? In sum: microeconomics is a good explanation of what happens here. At the macroeconomics level, however, the supply-demand fulcrum is overswung by the usual "madness of crowds" et al.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Fred_A]

mod parent up: insightful

Done.

Wait. Damn.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[feepness]

Oy. Warmed-over Libertarianism, anyone?

I just found the juxtaposition amusing.

It seems the only time we're not concerned they're under-charging us is when we're concerned they're over-charging us.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[bertok]

I offer SSDs as an option on my new builds, and even after explaining the speed benefits don't have any takers, why? Because folks want bigger more than they want faster, that's why.

I do the same, and 90% of the time, they want SSD. You're not explaining it right.

Some people sell it as an "either" option: You can either have the speed or the capacity, choose one.

The solution is to simply sell both. An SSD for boot, apps, and speed sensitive data (think Outlook PSTs), and a huge spinning disk for everything else.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[drsmithy]

Keeping the games on an SSD is silly since the only thing it will effect is loading times.

Which is still a lot more than it will affect holding only the OS files (nearly all of which will be cached in RAM shortly after the system has finished booting, assuming you aren't memory starved).

Re:Price Fixing, Oligopoly, Collusion, Etc.

[AHuxley]

Its way off topic but the oil fixing vs tech fixing is fun.
http://exiledonline.com/a-peoples-history-of-koch-industries-how-stalin-funded-the-tea-party-movement/
The idea to create wealth out of been in the middle of any new tech is always a good read.
You would think with quality from SandForce, trim and hardware supply from low wage regions things would be looking good from a software/firmware/hardware point for the end user.
No more controller cartels or pro sumer chip limits.
Just market forces, good PR for a low, medium or hi end unit.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[atamido]

Decreasing production in response to decreased demand is a fairly typical business practice

However in this case the numbers don't really match up. Demand for NAND isn't exactly going away, and the facilities are already there. With the cost remaining high, the logical move is to at least maintain production with current facilities, and look for ways to decrease costs. These types of manufacturing typically increase efficiency the longer they operate, which decreases production costs, and increases margins. If they could maximize their production with current facilities, in a market with high prices, they will still make a killing. And if they are lucky, they would overpower their competitors.

Now, if all of the companies produce at maximum rates, then they might risk flooding the market and killing their margins. But cooperating to keep manufacturing down and prices up is called collusion, which is sort of the point.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Eivind]

Straight on about the superstition among corporate users. I've seen suggestions for $100K backup-systems for 3TB worth of data, when I've pointed out that actually I handle a larger amount of data at -home-, and I've spent about 1% of that cash for doing so, and yes that includes daily backups for the last week, weekly backups for the last month and monthly backups for the last year stored at two physically distinct locations (actually physically distinct continents :D) they are stunned.

$1000 is a lot of disc, in a world where disc is $70/TB.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[metaforest]

Maybe subconsciously you are steering your custies towards HDDs because your margins are better on those than SSD?

Re:Price Fixing, Oligopoly, Collusion, Etc.

[sptbear]

They will find a niche in selling SSD's for booting OS's and Outlook PSTs. With the huge influx of HD movies and DVR recording, use spinning disks for that. Doubtful they will try to compete with the media market the HDD vendors crave.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Phoghat]

+ 1 common sense

Re:Price Fixing, Oligopoly, Collusion, Etc.

[poetmatt]

you're completely off because manufacturing cost has a cost floor, and the minimum price is not even remotely near that cost.

The $1000+ hard drives that exist today realistically don't even cost $100 to make in materials, assembly, marketing and packaging.

The resistance to downward push is just resisting existing market influences. Market demand will ALWAYS push cost towards zero on a product. Basic economics proves that.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Just Some Guy]

Waking from hibernation, OTOH, uses the hard drive a lot.

And even then, mainly in long sequential reads where SSDs aren't vastly faster than HDDs.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[pebs]

So I'd say the simple fact is SSD simply isn't needed on the desktop. Mobile is another story, with the non volatile nature of SSDs making them a good choice, but since most of my customers are simply doing the basics on their laptops (word processing, surfing) they really don't need anything bigger than the basic bottom of the line SSDs

I kinda see it the opposite, but I am not a typical user: on the desktop, you can easily have multiple drives, so having one SSD for the OS and apps, and having one or more HDDs (either internal or external) for general storage would do the trick. On most laptops, you can only have one drive, so better make it a big one.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[navyjeff]

I got a Kingston 64GB SSD for ~$150 from an online retailer about a month ago. SSD prices are already coming down, just a bit more slowly than people would like. Aside from memory, it's the best upgrade I ever bought for my laptop, too.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[ebuck]

... most PCs have gone waaaaay past good enough several years back, and for most have reached ludicrous speed...

I won't be happy till my PC goes plaid!

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Thundersnatch]

Snapshots are not a substitute for off-site backups. Think fire, theft, flood, filesystem or disk controller corruption. Do you rsync those snapshots to an off-site system?

Re:Price Fixing, Oligopoly, Collusion, Etc.

[Thundersnatch]

Since moving to an Intel X-25M in my laptop, I must say I can never go back to a spinning disk for my worksation. It's now so much faster than my work desktop (which has a 2x7200K RAID-0 setup and more RAM and better CPU) that I will probably dump my desktop at the office.

But anyway, the real big-win for SSDs is in the server space. We can't get enough IOPS for some of our databases, which already have 64 GB of RAM as cache. Even some of our web servers that have massive content libraries are disk-bound. Our SAN vendor (LeftHand now HP) has really dropped the ball on this one. We're shopping only because they don't offer SSD cluster nodes yet. They probably don't want to cannibalize sales on their higer-end monolithic SANs I suppose./p

Re:Price Fixing, Oligopoly, Collusion, Etc.

[petermgreen]

RAID-0 pretty much doubles your risk of failure.
Worse than that because as well as the risk of either drive failing you have the risk of something going screwy with the raid implemenation (especially with the cheap fakeraids seen in desktop systems) and breaking the array. Maybe you can rebuild it somehow but it's likely to be rather painful.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[petermgreen]

Agreed, you can get SSDs that are nearly as big as the biggest laptop drives but expect them to cost about as much (sometimes a bit more, sometimes a bit less depending on your taste in laptops) as the rest of the machine put together.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[pla]

Snapshots are not a substitute for off-site backups. Think fire, theft, flood, filesystem or disk controller corruption. Do you rsync those snapshots to an off-site system?

You'll probably never see this because I took so long to reply, but...

Not the snapshots, but by "mirror" I didn't mean RAID-1, I meant a separate machine, in a separate building (garage).

I figure, if both my house and my garage completely burn to the ground at the same time, I'll have a whole lot more to worry about than whether or not I still have my collection of slackware ISOs going back to V1.0.

I did, once upon a time, have "good" offsite backups, in the form of DVDs I'd drop off at my parents house every now and then. But offline storage technology (in a price range suitable for home users) simply has not kept up with drive capacity. I still do that with my truly irreplaceable content (personal records, code I've written, and assorted tidbits I most likely couldn't just re-download or re-rip), but that only accounts for less than one percent of my total disk usage.

Re:Price Fixing, Oligopoly, Collusion, Etc.

[billcopc]

Amen! I am also a convert, I had been holding off for a very long time, but since I bought my Vertex 6-7 months ago I feel like everything else is crap. I still use spinning disks for storage, but for a boot drive I gotta have my solid state.

Because...

[TheKidWho]

SSD cost is limited by the cost to refine and turn Silicon into Flash Memory.

The price will only go down as the process size goes down, currently at 32nm with Intel's Latest drives. Once it reaches 8nm or the like then the cost will truly be comparable to Hard Drives. Until then, don't expect a miracle.

Re:Because...

[Microlith]

At 8nm NAND will be lucky to function properly. MLC will be impossible.

SSDs will reach hard disk prices eventually, but not quite so soon. Rotating media isn't going anywhere for the time being, simply due to the capacity and how cheaply it can be had.

Re:Because...

[MoonBuggy]

I assume that the cost amortisation of the fab plants comes into it somewhat, and presumably other factors too (not least the 'whatever the market will bear' coefficient).

Of course, there is a finite manufacturing cost floor and when you hit that you're only going to improve by altering the technology, but I was under the impression that we're still a decent way off from that point.

Re:Because...

[Firkragg14]

Especially since as we wait for SSD prices to drop and capacities to increase the size and price of rotating media is going to improve at the same time. The way i see it SSDs will eventually become what people use for their os and on portable devices and rotating drives will remain as large mass storage devices.

Re:Because...

[TheKidWho]

A lot of the cost is the raw cost of the processed silicon wafers. Making the pure crystals and then slicing/cutting/polishing them to a workable state takes a lot of time.

Where the processing technology comes in is the ability to make more chips for a given wafer size.

Re:Because...

[timeOday]

The price will only go down as the process size goes down, currently at 32nm with Intel's Latest drives. Once it reaches 8nm or the like then the cost will truly be comparable to Hard Drives. Until then, don't expect a miracle.

Really, that's just re-stating the matter. According to Moore's law, the process size decreases like clockwork, like the sun rising in the east every morning, and end-user pricing decreases correspondingly. Well, it turns out the predictive value of a single-parameter model like Moore's law is pretty weak over the short term. Process size reductions don't just appear like manna from heaven. Market dynamics, including the "great recession" of the last couple years can and do get in the way of progress, even if there's no underlying physical barrier in the way (as we saw when clock speeds topped out).

Re:Because...

[maxume]

To an extent. There are a lot of people who will never fill a 500 GB drive, so once 500 GB SSDs get to their 'reasonable', they won't use spinning media anymore.

Re:Because...

[TheKidWho]

even if there's no underlying physical barrier in the way (as we saw when clock speeds topped out).

Hrmm? There is indeed an underlying physical barrier, you know, the size of the atoms that make up the darn things. It can only get so small.

Moore's law has held up for the past few decades because we've been picking the low hanging fruit, it's going to be really hard to shrink the process size in the near future. Besides the fact that Moore's law states nothing on the processing technology and only on the fact that feature(transistor) count will double every 2 years.

Re:Because...

[Bengie]

SSD cost is limited by the cost to refine and turn Silicon into Flash Memory.

The price will only go down as the process size goes down, currently at 32nm with Intel's Latest drives. Once it reaches 8nm or the like then the cost will truly be comparable to Hard Drives. Until then, don't expect a miracle.

I agree with this. Most of the price drops for SSDs was Flash memory suddenly going from 50 to 40 to 30nm as they quickly caught up with chip sizes. Now that SSDs are caught up with current fabrication processes, the prices will stagnate until a new better fabrication comes out.

