F E B R U A R Y 2 0 1 0 i s s u e n ° 1 4
N e w s l e t t e r o n 3 D I C , T S V , W L P & E m b e d d e d T e c h n o l o g i e s
C O M P A N Y V I S I O N
Elpida looks to TSV to move beyond the DRAM business
Elpida Memory, Inc. is trying to diversify beyond the memory business to higher value-added systems by using its own
through silicon via technology to make vertically interconnected devices. Yole Développement partner Nikkei Microdevices
asked Elpida Director and CTO Takao Adachi about the company’s new business strategy.
N
ikkei Micro Devices: We hear you’ve 16 or 32 bits to as much as 1000 or 2000 bit width. about six months after we start initial production
developed an 8G multilayer DRAM using So we can achieve far faster graphic management of DRAMs we develop in Japan. However, we
TSV. circuits, or, alternatively, can greatly lower power are thinking of focusing on Japanese partners for
Takao Adachi: The DRAM we developed is a consumption. the TSV system solutions business, and doing
nine-layer device, with eight 1G DRAM chips and the whole process from design to production in
an interposer layer. In 2010 we are developing a NMD: So TSV means that memory will determine Japan. We are currently doing the design at our
16G device composed of eight stacked 2G chips, the system capabilities. development center in Kanagawa, the DRAM and
which is currently very close to the commercial TA: All DRAM makers are typically whipped about TSV processing at the Hiroshima plant, and the
product stage. However, our aim in developing by capex costs and spot market prices. We win or packaging at Akita Elpida. We have capacity for
TSV stacking technology is not just to make higher lose on production capacity and price. We now aim 10,000 300mm wafers per month when demand
density memory devices. We intend to use this key to establish a better position in the electronics value warrants.
technology to make systems solutions, by stacking chain.
memory together with other kinds of chips like logic, NMD: Is the system solutions business easier
sensors, and RF devices. We aim to make systems NMD: What are you developing in logic and than the DRAM business?
chips with TSV instead of on-chip integration. other areas besides memory? TA: The problem is that the DRAM supplier is so
TA: We are forming cooperative relationships with far removed from the final customer. Without direct
NMD: Usually it had been the logic makers that other companies to supply the other chips for our contact with the end user, we cannot respond rapidly
have developed system solutions. stacked systems devices. We are not planning to enough to the market demands. By being in the
TA: Up to now it has been primarily logic makers develop our own logic or RF or sensors. But in order system business, in direct contact with the end users,
that have taken the leading role and come to the to offer our own system solution, I think we will need we can make sure we’re developing the right products.
memory suppliers for components. Memory to develop the software ourselves.
makers couldn’t take the lead, because memory NMD: What’s the progress in solving the
chips were all JEDEC standard products. But TSV NMD: Will the partners include companies from technical problems of interconnecting
technology changes all that. By using TSV we can Taiwan and elsewhere overseas? multilayer device stacks with TSVs?
easily expand the DRAM I/O interface bus beyond TA: We currently move volume production to Taiwan TA: The devices we will ship in 2010 are commercial
Elpida Memory TSV Roadmap
Elpida Memory TSV roadmap. The 8G chip developed in 2009 uses a 70nm generation process. The 16G product slated for 2010 uses a 50nm process.
(Source: Elpida Memory, Nikkei Microdevices)
8
Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20