Nano and SEMI Mix it up in Chi Town
Next week I will be blogging from the Windy City, at the NanoCommerce and SEMI NanoForum (www.nanocommerceforum.com). This is not just another nanoconference because it mixes talks from the leading lights of the nanotech with those of big shots from IBM, Motorola, Texas Instruments, ASML etc. The conference will feature: Northwestern’s Chad Mirkin, founder of Nanosphere and NanoInk; James von Ehr, Founder and CEO of Zyvex; Jim O’ Conner, Vice President Technology, Motorola; David Macdonald, CEO of Nanomix; and the Big Daddy of Dendrimers, Don Tomalia, among others.
The semiconductor industry forms a big part of the famous one trillion dollar nanotech industry predicted by NSF, a figure that has been seen and raised by Lux Research. Of course, in some ways, the semiconductor industry is already nanotech-enabled, in that nanoparticles are used to polish silicon wafers and chips, and lithographic feature sizes on commercial chips have now cracked the 100 nm barrier. But this is all normal, everyday top-down engineering. It is not the type of quantumly weird, molecular scale, bottoms-up stuff that the elite would refer to as true nanotech.
But chips of the future are going to be different. The semiconductor understands that the way of the future is down still further in scale: Moore’s law demands it.
Memory
The first “true nanotech” chips will probably be for memory storage. Today, the memory market is fragmented because of the various types of memory required by different products.
DRAM, or dynamic access memory, is the type of memory chip used in computers and accounts for the biggest market, about $24 billion in 2004. SRAM, or static random access memory, is a conventional memory chip that is faster than DRAM but lower in density; it used is for operations such as digital signal processing in cellphones and caches in computers, where speed is critical. This was about a $3 billion dollar market in 2004. Flash memory is non-volatile, which means it doesn’t go away when the power goes off. Flash memory is used in cell phones to store data and in memory sticks, personal digital assistants (PDAs), MP3 players and digital cameras. This was a $12 billion market in 2004 and growing rapidly.
There are two types of nanotech memory chips, called MRAM and NRAM. that haven’t made it to prime time yet, but have the potential to replace all the others.
Ever since Thomas Edison created the electric light, the electronics industry has relied on the fact that negatively charged electrons flow naturally through a metallic conductor toward a positively charged pole. MRAM (for magnetoresistive RAM), on the other hand, relies on a quantum property of electrons called “spin.” Spin comes in two flavors, ingeniously called “up” and “down.” MRAM uses these two states to encode the 1’s and 0’s of computer memory. According to Daniel Baker, CEO of NVE Corp., MRAM will have the speed of SRAM, the density of DRAM, and the nonvolatility of flash memory. Therefore, the whole $40 billion memory market is a target.
NVE Corp. held the early lead in MRAM development and has licensed its technology to Cypress Semiconductor and Freescale Semiconductor (by way of Motorola). Both Freescale and Cypress have produced modest quantities of MRAM chips. But NVE Corp. received a setback to its plans for monetizing its technology when Cypress announced in April that it would discontinue MRAM development. Moreover, a Who’s Who of semiconductor manufacturers, including Fujitsu, Hewlett Packard, IBM, Infineon, NEC, Samsung, Sony, Taiwan, and Toshiba are reportedly interested in developing their own versions of MRAM.
NRAM will be explained in the next post….
The semiconductor industry forms a big part of the famous one trillion dollar nanotech industry predicted by NSF, a figure that has been seen and raised by Lux Research. Of course, in some ways, the semiconductor industry is already nanotech-enabled, in that nanoparticles are used to polish silicon wafers and chips, and lithographic feature sizes on commercial chips have now cracked the 100 nm barrier. But this is all normal, everyday top-down engineering. It is not the type of quantumly weird, molecular scale, bottoms-up stuff that the elite would refer to as true nanotech.
But chips of the future are going to be different. The semiconductor understands that the way of the future is down still further in scale: Moore’s law demands it.
Memory
The first “true nanotech” chips will probably be for memory storage. Today, the memory market is fragmented because of the various types of memory required by different products.
DRAM, or dynamic access memory, is the type of memory chip used in computers and accounts for the biggest market, about $24 billion in 2004. SRAM, or static random access memory, is a conventional memory chip that is faster than DRAM but lower in density; it used is for operations such as digital signal processing in cellphones and caches in computers, where speed is critical. This was about a $3 billion dollar market in 2004. Flash memory is non-volatile, which means it doesn’t go away when the power goes off. Flash memory is used in cell phones to store data and in memory sticks, personal digital assistants (PDAs), MP3 players and digital cameras. This was a $12 billion market in 2004 and growing rapidly.
There are two types of nanotech memory chips, called MRAM and NRAM. that haven’t made it to prime time yet, but have the potential to replace all the others.
Ever since Thomas Edison created the electric light, the electronics industry has relied on the fact that negatively charged electrons flow naturally through a metallic conductor toward a positively charged pole. MRAM (for magnetoresistive RAM), on the other hand, relies on a quantum property of electrons called “spin.” Spin comes in two flavors, ingeniously called “up” and “down.” MRAM uses these two states to encode the 1’s and 0’s of computer memory. According to Daniel Baker, CEO of NVE Corp., MRAM will have the speed of SRAM, the density of DRAM, and the nonvolatility of flash memory. Therefore, the whole $40 billion memory market is a target.
NVE Corp. held the early lead in MRAM development and has licensed its technology to Cypress Semiconductor and Freescale Semiconductor (by way of Motorola). Both Freescale and Cypress have produced modest quantities of MRAM chips. But NVE Corp. received a setback to its plans for monetizing its technology when Cypress announced in April that it would discontinue MRAM development. Moreover, a Who’s Who of semiconductor manufacturers, including Fujitsu, Hewlett Packard, IBM, Infineon, NEC, Samsung, Sony, Taiwan, and Toshiba are reportedly interested in developing their own versions of MRAM.
NRAM will be explained in the next post….
<< Home