There will obviously be price drops as the cost of initial investments are covered and when a slightly newer more competitive model comes out and they need to clear old-stock. But overall, they're like CPUs now. Expensive when they first come out, drop sharply shortly after and level out with a minor decline until the next big thing.

Re:Because...

[hypergreatthing]

o'rlly? So, the cost of silicon is so astronomically high that a inch by inch wafer costs several hundred to manufacture. Only if we could get it down to a .1 inch by .1 inch our problems would be solved?

I sincerely doubt that the cost is proportional to the amount of silicon used. I'm sure the materials play a factor, but lets not kid ourselves here. And i'm also positive that retooling the fabs is a major cost that has to be recouped.

Re:Because...

[TheKidWho]

SSD cost is limited by the cost to refine and turn Silicon into Flash Memory.

Yeah, because turning that silicon into flash memory has nothing to do with tooling costs. That's the magical fairy step.

Re:Because...

[timeOday]

Hrmm? There is indeed an underlying physical barrier, you know, the size of the atoms that make up the darn things. It can only get so small.

And yet, if the article is to be believed, that is not the cause of the current stagnation in SSD pricing - rather, it's market dynamics. Of course physical limits exist, one might even suspect them of dooming SSD's to forever lag hard drives in capacity, but the article says on the contrary recent pricing is just due to market dynamics.

Moore's law states nothing on the processing technology and only on the fact that feature(transistor) count will double every 2 years.

In other words, it's actually more applicable to memory than it is to processors. Memory is direct a function of transistor density as you can get.

Re:Because...

[Firethorn]

For that matter, there's lots of people who won't fill a 50GB drive.

Still, a 80GB HD is like $35-50, a 64GB SSD is $130+.

SSDs need to halve in price a couple more times to get them economical.

Re:Because...

[LWATCDR]

SSDs will take over for "smaller" HDs first.
A standard hard drive has a fixed low end price. It costs x to make any hard drive no matter how small the capacity. That is why an 80 gig hard drive is a rotten deal compared to say a 250 or 500. Not being an expert I would guess that is is somewhere around $40 dollars.
Right now on newegg you can get an 80GB sata laptop drive for $40, and 120GB for $45, and a 250 for $47.
It really doesn't make any sense to pick anything smaller than a $250 from the end users point of view but if you are selling to a price for say a netbook and your price point is $299 then the difference can matter.
But if you can get a 40GB SSD for $30 or 120 $40 then it makes sense.
The minimum price for an SSD is lower than for a flash drive.
I think you will see SSDs sell big in the very fast segment and in the small segment first. A standard USB flash drive is nothing but a slow SSD.

Re:Because...

[Zencyde]

I wish I could mod you down.

Re:Because...

[TheKidWho]

Don't worry, it's been done before.

Re:Because...

[Dogtanian]

According to Moore's law the process size decreases like clockwork [etc]

That last part borders on putting words in Moore's mouth and holding him to a higher standard than he ever claimed originally.

In fact, you're generally reading way more into Moore's "Law"'s predictive power than was ever likely originally intended. It's not, and never was a law, merely an observation and a prediction that progress at the then-current rate would continue for "at least 10 years", i.e. until 1975.

The fact that it's broadly held up for 35 years longer than that is pretty astounding.

Re:Because...

[pwnies]

At 8nm NAND will be lucky to function properly. MLC will be impossible.

Can I ask why? I'm not debating it, i'm just curious.

Re:Because...

[drinkypoo]

SSDs will take over for "smaller" HDs first.

Uh, they already have. Netbooks with 2 and 4GB flash disks prove this. Unfortunately, they also demonstrate that there is a gap between hard disks and SSDs, there is no point at which they yet overlap. Thus, you have 2GB SSD, 4GB SSD, 8GB SSD... 30 GB HD, etc. So when the average computer has 2TB HD, perhaps it will be offered with 64GB SSD for the same price... except I think it's more like 4 or 8TB HD matching up to the puny SSD.

Re:Because...

[petermgreen]

Netbooks with 2 and 4GB flash disks prove this.
Notice that those (and the slightly later models with about 20GB of flash and a screen that actually filled the space available) have mostly dissapeared replaced by slightly larger ones with 80GB-160GB HDDs? Performance (particulally write performance) on those things was shit and the capacity limit was pretty crippling.

One thing to remember is that individual flash chips aren't particulally fast but by combining lots of them in paralell with a good controller high performance can be acheived. OTOH devices like USB sticks with a cheap controller and a handfull of flash chips tend to be slower than HDDs (especially when writing). This along with mounting requirements means that the price floor for SSDs that actually perform better than hard drives and can be fitted in place of a traditional hard drive will be higher than USB sticks, CF cards etc.

Re:Because...

[petermgreen]

Right now on newegg you can get an 80GB sata laptop drive for $40, and 120GB for $45, and a 250 for $47.

There seems to be a similar effect with SSDs while there are 8 and 16GB drives on the market they cost as much (sometimes more) as the cheap 30GB ones (admittedly the smaller drives are SLC but the smaller number of chips and older controllers means this doesn't bring a performance advantage afaict). The sweet spot price for SSDs seems to be arround the 60-120GB mark.

Yeah, why hasn't the price gone down?

[BadAnalogyGuy]

fabricators pulled back on production and investment in new facilities

Joseph Smith once said that a man with one wife was blessed, but a man with more than one was cursed. I guess he meant that as the supply goes down, the more profit can be realized per unit.

Re:Yeah, why hasn't the price gone down?

[mswhippingboy]

I guess he meant that as the supply goes down, the more profit can be realized per unit.

Or maybe he just didn't count on m = (b + g - x) ^ n (where m= misery factor, b=bitching factor, g = carryoutthegarbage factor, x = (take a guess) and n = number of wives).

Give it time

[uncledrax]

Give it time.. the $/GB price will go down.. SSDs are only a few years old and they are still working on achieving density comparable to spinning-disks, instead of focusing on cost reduction.

If you're in a tizzy because SSDs are expensive, then continue to use tried-and-true conventional disk until they meet your price point.

I'm sure there are other considerations.. prob some monopoly on NAND manufacturing or something.. but that'll eventually sort itself out and cost will go down. Not like you can open a new NAND fab overnight or anything.

Re:Give it time

[nschubach]

and that is not expected to change until 2011.

But 2011 is so far away! OMG!

Maybe I'm getting old, but 2011 doesn't seem like an inordinate amount of time to wait. It's roughly 8-20 months to wait. While it's not going to enthuse those looking to upgrade this year, it's not like the world will end if SSDs don't take off in a month.

They'll go down eventually

[mr_flea]

They'll go down eventually if you give it time. SSDs are now just getting popular. Larger LCDs are finally affordable now, and how many years did that take? They just need more time to get the manufacturing procedure and the like down. I'm sure advances in SSD manufacturing will bring them down in price eventually. Just be patient.

Re:They'll go down eventually

[rm999]

I think how long it will take is a valid question; the impact on the industry will be *huge* when these things become mainstream.

Part of the problem is that unlike LCDs, SSDs are hidden from view. Average Joe doesn't demand it because he doesn't know the impact it will have on him.

Re:They'll go down eventually

[jedidiah]

> Part of the problem is that unlike LCDs, SSDs are hidden from view. Average Joe doesn't demand it because he doesn't know the impact it will have on him.

Another part of the problem is the fact that Average Joe won't notice and won't care.

Re:They'll go down eventually

[QuantumRiff]

Well, the LCD screens really dropped in price, right after a bunch of manufacturers were busted for Price Fixing their products. So I'm thinking you're not too far off...

They have...

[MistrBlank]

The same 60GB drive I paid $230 for 6 months ago is now $130 after rebates and $160 before.

Re:They have...

[b0bby]

But flash in general seems to be holding steady, or even going up a bit. The same 16bg SDHC card I bought a year ago from Newegg for $30 is right now around $34. That could be explained by a combination of reduced capacity & larger SSDs taking up more of the reduced production.

Question

[Anonymous Coward]

I know this topic is about SSDs, but I remember back in the day we had full height 5.25" drives that sounded like jet engines and had several platters. Why hasn't anyone made bigger platters- are we really constrained to the 3.5" form factor? I'd think they could make big platters with some extra ECC, have several platters, or even have internal platter mirroring or something l like hardware raid6 at the platter level?

Re:Question

[HFXPro]

A 5.25" drive would have significantly higher stresses placed upon the platter. There would also be more area to have to physically move the head across.

Re:Question

[powerlord]

I know this topic is about SSDs, but I remember back in the day we had full height 5.25" drives that sounded like jet engines and had several platters. Why hasn't anyone made bigger platters- are we really constrained to the 3.5" form factor? I'd think they could make big platters with some extra ECC, have several platters, or even have internal platter mirroring or something l like hardware raid6 at the platter level?

Putting RAID inside the drive doesn't buy as much (from a Redundancy perspective) as putting multiple drives into the same space does.

With multiple drives you can:
- use different manufacturers
- replace each independently

With RAID in a drive, you are probably using the same circuitry for those whole enclosure (comparable to the Disk Controller now), and you do not have the option of replacing a failed platter (without a new HDD architecture of some sort).

Add to that the limited number of environments willing to spend the extra money on RAID, and the non-techie obsession with "smaller is better tech" being seen in NetBooks, and NetTops, leaving less room for this sort of thing, and you're left with a limited market that could alternatively be served by existing RAID controllers/setups, using "standardized" components (3.5" drives), that can be mass produced due to the large demand for them in non-RAID applications.

Re:Question

[kingofwaldos]

I had a friend that had a 5.25" drive about 10 years ago and it was absolute trash. I think Maxtor made it. Everything that is good about a SSD is terrible about a drive with huge platters like that. First, you can't spin them very fast -- his was 3600 rpm. Second, slow rotation + long distances traveled by the read/write arm = horrific latency. His computer took forever to boot up and word might take a minute to start. Be glad they're gone.

Re:Question

[TheKidWho]

When you have something spinning, the smaller the better.

Let's do some simple Math:

For a 7200RPM Hard Drive@ 3.5" Diameter or .0889 meters you have a velocity of
V = Pi*D*RPM = 3.14*.0889*7200 = 2009m/min or 33.5m/s.
Now the centripetal acceleration:
a = v^2/r = (33.5m/s)^2/.04445m = 25243m/s^2.

Or in other words: 2573g

Yeah, that's a hell of a lot of acceleration at the outer edges. The smaller the better.

Re:Question

[compro01]

Larger platters means more centrifugal force. Too much force and the magnetic emulsion starts to come off the platters, so you either need to find a better way to make the emulsion stick (which does not appear forthcoming. They have to shrink the platters for higher speed 10k and 15k drives) or run the drive slower, which will clobber performance.

