While i agree that 256Gbyte <1k wont be feasible, SSDs currently are overpriced.
There is no economic reason for Flash SSDs to be that expensive, except milking the early adaptors.
SLC cells are around $8 /GByte on the spot market now, MLCs at $2.50.
Maybe economy of scale and increased competition will help that.
(hint: Samsung and Mtron are in one bed, OCZ and Hama also only offer rebadged Samsung respectively Mtron drives. So no real competition going on, currently)
It's not just a matter of competition. Mtron has been practically the only maker of really fast SSDs. For those in need of cheaper ones, there are many. If you buy those MLC chips at US$2.5 per GB, you can make 256GB MLC-based SSD and sell one for, say, US$625, or even US$500. However, what about controllers? Case? Marketing? I mean those are targeted at different markets, and do you really believe it takes US$10,000 for Intel to assemble that workstation based on Skulltrail platform? Anyone here can probably build two 9300-based systems for under US$1K. I'd rather go for 10 or more E1200-based systems for the same amount of money since I can buy 256GB DDR II RAM for US$2, 40GB HDD for like nothing, E1200 (which is nearly as fast as those in the Skulltrail for most things most people do) processor for about US$40, a case plus power supply for U$10, and so on.
It's no coincidence that both Samsung and Mtron are in South Korea, and there are some other makers. The cheapest SSD costs under US$50. There's a 128GB SSD that costs over US$3,000.
Such products are "overpriced" in the sense some luxury cars are overpriced because they are handmade.
Intel will help to make SSD market thrive, but it's also commonly known Intel has worked with Samsung and Microsoft and though Intel makes CPUs better than anyone else on the planet, Samsung is the one with flash. Intel can probably sell 8-core 32nm processors for under US$100 if they want to by the end of 2009, but that's only because there's enough room yet to maneuver. Such options usually don't exist in the DRAM and flash industries. Which company's investing how many billions in USDs this year is no secret, though there are many secrets like exact yield rate, and that's the only information needed to predict roughly how SSD market will evolve in the next 5 to 10 years. Intel is not, Apple is not, AMD is not investing much, so don't expect a miracle.
For some, the biggest challenge has been Gigabyte's I-RAM-like products. I-RAM uses DDR modules. I haven't seen I-RAM II yet which should use DDR II. 2GB DDR II costs US$40 in Seoul which isn't that much considering it takes US$160 plus the cost of I-RAM II card versus the cost of 8GB USB stick (about US$30.) I'd rather buy I-RAM II-like cards based on DDR II than luxury SSDs like the 256GB Samsung one mentioned here. Spending a few thousands of USD's on a single hardware that combines storage and access time and transfer rate all at once isn't as good as spending US$1K on speed (imagine FOUR I-RAM II RAIDed to SATA, each with 8GB PC2-6400) and another US$1K on storage like five 1TB HDDs.
The bottomline's that 256GB SSD's not for you. Theoretically, you can have 256GB PC2-6400 DDR II memory with lower access times and faster transter rates for about US$4,000 which is cheap considering how much the fastest 64GB SSD costs now. The world has more fabs for DRAM than flash as well.