Matsushita and Intel started mass-producing 45nm chips in late 2007, and AMD started production of 45nm chips in late 2008, while IBM, Infineon, Samsung, and Chartered Semiconductor have already completed a common 45nm process platform. At the end of 2008, SMIC was the first China-based semiconductor company to move to 45nm, having licensed the bulk 45nm process from IBM. In 2008, TSMC moved on to a 40nm process.
Many critical feature sizes are smaller than the wavelength of light used for lithography (i.e., 193nm and 248nm). A variety of techniques, such as larger lenses, are used to make sub-wavelength features. Double patterning has also been introduced to assist in shrinking distances between features, especially if dry lithography is used. It is expected that more layers will be patterned with 193nm wavelength at the 45nm node. Moving previously loose layers (such as Metal 4 and Metal 5) from 248nm to 193nm wavelength is expected to continue, which will likely further drive costs upward, due to difficulties with 193nm photoresists.
High-κ dielectric
Chipmakers have initially voiced concerns about introducing new high-κ materials into the gate stack, for the purpose of reducing leakage current density. As of 2007, however, both IBM and Intel have announced that they have high-κ dielectric and metal gate solutions, which Intel considers to be a fundamental change in transistor design.[1]NEC has also put high-κ materials into production.
Technology demos
In 2004, TSMC demonstrated a 0.296-square-micrometre 45nm SRAM cell. In 2008, TSMC moved on to a 40nm process.[2]
In January 2006, Intel demonstrated a 0.346-square-micrometre 45nm node SRAM cell.
In April 2006, AMD demonstrated a 0.370-square-micrometre 45nm SRAM cell.
In November 2006, UMC announced that it had developed a 45nm SRAM chip with a cell size of less than 0.25-square-micrometre using immersion lithography and low-κ dielectrics.
Intel shipped its first 45nm processor, the Xeon 5400 series, in November 2007.
Many details about Penryn appeared at the April 2007 Intel Developer Forum. Its successor is called Nehalem. Important advances[4] include the addition of new instructions (including SSE4, also known as Penryn New Instructions) and new fabrication materials (most significantly a hafnium-based dielectric). Intel's 45nm process has a transistor density of 3.33 million transistors per square millimeter (MTr/mm2).[5]
AMD released its Sempron II, Athlon II, Turion II and Phenom II (in generally increasing order of performance), as well as Shanghai Opteron processors using 45nm process technology in late 2008.
The Xbox 360 S, released in 2010, has a Xenon processor fabricated in a 45nm process.[6]
At IEDM 2007, more technical details of Intel's 45nm process were revealed.[8]
Since immersion lithography is not used here, the lithographic patterning is more difficult. Hence, a line-cutting double patterning method is used explicitly for this 45nm process. Also, the use of high-κ dielectric dielectrics is introduced for the first time, to address gate leakage issues. For the 32nm node, immersion lithography will begin to be used by Intel.
160nm gate pitch (73% of 65nm generation)
200nm isolation pitch (91% of 65nm generation) indicating a slowing of scaling of isolation distance between transistors
Extensive use of dummy copper metal and dummy gates[9]
35nm gate length (same as 65nm generation)
1nm equivalent oxide thickness, with 0.7nm transition layer
Gate-last process using dummy polysilicon and damascene metal gate
Squaring of gate ends using a second photoresist coating[10]
9 layers of carbon-doped oxide and Cu interconnect, the last being a thick "redistribution" layer
1.07mA/μm pFET drive current, 51% faster than 65nm generation, with higher hole mobility due to increase from 23% to 30% Ge in embedded SiGe stressors
In a 2008 Chipworks reverse-engineering,[11] it was disclosed that the trench contacts were formed as a "Metal-0" layer in tungsten serving as a local interconnect. Most trench contacts were short lines oriented parallel to the gates covering diffusion, while gate contacts where even shorter lines oriented perpendicular to the gates.
It was recently revealed[12] that both the Nehalem and Atom microprocessors used SRAM cells containing eight transistors instead of the conventional six, in order to better accommodate voltage scaling. This resulted in an area penalty of over 30%.