Re:Question

[obarthelemy]

my guess is that several issues scale exponentially with diameter:
- spin up torque and strain on the motor
- precision required for heads placement
- vibrations
- vertical tolerances (head positionning, platter warp...)
Solving those is probably too complicated and expensive to make sense when most people aren't even buying today's top-ine 3.5", 2TB drives.

Internal raid and platter mirroring doesn't make much sense either: eveything remains dependent on a single motor and head mechanism: waht's left must not be a very large proportion of HD dysfunctions.

Re:Question

[karnal]

Throughput isn't the only thing that makes a desktop feel slow. It's the random access. Given your scenario, putting a larger drive (i.e. more dense) but spinning it slower could result in similar if not better sequential transfer times. However, you go to seek and if the data is 1/2 or more way around the other side of the platter relative to the current head position, you're screwed.

Re:Question

[petermgreen]

but couldn't the 5.25" get away with a slower rotational speed end up having the same angular velocity as a 3.5" at the outer parts of the disk?
The angular velocity is by definition the same across the disk. The linear velocity increases the further you move from the center (assuming a constant motor speed). So a larger disk can in principle get higher transfer rates from the same spin speed.

However transfer rate is not the dominant issue in many applications. Total request time is made up of

1: time to move to the right track (seek time)
2: time to wait for the sector to come under the head (rotational latency)
3: time to actually read the data

For the small reads that are common in many desktop applications 1 and 2 dominate on a hard disk. Making the disk bigger will increase the time to move to the right track while slowing the rotation speed (needed to physically realise a larger platter) will increase the time waiting for the right sector to come under the head.

Re:Question

[juuri]

I like your thinking, just how much data could we cram in one of those old full height scsi drive bays!

Re:Question

[sirsnork]

You speak of the Quantum Bigfoot, the last 5.25 drive and at the time the cheapest drive you could buy. Of course there was a reason for that

Re:Question

[cpghost]

Not just HDDs got smaller and acquired higher capacity and reliability. FDDs also went from 8" to 5 1/4" to 3 1/2"... until they were all but replaced by solid state devices (USB drives), of much higher capacity and reliability. However, some old computer tech is superior to what we have today for some particular applications: think computers used in space travel (satellites): the less densely packed and the slower the electronics are, the less vulnerable they are to cosmic rays and other interferences. Same for EMP-hardened computers used in some military aircraft and tanks. It's all a matter of choice and there are some trade offs that application designers have to take.

Re:Question

[Amouth]

It was a Quantum Bigfoot

http://en.wikipedia.org/wiki/Quantum_Bigfoot_(hard_drive)

Ranged from 2-20gb over it's life (2-6 for the 3600rpm)

although i will say they don't make drives like they used to.. i currently have an old Linux box running using a Seagate 50pin Wide 9.1 SCSI drive that is 5.25 wide monster.. the damn thing has a strip thermometer on the front so you can see if it's over heating.. but you know.. it still works

here is a good picture (not mine but of the same drive model)

http://mail.lipsia.de/~enigma/neu/pics/st410800n_overview.jpg

 

Re:Question

[aztracker1]

I seem to remember a drive like that, Quantum Bigfoot iirc. Thing was dog slow, and painful to use, also had a higher rate of failure, compared to similarly priced, similar capacity 3.5" drives.. Now, a drive caddy in a 5.25" bay that fit 4-6 2.5" drives could be worthwhile though.

Re:Question

[Capt.DrumkenBum]

It was done.
http://en.wikipedia.org/wiki/Quantum_Bigfoot_(hard_drive)
They were crap.

Re:Question

[twistedcubic]

Putting RAID inside the drive doesn't buy as much (from a Redundancy perspective) as putting multiple drives into the same space does.
But it sure does kill roaches..., and that's something, isn't it?

Re:Question

[drinkypoo]

although i will say they don't make drives like they used to.. i currently have an old Linux box running using a Seagate 50pin Wide 9.1 SCSI drive that is 5.25 wide monster.. the damn thing has a strip thermometer on the front so you can see if it's over heating.. but you know.. it still works

Ah yes, the classic barracuda. You NEED a thermometer on it to see if it's overheating, those were literally the hottest 3.5" hard drives ever made. They were amazing speed demons for the time when they were released; I installed a couple then. I had to put them into the space of a full-height disk to keep them from going nuclear even in well-ventilated disk enclosures. From my standpoint, that disk was a massive failure, and they put me off Seagate for years. Of course, I grew up with Seagate; they in Scotts Valley, me in Santa Cruz. When I was a kid I was jumping up and down excited when used MFM disks hit a buck a megabyte, which they did there back when I had my Amiga 2000... with an MFM controller. Boot from floppy, switch to hard. Those were the days. Pulling your ST225 out of the box to whack it with a screwdriver once a week, oh, the memories. In fact, we called them Seizegate back then...

Gartner is wrong

[wonkavader]

OK, I think this bit from Gartner

Garnter analyst Joseph Unsworth [says] "...The point here is SSDs will never, ever be able to match hard disk drives on price per gigabyte."

is wrong. Flash is simpler than drives. The manufacturing requires less machining, materials, and human labor. I'm not saying next year or even five years out, but as SOME POINT, I am sure that memory devices like flash will be cheaper than disk drives on a per bit basis, or at least close enough that innovation on spinning drives will stop and that will allow flash/memory devices to pass them.

Anyone agree/disagree?

Re:Gartner is wrong

[daveime]

Bearing in mind that perhaps 2 years ago, a 256MB USB Keydisk cost me $10, and today, I can get a 16GB for the same price, yes, I'd say it's just a matter of time.

Re:Gartner is wrong

[nschubach]

Agree, but for no more reason that you state. Eventually we won't be able to tweak bits on a metal platter easier than on silicon, nano-tubes or whatever else seems to be coming down the line. I think it's more a lack of patience that this story was posted.

Re:Gartner is wrong

[Antimatter3009]

To put it simply, never is a long time. Saying "never" in regards to technology is always a stupid thing to do.

Re:Gartner is wrong

[diskofish]

I neither agree or disagree.

Re:Gartner is wrong

[Tridus]

Wouldn't it be more newsworthy if Gartner was right?

They're generally the most reliable source: bet on the opposite of whatever Gartner is saying.

Re:Gartner is wrong

[thijsh]

I respectfully disagree, even though your theory is sound. You fail to account for the likely possibility that before we reach the point where SSD is cheaper than HDD in practice we will have moved on to the next breakthrough technology (holographic, or even molecular or genetic storage). Especially when new breakthroughs like optical and quantum computing change the computing world both SSD and HDD will become obsolete around the same time.

Re:Gartner is wrong

[jedidiah]

Yeah... but that 16GB USB drive still doesn't have enough room for a single HD movie on it.

Meanwhile, I can get 2TB 3.5" spinny disks and a 250GB 2.5" spinny disk is $60.

It will still be a bit before I can cheaply turn my EEE PC 900 into an Archos 5 knockoff.

Re:Gartner is wrong

[fuzzyfuzzyfungus]

Whether he is right or wrong really depends on how large a disk you want.

Barring truly revolutionary advances in silicon device fabrication, and(I'm not sufficiently up on my physics to know for sure) possibly a change in physics, sputtering a thin metallic film with the appropriate magnetic properties will always be cheaper, per square centimeter, than fabbing a complex integrated circuit. Further, it is quite likely that the smallest possible magnetic domains will continue to be smaller than the smallest possible flash cells, so you get more bits per square centimeter, and you pay less per square centimeter with the magnetic stuff.

However, as you note, Flash is pretty much ready-to-go. Virtually all the cost is the silicon. Packaging and soldering are relatively cheap(and, since every HDD also has a controller board, both technologies pay the "assemble a PCB" cost). With magnetic storage, though, whether HDD or tape, you have to build a fairly complex and expensive machine to enclose the cheap magnetic medium, and read/write, and keep dust off, and so forth.

If you adopt the naive strategy of comparing each technology's "sweet spot cost"(ie. the cost/GB of the device with the lowest cost/GB of each tech), I suspect that Mr. Unsworth is correct, if not forever, at least for a long time. However, a great many applications don't actually care about that metric. If you know how many GB you need, you don't care about "what is the lowest cost/GB?" you care about "how can I most cheaply get X GB?"

In the case of HDDs, the "sweet spot price" is somewhere between 5 and 10 cents a GB. However, the sweet spot is measured in 1-2 TB devices. If you, say, only needed 20GB, you would be unable to find anybody to sell you a 20GB drive for $1-$2. A quick look a newegg suggests that the cheapest retail HDDs are around $30-$35. You do get 80GB to 160GB for your $35; but you basically can't spend any less. The cost of a machined housing, hiqh quality spindle motor, packing, shipping, etc. just make that impossible. For the same $30-$35, retail, you are looking at around 16GB of flash(less if you want AES encryption and stuff, a little more if it is on sale). Thus, for any application that needs 16GB or less, SSDs are, in absolute $/GB terms, actually cheaper than HDDs(in addition to their other virtues: quiet, low power, shock resistant, small size, etc.)

I suspect that, over time, HDDs will be cheaper than SSDs in "sweet spot price" more or less forever; but the capacity(currently around 16GB, was more like 8GB the last time I wrote something like this) below which the absolute cost advantage lies with SSDs will continue to creep up. If it manages to creep up faster than software bloats, we may reach the dramatic tipping point where an SSD is cheaper, as well as better, than an HDD for the boot volume of a "normal computer", as opposed to just embedded systems, the occasional netbook, and space/power constrained devices.

Re:Gartner is wrong

[Lumpy]

I have 3 HD movies on a 16 gig thumb drive here. They all look fantastic on a 42" set. Mpeg4 compression done the right way from a clean source works GREAT. Oh and I downconverted the audio to 5.1 AC3 audio... I have yet to find a person that can hear and see the difference from the original and the ripped version. Quantum of Solace, Transformers and Watchmen are what I used for testing ripped directly from bluray discs.

Re:Gartner is wrong

[imsabbel]

Gartner is most likely right.

Your main error is that you assume flash is simpler than HDs. The opposite is true. You cannot even imagine how complicated 32nm semiconductor fabrication is. And every step down will be worse. If it were easy, it wouldnt cost billions to get a manufacturing plant up. And they would not need Phds for process control.

  Storage cells on a HD surface are not that much bigger than flash bits on a chip. Maybe an order of magnitude. On the other hand, flash requires a HUGE amount of processing to make the chips, and scaling the structure size down will only increase the investment cost required for fabs. Yes, they will more than break even, but 15nm will not be 25% of the price of 30nm.

