Skip to main content
Message Font: Serif | Sans-Serif
No. of Recommendations: 11
Memory companies long dreamed of conjuring up a memory chip that offered the best of all possible worlds. (1) From the workhorse DRAM (dynamic random-access memory) they would take the ability to store vast quantities of data at rock-bottom prices. The thoroughbred SRAM (static random-access memory) would donate raw speed and low power consumption. And the EEPROM (electronically erasable programmable read-only memory) and its cousin, "Flash memory" would provide the special magic of being non-volatile: able to hang on to data when the power is switched off.

Such a chip would have numerous applications. It could be fitted in "smart cards "--the much-hyped devices that may one day replace the cash, keys and personal records that people currently carry in their pockets.

In fact, the FeRAM can store data for ten years or more without power. But the ferroelectric chips produced so far still have a laughably small storage capacity compared with even the ageing 64-megabit DRAMs used in today's personal computers. The biggest ferroelectric chips at the moment are 1-megabit experimental devices made by Fujitsu and NEC. Designers will have to learn to shrink their chips' innards dramatically, so that hundreds of times more data can be crammed in, if FeRAMs are to fulfil their destiny.

The only trouble is that ferroelectric chips come in two rival formats -- both invented in Colorado Springs in America. Hitachi, Rohm, Toshiba and Fujitsu have bought Ramtron International's PZT technology, which is based on the lead-zirconium titanate group of ferroelectric materials that have been around since the 1950s. Matsushita has taken out a licence from Symetrix, which uses a more advanced design based on so-called Y-1 super-lattice materials. These are layers composed of strontium-bismuth-tantalum oxide, and strontium-bismuth-niobium oxide. (10) Meanwhile, NEC is going ( A ) alone with a strontium-bismuth-tantalum oxide design of its own.

Ramtron (RMTR) who has several patents for their Trademarked FRAM (120 to be correct, although i'm not sure how importnat these are), has several licensing agreements. Besides the Japanese foursome, most of the world's non-Japanese heavyweights --including Samsung of South Korea and SGS-Thomson of Europe --are backing Ramtron's technology (IBM is also a customer).

Rohm also foresees FRAM's use in smart cards and other products that do not use a battery. The power from a magnetic field is all that is required to read/write/rewrite on the chip.

FRAM has miniscule power requirements, and is a nonvolitile type of memory that can be used to replace Flash memory. Because of it's tiny power requirements, it is an ideal chip for wireless devices such as cell phones, PDA's, laptops, etc.

The market for FRAM is expected to be $5 Billion for 2001 and $15 Billion by 2005.

The company has recently reinvented itself and is gearing up for the surge in FRAM demand. I have read that FRAM is expensive, so i'm not sure if it'll replace DRAM or flash ram anytime soon, but this is a very exciting technology IMO.

Here are some links for extra DD..

Just some info on FRAM

Samsung preps FRAM as possible flash replacement

A good primer from fujitsu
Print the post Back To Top
No. of Recommendations: 0
Here is another good link on FRAM.

Also, i was looking at Ramtron's annual report and as of right now the manufacturing costs of serial FRAM products are presently higher than competing EEPROM and Flash products. Ramtron and its strategic alliance partners are working to reduce such manufacturing costs. I'd think they'd be able to lower the costs..just my opinion, tho.

Even though this doesn't mean much, Ramtron has trademarked the word 'FRAM'.
Print the post Back To Top