Flash Likely to Get More Expensive; Samsung Builds 10nm Fab
If you are looking to buy Flash memory, you may want to purchase rather sooner than later.
There are more signs of supply constraints in the flash memory market, which could lead to price increases. Following Toshiba's announcement that it would cut flash memory production by 30 percent, we have seen companies that rely on these chips warn of demand exceeding supply.
For chip makers such as Toshiba, that may be good news, but the product chain is likely to see higher prices. Among the companies that warned of revenue shortfall due to supply constraints are OCZ as well as Sandisk. OCZ said that it had record orders, but was not able to meet the demand because of flash shortage, according to Seeking Alpha. The publication noted that Samsung, the world's largest flash chip maker "indicated its intention to convert to the production of logic products from memory chips at its Austin, Texas facility."
Add to that the fact that Apple has largely secured first dibs on flash memory from all major manufacturers, and a greater than expected success of the iPhone 5 could make the current constraint situation even more serious.
On the brighter side, Samsung is reportedly preparing the construction of a fab in Xi'an, China, which will be dedicated to 10 nm flash chips initially. At a cost of $7 billion, the factory will be built in two phases and have floor space of 10.8 million square feet - the equivalent of about 187 football fields. The factory is scheduled to go online in 2014 with an initial floor space area of almost 5 million square feet. The Xi'an is home to a total of 37 universities and more than 3,000 IT R&D facilities.

*cough*Advertisements*cough*
Exactly.
by 4 to 6.5 times
basically 100gb now costs 100$
when the 10nm evens out 100gb will cost between 16-25$
*cough*Advertisements*cough*
What it also indicate is that Samsung might be able to make processors on 10nm by 2015, a full year before Intel.
Direct factory prices are up for new contracts. However, most suppliers have long contracts (meaning, deliver x amount at x price in x number of batches) and therefore aren't affected yet.
Probably because the history of collusion within the industry to fix prices (DRAM, Monitors, etc.)
I mean, since Flash Memory has been trending higher each year as it's included in various forms (phone flash memory, SSD, I could understand that the prices would go up in the short term, but like others have said, it's good over the long run.
As for hihi - this is flash memory on a fab/pre-assembly level, meaning that it's going to affect all products across the board even though there are differences in final product and specifics used per product. Probably why the made mention of OCZ, Samsung, and Apple.
The article is talking about memory not CPU, i dont think its even possible to hit 10nm on a CPU. "The problem is that at a certain point you run into quantum effects that simply can't be controlled with traditional lithographic methods. Quantum tunneling for example - where an electron is supposed to be on one side of a dielectric but magically appears on the other. Tunneling is a probability function that varies with distance. When you start hitting distances that permit the effect, you're screwed and there's nothing you can do about it. I just don't happen to know at what level that starts to become a problem - or rather, an unmanageable problem" by Twilyth on Techpowerup forum.
Why don't you go to the local library and look at a 10 year old issue of PC Magazine for example.
Look at the products and prices; than tell me again you do not get anything different for your money.
I bought a then state of the art 10GB HDD in 1998; cost me $1,000
Just picked up a USB keychain; 16GB for $9.95.
Competition?
Apple has no manufacturing capability of their own.
Yeah, it's 2014 for memory. If you read another article here you see that Intel released roadmap for 5nm. And it's expected that 10nm processors comes in production in late 2016/early 2017.
Quantum tunneling effects makes traditional computing techniques impossible at about 2-3nm. However, at 4-6 nanometer there's already the problem of accurate doping techniques. So the limit of traditional computing is in that area.
However, both doping technique limitations and quantum tunneling can be overcome by shifting to probabilistic computing techniques instead. That'll shift the limit to sub-nm levels.
Probibalistic computing is when you design a computer to guess the correct result instead of calculating. That allows for much much smaller and much much more efficient individual cores, and based on the level of accepted inaccuracy in the guesstimates, you just parallel several such cores combined with sophisticated algorithms to bring the inaccuracy down to 0% on the output.
hehe.. and if people listen and go out and buy more than they would have, they will create the increase in prices by suddenly hiking up demand!