Also, scaling is limited for flash. There will never be a point were a 3nm process gets shrunk to a 2nm process. While it is true that HDs are closer to the scaling brick wall, currently the price difference is still about a factor of 40. And HDs wont stop dropping in price.

The only way I see FLASH ever getting cheaper than HDs is that if at some point of time it will be cheap enough for most people.

- In one scenario, the $20-30 mechanical parts base cost of an HD could make the difference. A 2TB drive might be not much more expensive than a 300GB, but if you only need the 300GB, flash might be cheaper.

- Similar, if production rates of HDs go down because of flash being "cheap enough" for most people, economy of scale can drop. This may end in a scenario where HDs are still cheaper per GB if you need 500TB of storage, but you will only find them in media servers or in datacenters, similar to streamer tapes in the past.

The main point is that you can get huge densities in flash, you can stack dies many times beacuse of the low power, it should be no problem to fit 10TB in the volume of a 3.5" drive even today, etc.

But in the end, for every GB, a certain size of silicon has to be processed in some of the most expensive and complicated production processes the world has ever seen. While for HDs, (simplified) you just do some vacuum sputtering.
They WILL stay cheaper.
But maybe at some point they will die out because they are big, sensitive and slow.

Re:Gartner is wrong

[JustinOpinion]

Barring truly revolutionary advances in silicon device fabrication, and(I'm not sufficiently up on my physics to know for sure) possibly a change in physics, sputtering a thin metallic film with the appropriate magnetic properties will always be cheaper, per square centimeter, than fabbing a complex integrated circuit. Further, it is quite likely that the smallest possible magnetic domains will continue to be smaller than the smallest possible flash cells, so you get more bits per square centimeter, and you pay less per square centimeter with the magnetic stuff.

Apparently sputtering of magnetic films is in fact approaching a limit. As the bit density increases, the magnetic domains in the film need to keep getting smaller. The smaller domains become more and more susceptible to being randomly flipped (thermally or due to interactions between neighboring magnetic domains). Currently the envisioned solution is to create small isolated magnetic domains: basically magnetic nano-islands on a non-magnetic platter. This presents a whole bunch of new problems (e.g. getting the read/write head to repeatably track to such small locations).

But this proposed 'bit patterned magnetic media' obviously increases the cost of HDD fabrication since you're no longer just sputtering a magnetic film; you have to pattern the disk platter (albeit with a relatively simple pattern compared to a microchip or even flash storage).

This barrier is "close enough" that HDD companies are seriously researching how to make bit-patterned drives. (Hitachi is the company that I know for sure is working on it; no doubt others are, too.) Of course it's always possible that someone makes a discovery in magnetic thin films that lets us keep using simple layers for hard drives platters... (It's always difficult to predict such things!) ... but currently it looks like hard drives are going to become patterned in the future.

If this happens, then it will close the gap, in a sense. With hard disks giving up one of their manufacturing advantages in order to push to greater bit densities, SSD will probably catch up and overtake.

Re:Gartner is wrong

[petermgreen]

In the case of HDDs, the "sweet spot price" is somewhere between 5 and 10 cents a GB. However, the sweet spot is measured in 1-2 TB devices. If you, say, only needed 20GB, you would be unable to find anybody to sell you a 20GB drive for $1-$2. A quick look a newegg suggests that the cheapest retail HDDs are around $30-$35. You do get 80GB to 160GB for your $35; but you basically can't spend any less. The cost of a machined housing, hiqh quality spindle motor, packing, shipping, etc. just make that impossible. For the same $30-$35, retail, you are looking at around 16GB of flash
Sort of

You can get a USB stick for that price but USB sticks are notdesigned to be mounted in a PC case (you could probablly tie one in and link it to an internal port somehow but I would consider it a bodge), usually not optimised for speed or high write cycle use and the USB interface and the fact they show up as removable drives will cause complications with using them as a boot drive (especially if you are a windows user).

Compactflash cards can also be got in that price range but you need an adaptor to connect them up and mount them and then since those adaptors are usually designed for laptop or small form factor use you will probablly need an adaptor to mount the adaptor driving up the overall cost. Once again they also don't tend to be optimised for speed and high write cycle use like proper SSDs.

Proper SSDs with SATA interfaces and a standard drive form factor seem to start at arround $90 (slightly less if you count after rebate prices) for 30GB (there are 16GB ones on the market but they cost MORE than the 32GB ones) . Further the form factor is a laptop one meaning you may also need to buy an adaptor plate (some but not all SSDs come with this in the box afaict) which costs another $15.

You also talk about the "boot volume" as if it's a dedicated drive. There are probablly some enterprise conditions where that is the case but afaict for most desktops the same drive serves as boot drive and data drive. Granted some people don't have much data though.

Re:Gartner is wrong

[Just Some Guy]

Barring truly revolutionary advances in silicon device fabrication, and(I'm not sufficiently up on my physics to know for sure) possibly a change in physics, sputtering a thin metallic film with the appropriate magnetic properties will always be cheaper, per square centimeter, than fabbing a complex integrated circuit.

But what will the price difference be per cub centimeter? As chips (eventually) transition to stacked designs, I'd be willing to bet that you can stack a lot more layers of silicon per unit height than you can spinning platters that allow enough room for air flow, head thickness, etc.

Re:Gartner is wrong

[Just Some Guy]

That was supposed to be cubic. You get the idea. :-)

Re:Gartner is wrong

[fuzzyfuzzyfungus]

Again, it depends on exactly what you want: If you are significantly space constrained, SSDs may well be your only option(HDDs are either simply impossible, or you would need an ultra-thin variant custom designed and produced just for you, which, with lousy economies of scale, could mean paying hundreds or thousands of dollars per unit).

However, if you are not significantly space constrained, die stacking doesn't do much to change the cost of SSDs(packaging costs go up; because the die bonding gets harder and more complex, PCB costs go down because you can get away with smaller boards and fewer layers and traces). The cost per square centimeter of functioning silicon, which is the bulk of flash's cost, stays the same.

Already, flash has slaughtered the more space-constrained HDDs(the CF-card sized Microdrives are basically extinct. 1.8 inch HDDs are an endangered species). However, at the 2.5 inch size, the cost of silicon really starts to hold them back. Even by the standards of today's die packaging and PCB tech, a lot of 2.5 inch SSDs are shipping "half-full" or less; because it would be too expensive to purchase the extra flash.

With either contemporary, or plausible near future, packaging, SSDs can definitely win the density battle; but that only changes the cost game in very density constrained situations.

Re:Gartner is wrong

[eharvill]

OT, but what software do you recommend for ripping your Blu-Rays? I just recently made the upgrade to BD and want to start filling up my NAS. A 4 year old and CD/DVD/BD are not a good combo.

Re:Gartner is wrong

[aztracker1]

Probably AnyDVD HD for the ripping.. as to the re-encode, I've been using Nero Recode for my DVD re-encodes to mp4+avc (h.264), but haven't made the plunge to BluRay, so not sure now Nero does there. If you are running linux/mac would just stick with VLC + Handbrake. I find the AnyDVD + Recode is a bit more reliable though. The only issue I have is making sure the cropping results in a final resolution divisible by 16 (otherwise you get a weird yellow line at the bottom).

Re:Gartner is wrong

[daveime]

I'll reply to myself, as everyone seemed to have missed the point of the parent's question, and my response.

I was merely saying that in the space of 2 years, the capacity of a typical keydisk has gone up by a factor of 64 while the price has remained constant.

Assuming that same rule for the next 2 years, and we end up with a 1TB keydisk for $10 in about 2012.

Re:Gartner is wrong

[mjwx]

is wrong. Flash is simpler than drives. The manufacturing requires less machining, materials, and human labor. I'm not saying next year or even five years out

The construction is more complex and requires more skilled labour. Right now in a Western Digital factory in Issan, Thailand young Thai's are assembling mechanical hard drives as an alternative to subsistence farming. Netting approximately 1000 Baht (US$35) a week, well paid for an Issan Thai. How much would a single worker in one of Intel's flash memory fabs cost.

The cost of the human labour will not reach parity for some time, neither will the machines required for assembly. What will eventually drive the price down is:
1. increased demand, leading to increased production.
2. this is where economies of scale kick in, even if the price to turn sand into silicon does not decrease, profit margins can be shaved as more is being moved.
3. Increased competition also from increased demand.

In fact I don't see SSD's reaching per GB parity with mechanical hard drives until we can no longer make larger mechanical hard drives. Where SSD's win is in performance, faster with less power draw, even at thrice the per GB cost they become a bargain for the performance gain. Right now I can buy a 1 TB HDD for A$110, that's A$0.11 a GB while a cheap SSD costs A$219 for 64 GB which is A$3.42 per GB (A-RAM brand, and Intel 80 GB is A$400). Even once we start paying A$1 per GB SSD's become worth it.

Re:Gartner is wrong

[metaforest]

It won't be flash. It will likely be a PCM or some variant of QF that does it. A film coating over lower density CMOS gates is simpler and potentially more reliable than Isolated Gate approaches. In much the same way that DRAM variants have relegated SRAM to specialized applications, FLASH is likely going to get replaced before too long. Decreasing the cell foot-print by using a more space efficient cell is usually more effective at reducing cost per bit than decreasing feature size.

Rethinking storage and memory tech is cheaper(in the long run) than shrinking feature size on an existing design.

By the same reasoning ARM is likely to hand Intel it's ass in a few more generations, on several critical specs.

Re:Gartner is wrong

[drinkypoo]

You can get a USB stick for that price but USB sticks are notdesigned to be mounted in a PC case (you could probablly tie one in and link it to an internal port somehow but I would consider it a bodge),

If you add a $5 USB card, you can get an internal USB connector, and plug in there. I've done this for one system I have that runs entirely from 4GB flash.

usually not optimised for speed or high write cycle use

Get a real stick, like an OCZ Rally2 or Diesel. Or whatever the new hotness is now. I get HD-comparable speeds using one. Nothing screaming, but the lack of seek time doesn't hurt, and in the balance it's a little faster than a cheapass 5400 rpm disk.

and the fact they show up as removable drives will cause complications with using them as a boot drive (especially if you are a windows user).

Only if you are a Windows user. I don't know how they're handled on the Mac — on a system with Firewire boot, I'd hope they'd work as well as any other volume — but for free operating systems which use grub, this is basically now a non-issue. Current versions of grub allow you to specify a UUID as the groot, which eliminates the last booting problems that used to drive people to still use syslinux (ugh.)

Compactflash cards can also be got in that price range but you need an adaptor to connect them up and mount them and then since those adaptors are usually designed for laptop or small form factor use you will probablly need an adaptor to mount the adaptor driving up the overall cost.

I usually just mount them with double-sided foam tape. I use the 3M stuff, it lasts in an environment like a PC case.

Once again they also don't tend to be optimised for speed and high write cycle use like proper SSDs.

Write cycles depend on the type of flash. Speed? If you pick which card you get, you can get CF with an ATA UDMA4 interface. I have an 8GB card like that in my DT360. Not as good as a PCI-E interface, though most PCI-E SSDs have been crap; also not as good as a SATA interface, but you can't have everything.

Proper SSDs with SATA interfaces and a standard drive form factor seem to start at arround $90 (slightly less if you count after rebate prices) for 30GB (there are 16GB ones on the market but they cost MORE than the 32GB ones) . Further the form factor is a laptop one meaning you may also need to buy an adaptor plate (some but not all SSDs come with this in the box afaict) which costs another $15.

You can get adaptor rails for under $5.

You also talk about the "boot volume" as if it's a dedicated drive. There are probablly some enterprise conditions where that is the case but afaict for most desktops the same drive serves as boot drive and data drive. Granted some people don't have much data though.

What's the cutoff, anyway? You end up with applications (or at least parts thereof) on the boot drive in practice because pretty much every operating system demands it... well, except Unix :)

Re:Gartner is wrong

[squiggleslash]

16GB is plenty of room for a single, Blu-ray or HD DVD quality, HD movie of average length, and easily more than enough for an ATSC quality movie, even using MPEG-2. And if you want to go for "HD download" quality of the Apple TV variety (720p H.264) then all bets are off.

Remember that BD movies generally only weigh in at 20GB+ because of the extra overheads of having to carry a pseudo-lossless soundtrack or two, plus several other languages, plus a possible extra PIP video, etc. The actual "video" portion of a Blu-ray (or HD DVD) release tends to be in the 5-7GB/hour range.

The myth it requires some huge amount of storage dates back to the original BD vs HD DVD wars in which fans of the former kept stating as fact that you had to have some ungodly amount of storage to store the video that just happened to be slightly more than a single HD DVD layer. You don't. The original reason BD went for 5X the storage capacity of DVD was that it expected to have to store 5X as much data - originally BD was slated to use, and only use, MPEG2. The success of H.264 turned that on its head, with much higher quality video being stored at a much lower bit-rate. As if to add insult to injury, even MPEG-2 circa 2005-2007 is not MPEG-2 circa 2010, with ATSC's adoption of the format for HD video doing astonishing things to the development of better compressors.

I'm disappointed flash memory prices haven't fallen faster over the last year or two, and wouldn't be surprised if part of the reason is a desire to save the optical industry for a little bit (or give it a chance to establish itself) as we were getting very close to a point that HD movies could be distributed on SD cards, either owned by a customer (buy a 128GB drive, take it to a store and have them copy the movie you want with the soundtracks and extras you want, for a fee, or use some kind of studio-approved software to download and store the movie in the same way) or actually mass produced for a price not far off an optical disk. I suspect the magic figure to get that going is around 2G/$ in small sizes (eg 20G for $10, not 200G for $100.) With this kind of technology, we'd be looking at the players themselves costing $10-20, perhaps even players built into TVs rather than appearing as separate boxes.

The sooner we get rid of spinning plastic discs, the better, and I'm hopeful that flash will take us there.

Re:Gartner is wrong

[Lumpy]

AnyDVD-HD for the BREAKING of the protection. You cant rip with AnyDVD.

I then use RipBot264 to have an all in one Ripping setup. works great.

Re:Gartner is wrong

[DusterBar]

What's the cutoff, anyway? You end up with applications (or at least parts thereof) on the boot drive in practice because pretty much every operating system demands it... well, except Unix :)

So, you are saying "pretty much every operating system demands it... well, except for all those that are not Windows" (Ref MacOSX is Unix, Linux is like Unix, Solaris, FreeBSD, AIX, QNX, SCO, etc.) And then there are the Mainframe OSes (not Unix like) that also support volumes and isolation of OS from Applications from Data.

Re:Gartner is wrong

[petermgreen]

It's difficult to say.

Hdds require some complex precision parts, namely the heads and the assemblies that move those heads. They also require very stable motors.

BUT afaict the actual data storage platters are relatively simple to make just a uniform coating of material on a platter base.

A SSD OTOH requires a very large silicon area all patterned in many stages at an extreme density. Furthermore since huge ICs have unnaceptable failure rates many seperate ICs must be used to make up the SSD. These in turn need to be mounted on circuit boards which need to be stacked together and connected to further chips that provide a fast interface between all those flash chips and the host computer.

Very different types of manufacturing so it's hard to say which is more "complex". Complicating matters further is the fact that a lot of the costs (on both sides) are in R&D.

Flash scales down better which is one of the reasons it's taken over the portable market.

Patents

[tepples]

I'm sure there are other considerations.. prob some monopoly on NAND manufacturing or something.. but that'll eventually sort itself out

Monopolies tend to take 20 years to sort themselves out.

SSD are not more expensive

[Bardwick]

If your doing cost per gig, okay. If your doing cost per I/O, SSD is actually a tad cheaper. Typical 300g 10k RPM drive will net you about 150-180 IOPs. SSD will be ~70X.

Cheap competition? Niche benefit? Paredo Optimal..

[Orga]

When platter HD's were coming out we didn't have much of an alternative to look at, today Solid State drives have a well established competitor. As TFA states, people benefiting from these drives are a niche, the general person spending all day on facebook, youtube doesn't really see any benefit from spending extra money on a SSD. And when people do even care about performance.. well it's coming at a hefty price. I build my own pc's but while I'd like a SSD I'd rather spend the extra 100's of dollars on a much superior CPU and/or RAM that that money could spent on. Prices will take a long time to come down imo because the demand simply isn't there to push it forward. The mass majority of the marketplace just don't store MASSIVE files that need super fast access.

Joseph Smith and the diminishing value of wives

[wonkavader]

Well, he'd know.

News from the Future

[Anita Coney]

April 19, 2011 - PC World Magazine

The Federal Trade Commission has launched an investigation of price fixing among manufacturers of solid-state drives. A spokesman for the FTC says that the top hard drive manufactures "have colluded to keep the prices of SSD drives artificially high." A representative from Western Digital has stated that his company takes the charges seriously and that it will fully cooperate with the investigation.

All electronics have stayed/increased in price

[odin84gk]

Thanks to everyone shutting down factories during the recession, there is a severe shortage of analog parts, electrolytic capacitors, and some FET's. It is typical to see a 4-8 week lead time on an order of 20k. A 16 week lead time makes you Very uncomfortable and you start looking for second sources or redesigns.

Some analog/digital companies are shipping at 16-24 week lead times.
Some electrolytic capacitors are at a 40 week lead time.
And at least one major company stopped accepting new orders.
In the mean time, some distributors are starting bidding wars on parts that they do have.

Right now, demand is far greater than the supply. It is going to be at least another year before prices start to come down.

Re:All electronics have stayed/increased in price

[misexistentialist]

It does seem that the price of magnetic drives has been on a plateau for a while too.

Re:All electronics have stayed/increased in price

[TheSync]

Thanks to everyone shutting down factories during the recession, there is a severe shortage of analog parts, electrolytic capacitors, and some FET's.

I got burnt on 4-port HWIC cards for 1841 routers (HWIC-4ESW) and had to start using the 2-port cards (HWIC-2FE). Supposedly due to factories being shut down during the recession...

Re:All electronics have stayed/increased in price

[omglolbah]

Simple answer: They didnt.

What they did is to shut down when demand started dropping. They didnt want to be left with a big stock of parts that -could- be obsolete in a year.

Since so few were willing to gamble on demand raising we are now stuck with them trying desperately to ramp up production to meet demand and failing miserably. Ramping up production is a heck of a lot slower than shutting down...

Re:All electronics have stayed/increased in price

[omglolbah]

It is really easy to save money in a factory by shutting it down. On the short term you get incredible profit...
Then you run out of stock and catch fire *snickers*

Unfortunately few were willing to go on producing goods during the recession and are now severely fucked when it comes to supply... Funny how that happens every time the market takes a dip... You would think someone would learn :(

Re:All electronics have stayed/increased in price

[Anonymous Coward]

Whatever you do, don't design with Maxim parts. You will get immediate samples, get it prototyped and then find out later the the parts are pure Unobtainium. Further, nothing from the company is available from a second source.

Re:All electronics have stayed/increased in price

[metaforest]

This is great for small runs with flexible designs. Large players can't hit their sweet spots, but small run designers can by designing layouts to support several different packaging footprints for passives. It requires a little fancy dancing to alter the design to keep up with available parts. The big guys can't do it.

Two things.

[fuzzyfuzzyfungus]

While the question of whether prices aren't being competed further down because of collusion, or just because of inescapable production costs is an interesting one(and hopefully somebody has their forensic accountants on it, just to be sure), it seems reasonably obvious why SSDs have settled into the niche that they have, and why the manufacturers are making the size/price decisions that they are.

Now that the initial round of epically bad JMicron controllers are mostly gone, and the boring Samsung reference ones are confined mostly to build-to-order options on corporate laptops, all but the ghastliest SSDs are embarrassingly superior to HDDs for the sort of random mixed read/write that makes such a difference for desktop responsiveness. At the same time, though, nothing short of alien nanotech is going to allow them to touch HDDs in price/GB. That being so, you would expect to see SSD capacities largely cluster around "enough for a Windows boot volume, with a few key applications on it; but not much more". Anything less is largely useless to the target market(or, more accurately, anything less is aimed at the embedded devices market, and probably uses entirely different connectors and isn't sold at retail) and anything more gets very expensive very fast. This is, also, the reason why a lot of the high capacity (512GB to 1TB+) SSDs that you see are actually 2 or 4 of the vendor's lower capacity boards stuck together behind a cheap RAID chip. The market for the super high capacity ones just isn't all that big, at least among systems that use SATA as a storage connection bus, so the high capacity drives being sold are practically low-volume engineering samples, just polished enough to be sold for the usual early-adopter premium.

The only real forces supporting the existence of SSDs larger than that are high-end laptops(if you only have one drive slot, you can't adopt the mixed SSD/HDD strategy), a few loony enthusiasts(if you are the sort of person who buys every highest-end video card on release day, you can probably be convinced to go for a couple of 512GB SSDs, in RAID of course, for your gaming machine) and some truly titanic databases run by the deep-pocketed(though it isn't clear how much of that is SATA connected, and how much is the directly PCIe attached stuff, which is even faster).

Re:Two things.

[misexistentialist]

That being so, you would expect to see SSD capacities largely cluster around "enough for a Windows boot volume, with a few key applications on it; but not much more"

Microsoft will continue to drive innovation with Windows 10 + MS Office Suite requiring 500GB.

Chinese, but others?

[AnonymousClown]

The links talk about China and Japan respectively. But then there is Singapore, Thailand, Malaysia, Viet Nam, Costa Rica, and I'm sure there are other countries where chip making is occurring. I really don't think the price fixing is that wide spread and even if it is, for the sake of argument, it cannot last. OPEC has tried for decades and always someone cheats - and I do mean always.

Echon 101.

[jellomizer]

If supply goes up and demand stays the same prices will go down. If demand goes up and supply stays the same prices will rise. What happened was the economy dropped so they lowered their supply otherwise what could happen is the Supply/Demand curve would fall under making profit. Now as we recover demand is rising again but they are unsure about the longterm projections so they are keeping supply still low for a while.

This isn't necessarily greed unless they are bean counters. Sure you may make more per unit but if you sell more units for less then you can make more money.

SSDs on the desktop: ReadyBooSSD anyone ?

[obarthelemy]

SSDs improve performance only in very specific scenarii: small random reads are their forte, large reads are OK, writes are bad, especially small random ones. On the desktop, that makes them good system drives, OK Apps drives, bad data drives, terrible log drives. On the desktop, basically, SSD are useful when booting, launching an app, loading a level or other ressources. Nowadays, I boot my PC twice a month, launch apps at most once a day, and don't really play anymore.

To complicate matters, most OSes and apps are made up of a good 50% "dead" files that are very rarely used, and also have log files that get written to frequently. Installing a whole OS, app or game to an SSD is majorly wasteful because of that. Manually segregating "frequently-read" from "frequently-written" and from "dead weight" files within an OS or app is at best cumbersome and difficult, at worst, impossible.

I'm wondering why OSes don't yet support some king of SSD ReadyBoost: it would make a whole lot of sense to use a smallish SSD as a cache for frequently-read (not written) files. One SSD maker has released a thingy that clones the first x sectors of an HD to a SSD. Though automatic and easy, that is very crude, as caches go. I seem to remember one of Linux's filesystem allows to easily use an SSD as an intelligent cache, but that filesystem is fairly marginal ( ZFS ? not sure). MS has not adapted ReadyBoost.

With an adapted ReadyBoost, I'm sure I could get 90% of the benefit of a large (64 Megs) SSD in a much smaller (16 Megs ?) one. I'm waiting for that, or, if MS doesn't wake up, for prices to go way down.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[thue]

> writes are bad, especially small random ones

Some of the first SSDs had big trouble with random writes. The good new ones handle them much better than a mechanical hard disk.

See for example http://www.legitreviews.com/article/1022/6/ . While a few SSD have 20KB/s random write speed, the good ones like Intel's x25-m have 60MB/s, way better than a hard disk.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[viper66]

I think your facts are out of date. The latest SSDs beat hard drives in every category, sequential reads/writes and random reads/writes. SSD random write performance has been superior to the fastest hard drives for quite a while now. Performance is even better with TRIM and 4K alignment in Windows 7. It is sequential write performance that has typically been weak, but even that is no longer the case.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[TheDarkMaster]

I agree. It is the main reason I avoid SSDs, besides the price: They are bad in important activities of an OS, such as keeping logs updated and hundreds of small writing that they are always doing (eg, update the date of last access of a file). And we can not ignore the problem of wear, which unlike a hard drive that can be written many many times, an SSD has a limit of write operations, and this is a small limit for an OS that writes things all the time during PC operation.

And of course, also have the problem of garbage SSDs sold in huge quantities (in my country is what most have), which even though they cost several times more than an HD equivalent, have worse performance than a HD.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[bertok]

I agree. It is the main reason I avoid SSDs, besides the price: They are bad in important activities of an OS, such as keeping logs updated and hundreds of small writing that they are always doing (eg, update the date of last access of a file). And we can not ignore the problem of wear, which unlike a hard drive that can be written many many times, an SSD has a limit of write operations, and this is a small limit for an OS that writes things all the time during PC operation.

And of course, also have the problem of garbage SSDs sold in huge quantities (in my country is what most have), which even though they cost several times more than an HD equivalent, have worse performance than a HD.

Your info is out of date. The latest SSDs outperfom every hard drive ever manufactured in every performance metric, often by orders of magnitude, and last long enough that you'll have replaced your computer before they wear out.

According to the reviews of the "Crucial RealSSD C300" drive, which uses SATA 6 Gbps - because SATA 3 Gbps just wasn't fast enough - it can do upwards of 60,000 IOPS if you have multiple threads hitting it with small random 4kB reads and writes. To put that in perspective, that's the same performance as about 300 of the 'enterprise' 15K RPM fibre channel drives in one huge SAN array, which would cost millions. No single mechanical drive of any kind can even begin to get anywhere near that kind of performance, and even huge enterprise arrays are being left in the dust.

The hard disk is the last mechanical component commonly used for computation. It is way overdue to be replaced by solid state.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[TheDarkMaster]

Sorry, not on the world below the Equator line. Here one SSD like you cited is "luxury", and with a price tag very, very very big for a PC storage (intel X25-M 160GB costs US$1000 here). The only good thing on SSDs acessible here is the access time, the rest is garbage at best.

And here, a PC needs to last much more than two years, someones are in use (and working well) for more than 10 years. No actual SSD can do this.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[BikeHelmet]

The latest SSDs beat hard drives in every category, sequential reads/writes and random reads/writes.

Yep, but last gen sure didn't! I remember looking at some reviews pitting those awful JMicron SSDs against that dual-head 2TB WD Black. Imagine my surprise when the HDD beat them in real-world IOPS tests. :/

As amazing as SSDs are, for me to buy one the price has to hit ~$70, and not be total crap.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[BikeHelmet]

According to the reviews of the "Crucial RealSSD C300" drive, which uses SATA 6 Gbps - because SATA 3 Gbps just wasn't fast enough - it can do upwards of 60,000 IOPS if you have multiple threads hitting it with small random 4kB reads and writes. To put that in perspective, that's the same performance as about 300 of the 'enterprise' 15K RPM fibre channel drives in one huge SAN array, which would cost millions. No single mechanical drive of any kind can even begin to get anywhere near that kind of performance, and even huge enterprise arrays are being left in the dust.

Great for forums and other database crap.

But not for file servers or video streaming.

Two words: RAM Cache.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[bertok]

According to the reviews of the "Crucial RealSSD C300" drive, which uses SATA 6 Gbps - because SATA 3 Gbps just wasn't fast enough - it can do upwards of 60,000 IOPS if you have multiple threads hitting it with small random 4kB reads and writes. To put that in perspective, that's the same performance as about 300 of the 'enterprise' 15K RPM fibre channel drives in one huge SAN array, which would cost millions. No single mechanical drive of any kind can even begin to get anywhere near that kind of performance, and even huge enterprise arrays are being left in the dust.

Great for forums and other database crap.

But not for file servers or video streaming.

Two words: RAM Cache.

Two words: price/performance.

RAM cache is much more expensive than an SSD per gigabyte. Secondly, cache does't help for "cold start" scenarios, random reads across data that's even slightly bigger than the available memory, and does nothing for write latency.

My point was that SSD performance vastly outstrips mechanical hard drives. I didn't say they were most cost effective storage per gigabyte if performance doesn't matter.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[bertok]

Sorry, not on the world below the Equator line. Here one SSD like you cited is "luxury", and with a price tag very, very very big for a PC storage (intel X25-M 160GB costs US$1000 here). The only good thing on SSDs acessible here is the access time, the rest is garbage at best.

And here, a PC needs to last much more than two years, someones are in use (and working well) for more than 10 years. No actual SSD can do this.

I'd debate that. You can get small (64GB) SSDs, you can buy them directly from the US and have them shipped, so the cost shouldn't be much higher anywhere in the world. You can buy the latest & greatest 3rd gen SSDs online, it's not like they're magically restricted to "north of the equator".

Secondly, it's more cost effective to buy a new PC with a slightly lower overall spec and spend the remaining money on an SSD, because it makes such a huge difference that even a relatively sluggish CPU will appear blazing fast.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[obarthelemy]

Plural might be "scenarios" or "scenarii", but certainly not scenari. Italian plural, "o" -> "i", thus "io" -> "ii".

http://en.wiktionary.org/wiki/Talk:scenarii

"English must have borrowed scenario and scenarii before Italian dropped the second ‘i’ in the terminal ‘-ii’ (see this). As it stands, scenarii is formed identically unto concerti and virtuosi; this isn’t an example of “‘-o’ ‘-ii’”, but rather of the somewhat more familiar “‘-o’ ‘-i’” (showing itself as “‘-io’ ‘-ii’”). English is fairly open to absorbing little bits of grammar (such as how to form plurals) from other languages; I imagine that if it still had a case system, it would indeed take unto absorbing foreign inflexions for marking grammatical case. Raifhr Doremítzwr 23:56, 11 January 2007 (UTC
English absorbs lots of things, this is partly what makes it colourful in, which is a good thing of course. The problem comes when there is a clash in how to treat absorbed words i.e. how to inflect them. Early on in the borrowing phase all sorts of things can happen, later on most people come to a consensus by various means. That is in the case of scenario and other borrowed words we can have multiple plurals, the consensus would be scenarios and is formed regularly by adding -s and the alternative scenarii is not the consensus use (this can by assumed when teachers may take aversion to it in comprehension for example) and it would follow that it is formed irregularly due to the use of a foreign infelction system that is obvioulsy alien to most English speakers.--Williamsayers79 10:31, 12 January 2007 (UTC)
I see absolutely no evidence that English borrowed scenario and scenarii before Italian changed the spelling. It’s not even clear that Italian did shift from -ii to -i. As I understand it, -ii only occurs today, as one of the posters on the link you reference says, in the presence of a stressed i in the singluar (lo zio/gli zii) and particularly not in the case of -io (il gregario/i gregari). I just did a quick search through the Inferno. Several words ended in -ii, but most appeared to be verb forms. The exception was rii. On the other hand, Dante has avversari (adversaries), Tartari (Tartars). I’m pretty sure I’m looking at Dante’s actual spellings, but perhaps I’ve missed something?
Searching Project Gutenburg doesn’t turn up scenarii at all, while scenario appears in 231 books, scenarios in 69, and scenari in one (in Italian, written in 1885).

So, given that
scenarii doesn’t appear to be a valid Italian plural form in either modern Italian or Dante’s Italian.
Latin doesn’t work.
The English scenarii seems to be recent.
The Engiish scenarii only seems to appear in a limited, technical context, not in general use.
The meaning English scenarii is fairly far removed from the Italian meaning scenari (which has more to do with theater than operations research and formal logic)"

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[BikeHelmet]

It's called value. $180 for something that lets me start three games in 30 seconds rather than 60 seconds isn't good value.

It might be worth it for Crysis - that game was shown to have 50% higher minimum framerates with an SSD - but Crysis isn't one of my favourite games.

On the other hand, $260 for a 2TB HDD that lets me start a game in 40 seconds rather than 60, isn't bad. And that amount of space is pretty good value compared to a current gen SSD.

But you know what's even better than that? $80 for a 1TB HDD, and just waiting 60 seconds for a game to load.

It's no biggy. If it matters... there's always Superfetch.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[BikeHelmet]

Benchmarks have shown that in real-world scenarios, reading as little as 1MB in a strip puts HDDs very close to SSDs. Certainly close enough that 4+ of them in RAID would be faster and cost less. I'm guessing the sweet spot is around 4MB, since that's what Google uses.

Nobody cares if videos start streaming in 250ms or 1000ms, so it's fairly safe to say video streaming will be fine if handled by HDDs and RAM cache.

But you're right that more IOPS never hurts. We have L3 between L2 and RAM - SSD between HDD and RAM does makes sense, especially for comment pages and stuff. A few SSDs could manage the thousands of requests to comment pages with ease, but HDDs would be stopped dead in their tracks by all the seeking.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[bertok]

Benchmarks have shown that in real-world scenarios, reading as little as 1MB in a strip puts HDDs very close to SSDs. Certainly close enough that 4+ of them in RAID would be faster and cost less. I'm guessing the sweet spot is around 4MB, since that's what Google uses.

Nobody cares if videos start streaming in 250ms or 1000ms, so it's fairly safe to say video streaming will be fine if handled by HDDs and RAM cache.

But you're right that more IOPS never hurts. We have L3 between L2 and RAM - SSD between HDD and RAM does makes sense, especially for comment pages and stuff. A few SSDs could manage the thousands of requests to comment pages with ease, but HDDs would be stopped dead in their tracks by all the seeking.

Congratulations, you've found a niche application in which the hard drive performance limitations don't matter. Tape drives are still popular for the same reason: if all you care about is streaming speeds and capacity per dollar, then SSDs clearly don't make sense.

However, for 99% of desktop users, an SSD makes perfect sense. Operating system and application IO is extremely random and made up of very small reads and writes, so latency matters.

My laptop used to take 15 minutes to go from "cold start" to "logged in, all applications started and usable". I timed it. I took the exact same installation across to an SSD using a ghost image, and that startup time is now 35 seconds, half of which is the BIOS POST sequence. I'm not talking about "login prompt" visible, which is bullshit, but Outlook, Visual Studio, and maybe a couple of other applications loaded and responsive, which takes a long, LONG time on a hard drive.

Installing an SSD cost me 1/3rd of what a new laptop would have cost, and made my computer infinitely more usable. I can now properly multitask without being driven crazy by the endless waiting for the crunching sounds to stop. I once noticed that my CPU fan was going a bit loud all day while I was developing with Visual Studio, and it was only late in the afternoon that I realised that I had accidentally left a long-running SQL query running in the background. It was doing 50MB/sec of IO, and I didn't even notice, other than the sound of the fan. That's just not possible with a hard drive.

Please stop repeating some gibberish you heard years ago about 1st gen drives. Moving from hard drives to SSDs is a bigger difference than what it was like moving from floppies in the 8086 days to hard drives. It's like VHS to DVD! Nobody who's used a decent SSD in their day-to-day computer talks about going back. I do hear them whining about how broken their "other" computers feel afterwards, however.

Re:SSDs on the desktop: ReadyBooSSD anyone ?

[BikeHelmet]

Congratulations, you've found a niche application in which the hard drive performance limitations don't matter.

I could say the same about you. Congratulations on figuring out where SSDs absolutely shine.

My laptop used to take 15 minutes to go from "cold start" to "logged in, all applications started and usable". I timed it. I took the exact same installation across to an SSD using a ghost image, and that startup time is now 35 seconds, half of which is the BIOS POST sequence. I'm not talking about "login prompt" visible, which is bullshit, but Outlook, Visual Studio, and maybe a couple of other applications loaded and responsive, which takes a long, LONG time on a hard drive.

Oh my... 15 minutes! My old Athlon XP boots off HDD in 28 seconds. (Firefox open in < 35 seconds)

My new Phenom II X4 system boots off HDD in about ~2 minutes, but Firefox is open in about 50 seconds. Everything else gets shuffled to the background.

Benchmarks seem to indicate that perhaps some unknown factor was influencing your HDD boot times. Perhaps the drive was close to failure? Or was it one of those really low RPM drives? I'm talking about 7200RPM drives, obviously - some of them like the 2TB black are dual-head, and easily shave 40% of your boot time off, over a regular 32MB cache 7200RPM drive.

I'm not convinced everyone is better off with SSDs rather than a really fast HDD. I'm glad you've found your favourite new toy - but after playing with a few SSDs and high end HDDs, I decided the space and price was worth more than the performance, to me. Besides that, I'm talking about a desktop, and you're talking about a laptop. Of course you want an SSD in a laptop. They're power efficient and shock immune.

I'm also not compiling the linux kernel daily. I'm loading games, which benchmarks indicate would barely improve 20%. I'd much prefer to be able to put every game on a high performance drive, rather than just 3 games.

So I stand by my opinion: I'm not convinced everyone is better off with SSDs rather than a really fast HDD. But unlike you, I'm at least willing to examine all available options.

The problem is that quality went down too.

[sethstorm]

Larger LCDs are finally affordable now, and how many years did that take

The problem with that is that they cut out more than a few pixels and certainly a lot of of the quality in the panels.

Re:The problem is that quality went down too.

[drinkypoo]

The problem with that is that they cut out more than a few pixels and certainly a lot of of the quality in the panels.

WTF are you talking about? 1080p displays are cheaper than ever, have better contrast than ever, and have the same number of pixels they always did. You have no idea what you're saying. Or maybe you and I have different definitions of 'large', but a TV just like the one I paid $1500 for is now $1200. Well, it's not precisely the same. It has a new model number.

Re:The problem is that quality went down too.

[toddestan]

Nearly every panel on the market is a cheap TN panel, and while they play all kinds of games to get contrast ratios and color saturation up, have fun trying to properly calibrate one. And where are all the high resolution panels? Everything now is pretty much 1080p at best - maybe that's fine for TVs but why can't we push the pixel count higher for PCs, especially since 16:9 really doesn't work that great on a PC anyway? And where have all the good laptop screens gone? A few years ago you could get a 1600x1200 IPS screen at 15" - nothing like that today even comes close.

Same reason LCD TV prices aren't "going down"

[macraig]

It's bubblegum feature creep, the same tactic Microsoft used for years to justify prices for Windows. In the case of LCD TVs, they're using extra "Hertz" and LED backlighting (never mind the misleading marketing leads some people to mistakenly believe that the display itself is actually LEDs) to keep the prices artificially high. It's these industries' version of the fast food industry's coke-and-fries tactic: upsell the product with low-cost features or add-ons that add almost nothing to cost but add a mountain to profits. In the case of magnetic-media disk drives, they're adding more cache and tweaking this or that, which ultimately changes the total cost of production very little but helps the manufacturers stave off the price drops that SHOULD occur due to savings from mass production. In the case of SSDs, I duuno what the specific feature creep is, but you can bet that is what's taking place.

This is what manufacturers do to keep the full benefits of mass production from actually "trickling down" to consumers.

Re:Same reason LCD TV prices aren't "going down"

[ZFox]

Don't be so cynical--it's not like The Masters of Mass Production* are sitting in a smoke-filled room thinking of the next LCD TV feature, just to sock it to the consumer.

It could be the fact that feature-creep comes into play when they want to differentiate themselves in a saturated market.

With Microsoft I think it was the fact that they recognize if you don't make it do anything new, nobody will want it. The fact that it has new process management model isn't exciting for most consumers, so they tack on a lot of "bloat" features and for better or for worse, let the marketing teams do their thing.

*Trademark Pending. This is the perfect villain for today's Saturday cartoons.

Someone is buying massive quantities of them

[melted]

Someone is buying massive quantities of them. That's the only explanation. The enterprise is only now catching up to the benefits of low latency SSD access. You can use fewer servers and serve data with much better latency and throughput for IO bound tasks. Anyone who needs low latency random access to data (ads, search, data warehousing, OLAP, content distribution networks, hotspots in map data serving, etc, etc) are switching to SSDs right now as quickly as their budget allows.

Re:Someone is buying massive quantities of them

[perryizgr8]

let me say it: google.

Re:Someone is buying massive quantities of them

[owlstead]

I know a shop that also sells laptops with the Intel 40 GB SSD in high quantities. Don't forget, these cheap things (100 euro *retail* here) still push 170 GB read times. Together with higher reliability, these drives make lots of sense for businesses. Consumers of course are more interested in movies and pictures - and they don't know SSD's - so they care more about GB than speed and reliability.

Re:Someone is buying massive quantities of them

[melted]

Not sure about Google. They have so many servers, they can keep everything in RAM easily, and their servers are dirt cheap. More likely Bing, whose per-server costs are substantially higher and smaller ad networks whose number of servers is in the hundreds / thousands.

Re:Someone is buying massive quantities of them

[metaforest]

The latest showdown I read put the fastest HDDs a distant LAST in every category that matters for active storage devices versus SSD. So it makes sense that for cloud enterprise these drives even at high premiums are full of win in that market segment.

Hybrid drives

[Registered Coward v2]

Using a mix of flash and spinning disk technology you could build a single hybrid drive that supports fast operation by optimizing what uses flash and what uses the older tech; this would increase the perceived speed of a machine while keeping costs lower. The OS would need to support such a drive; but it would result in faster machines with reasonably large storage capacities at a lower price point. It's not really a new idea, variations of the solid state / hard drive memory mix use have been around a while.

I realize you could use two drives; but shoe horning it into one laptop form factor would make it a lot easier to incorporate into existing laptop form factors; as well as add backwards compatibility via an OS update.

Re:Hybrid drives

[toddestan]

Isn't this basically what Windows has been doing for 3 years now with ReadyBoost?

Re:Wait, what?

[petermgreen]

It depends on what you mean by SSD prices dropping.

Some manufacturers introduced some drives below the $100 mark but they were very low capacity. So the minimum price of entry dropped by the cost per gigabyte stayed about the same.

Rotating media to stumble on 4kb sector migration

[rwa2]

SSDs will probably see a slight boost in the next year or two due to rotating hard disks finally getting around to migrating to 4kb sector sizes (which still poses compatibility and performance challenges for Windows XP, and even many Linux utilities aren't quite prepared yet)

2TB drives have just started to come out, which is actually the limit to the 512b sector sizes hard coded into many OSs for the past couple of decades. Here's a good explanation:
http://arstechnica.com/microsoft/news/2010/03/why-new-hard-disks-might-not-be-much-fun-for-xp-users.ars/

So the SSD manufacturers could do well to advertise that people still running older systems may run into a fair amount of trouble upgrading to newer hard disks with 4kb sectors. Even in compatibility mode, they've found 3x-4x slowdowns in write performance if the file systems aren't aligned with the sectors just right. Makes the decision to upgrade with SSDs sound that much better if they can really increase their performance and not have to plop down a few hundred $$ and time on OS upgrades.

(yes, of course there are similar write performance gotchas with aligning file systems to the 128k SSD erase blocks, but those shouldn't bite quite as much)

Re:Terrible Windows OS development paradigm

[andreasg]

You can mount harddrive devices to directory paths in windows. SSD -> C:/ HDD -> C:/Program/Data

Why would prices go down? They shouldn't.

[m.dillon]

There's no reason for prices to go down on SSDs. Economic downturn? Not for flash chips, not when there are basically only two vendors making them, not when demand is high, and not when mass production has a low barrier to entry and is already extremely cheap... just plop a bunch of chips on a circuit board and you are done. That is what a SSD is.

-Matt

Re:Terrible Windows OS development paradigm

[Xeleema]

Linux has always had this problem solved, but Windows doesn't have an easy way of going about this.

Linux? Hell, *NIX has been that flexible for, oh I don't know...as long as Windows has been around? Heck, used to be that drives were so small, you had to be able to split applications from their data.

Most of Microsoft's products suffer from a cultural problem;

20_Yrs_Ago: "Hm, where should programs go? I know! Someplace called 'Program Files', yeah, with a space in the path and everything. It'll live in the root of the disk!"
Future_Thinkers: "Ah, um, excuse me, Mr. PHB. That might be a bad idea. See, the rest of the computing world is laying out Operating Systems like so and...."
20_Yrs_Ago: "Whoa?! Where the HELL did you come from?!?! *BLAM*"
Future_Thinkers: "Ah!! Ahh! You shot me! You shot me in the leg!!"
20_Yrs_Ago: *BLAM* *BLAM* *BLAM* *BLAM* *BLAM* *BLAM* *BLAM* *BLAM* *BLAM* *click* *click*
Future_Thinkers: .....
20_Yrs_Ago: "Geez, that's the fourth one this week. Where are these guys coming from?!"

Prices of LCD TV's aren't going down? Says who?

[N0Man74]

I bought my first LCD HDTV about 2 years ago. It was a 27" 720P, and I paid $650 it on a Black Friday Sale (it was normally $1,000). Now, in just a couple minutes, I can find several new 40" 1080p televisions for the same price I paid for my old one, or find 26" - 32" 1080p sets for significantly less than what I paid for mine.

LCDs (for the *equivalent* quality and features) *ARE* going down.

If I had to guess, your perception that they aren't going down is likely due to the fact that there are many new sets with higher specifications that keep the price point the same, but offer more features and quality.

If you compare 60hz 720p HDTVs from a few years ago with newer 1080p displaying at 120hz, then yeah, the prices don't seem to have changed much. However, if you ignore specifications in a price analysis, then computers haven't really gotten cheaper in the last 20 years either.

the U.S. dollar has not remained stable

[paulpach]

Between 2008 and the first quarter of 2009 there was a rally to the dollar. When compared to other currencies in the world, it went from 72 to 88 in the dolar index. This caused everything to drop in price in dollars. Since then, the US dollar has fallen sharply, fast enough to prevent improved technology from lowering costs. It has rebounded a little this year, but not because the dollar gained anything, but because the euro fell sharply.

But lets price it in something less erratic than the US dollar, euros or anything that comes out of a printing press. Lets price it in the most stable currency known to man: gold.

pricing the same 16GB in gold we get:

Q1 2008 0.006875 oz
Q1 2009 0.002777 oz
Q1 2010 0.002363 oz

As you can see, when priced in gold, you can see that flash has never really stopped going down in price. It simply slowed down probably due to some of the reasons explained in TFA.

Re:the U.S. dollar has not remained stable

[AHuxley]

Dont mention fiat currency, M3, U6, gold or any other real world data points :).
Any issues with SSD retail price are due to crony capitalists (not the good local heros who rescued your pensions and savings).
Communists running the mines, dictators running the mines, unions, cartels, price fixing, software fixing or some other external issue.
The US $ is strong and will always be so.

Re:Terrible Windows OS development paradigm

[nabsltd]

It's hard to separate all the "data" from the apps on Windows, but you can separate most on Windows 7 with no tools other than the original install media.

After installing the OS and adding your "data" drive, boot to the Windows 7 rescue mode and open the command prompt, and copy the "C:\Users" and "C:\ProgramData" to the data drive, fix the links so that everything points to the new drive, and edit a few registry locations.

There are also Microsoft tools to do this during the install.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:LCD TV prices *are* going down

[macraig]

While prices might be dropping, they're not falling to the point they would have in the absence of those gimmicks, ESPECIALLY at the low end. I've been paying very close attention to LCD TV prices (at the low end) for a year and a half; at just about precisely the points where there was an imminent and precipitous price drop, new "features" magically appeared on new models.

Re:In other words

[haruchai]

I'm with you on this. Friggin' rebates must have been invented by the Devil. I lambasted one of my favorite retailers for carry a $25 dollar rebate on a $700 drive. At present, my rebate threshold is $25 or 10% whichever is GREATER.

Crappy SSDs scares buyers.

[TheDarkMaster]

The cost-benefit of an SSD is still very bad. It is bad for write operations (the OS lives writing stuff all the time while using the PC), over time it loses performance and the problem of wear makes it has a limited shelf life (even with all the techniques of wear leveling) compared with an HD of the same size.

As if this were not enough, the market is saturated by junk SSDs with JMicron controllers and junky MLC memories, with performances so terrible they can be worse than a hard drive of same capacity. The buyer is afraid to take garbage for the price of luxury, and then just avoiding buying.

And the cost in most countries is really prohibitive: As an example, here in Brazil a 160GB SATA SSD Intel X25-M is US$ 1000. No kidding.

Re:Life isn't free

[Dogtanian]

If you think you can manufacture it in a less expensive manner, go for it.

Mmm, uh huh. Never mind the prohibitive cost of entry into a field of this sort and the ability of established players (as a general principle of markets) to shut down new competition by sheer might. In fact, they don't even need to be actively hostile, as their size dictates that merely acting in a "fair" competitive manner will kill new entrants before they grow into anything like a threat.

That's capitalism folks.

Your obviously intended implication here being "suck it up and accept it", but an equally obvious and valid implication (depending on one's view) could be that this demonstrates a serious flaw with free-market capitalism.

I don't *entirely* disagree with your implication that people should or shouldn't buy the products if they do or don't want to- within reason- but your silly rambling of context-free libertarian/free-market nonsense that implies any random Joe Public could- or should- compete with such massive operations is drivel.

Re:In other words

[Capt.DrumkenBum]

I don't do mail in rebates. If the regular retail price is too high then I do not buy. I value my free time very highly ($100 per free hour.) So filling out a rebate form neatly enough for them to read, then going to the store to buy a stamp. (I pay all my bills online.) So if I figure I would spend 1 hour getting the rebate, that rebate would have to be for over $100 for it to be worth my time.

Die Bonding Costs Go Up

[wonkavader]

"...packaging costs go up; because the die bonding gets harder and more complex..."

But is that really unit cost or RD cost (which gets amortized into the costs of the unit, but eventually disappear)? If the cost is developing a robot which stacks chips nicely and bonds them, but the robot costs a LOT of money to develop, then that should effect the ultimate cheapness of stacked stuff a few years after everyone buys one of these expensive monsters and starts cranking out gobs of stacked flash.

Or do you think that the difficultly will cost more over time because the number of defective units will go up?

Or does stacking require materials which are more expensive than non-stacked?

Re:In other words

[haruchai]

With only 1 exception, all the things I've bought with rebates were things I really wanted or would have bought anyway.
And I guess, I'be become efficient at processing them but I still hate it. I loathe doing it but cash have been tight for a long time and I don't want my systems, and experience to fall too far behind the curve.
But I dream of the day I'm well off enough to be able to tell companies that I would have bought their product if it wasn't for that damned mail-in rebate.

Re:Sell me some fucking SAS SSDs already!

[Wesley Felter]

STEC has SAS SSDs for only $50/GB. That might give you some incentive to reconsider your SATA-hating ways...

Re:Sell me some fucking SAS SSDs already!

[KonoWatakushi]

...and the STEC drives have the performance to match. If what you need are IOPS and not mere space, they are more than competitive, and you don't have to worry about them tossing transactions on power loss or performance degradation as they fill.

The same with HDs in the past

[Illogical Spock]

The same happened with HDs. Sure, for maybe 5 years or so now the price is dropping and the size increasing, but everyone that used a MFM or RLL hard disk - and even the IDEs for a long time - knows that we waited very long to see the prices really going down.

It will be like everything else: with the time passing they will get market share, then decrease the price. Even the RAM memory finally dropped, so why not the SSDs? But not tomorrow, I'm sure.

Re:Terrible Windows OS development paradigm

[drinkypoo]

Linux has always had this problem solved, but Windows doesn't have an easy way of going about this.

Uh, this is the programs' fault, not Windows. Programs choose to install DLLs to the Windows directory, which is furiously proscribed by basically everyone, unless you're updating an actual system component. Programs choose where to open their file dialog; some are hardcoded, and some are not. You can change the location of the Documents and Settings folder to relocate user accounts; many programs, including the entire Microsoft office suite allow you to change file save locations. As another poster has pointed out, Windows now supports mount points; but even in DOS, you could use assign and subst to turn folders into virtual disks and disks into virtual folders, I forget which is which.

So indeed, Windows has long had an easy way of going about this, but many windows applications pervert it. And meanwhile, I've had plenty of apps on Unix which always opened their file requester at the root, rather than $HOME or the cwd, so Linux doesn't exactly "solve" this problem either. You can mount your own filesystem on /usr/local, but if all the scripts point to /opt, you're either going to be editing a lot of scripts (with find -exec and sed -e, maybe, but still) or you're going to have to install it where it's meant to be installed. And a lot of the time, a symbolic link doesn't produce the desired result, so you can't use that as a solution either.