Tuesday, 4 August 2009

Thought Processor INTEL

The Intel CE 2110 Media Processor combines a 1 GHz Intel XScale processing core with powerful audio-video processing, graphics and I/O components. Single chip solution is important as consumer electronics manufacturers need to accelerate development/production process.
The MPEG-2 and H.264 video codecs maximizes system-level performance by enabling the Intel XScale processor core to be used exclusively for applications. In addition to the Intel XScale processor core, this highly integrated consumer electronics platform building block includes an Intel Micro Signal Architecture DSP core for audio codecs, a 2D/3D graphics accelerator, hardware accelerators for encryption and decryption, comprehensive peripheral interfaces, analog and digital input/output, and a transport interface for ATSC/DVB input.



Laser wafer marking tracks IC production

The making of semiconductor integrated-circuit (IC) chips--at one time a labor-intensive operation in which silicon wafers were hand-carried from machine to machine, aligned by eye through a microscope, and tracked by careful technicians--has become a highly automated process where milliseconds count and glitches cannot be tolerated. If a wafer goes astray, or information on the number of manufacturing steps it has gone through disappears, an entire production line may have to be idled while troubleshooters are called in to figure out what went wrong.
To prevent such problems and to keep tabs on the manufacturing process itself, most modern IC fabrication facilities ("fabs") require that each wafer be labeled with its own identification (ID) mark in the form of a string of characters, a barcode, or a two-dimensional (2-D) matrix of pixels (Fab equipment then automatically tracks the wafer through its manufacturing stages to the point at which it is diced into individual IC chips. Any inspection data accumulated along the way can be unambiguously tied to the proper wafer.
Laser marking, with its combination of speed, permanence, and reliability, has become the standard means of marking wafers. Although the technology has been around since the 1970s, it has, through steady improvement and the advent of new applications, continued to serve the semiconductor industry

IC makers weigh improvements

Because silicon has a higher absorption for green light than for near-IR, most manufacturers of laser wafer markers now offer frequency-doubled solid-state lasers as an option--or, as in the case of NEC Corp., as standard equipment. The disadvantages of a frequency-doubled laser--lower power and higher cost--can be offset by improved marking performance resulting from the fact that energy absorption of the doubled light occurs closer to the wafer surface. But the choice between green and near-IR is not clear-cut. Because each IC chip maker has developed its own proprietary methods, what works well at one fab may not pass muster at another.
In the case of backside die marking, the consequence of silicon`s lower near-IR absorption is more obvious. Although opaque to the eye, a silicon wafer transmits enough of the Nd:YAG laser`s 1064-nm fundamental wavelength that a small amount of light can reach all the way to the underside of the die itself, potentially causing damage. But damage of this sort "is uncommon," says Downes of General Scanning. He notes that of all the chips being manufactured at fabs where backside marking is used, only one type of chip at one fab suffered performance degradation due to underside irradiation. Even so, General Scanning offers optional frequency doubling of its lasers, he says.
When operating at high power and slow scan speeds, laser wafer-marking systems are capable of digging pits and trenches in silicon with depths of from a few to more than 100 µm, called "hard" marks. But this sort of marking creates particles that contaminate and ruin chips. In addition, when used for backside die ID, hard marking can produce raised kerfs up to 30 µm high that prevent a finished chip from adequately contacting its heat sink.



Solderless Flip Chip Using Polymer Flip Chip ProcessesA reliable and manufacturable flip chip infrastructure continues to develop worldwide.

A reliable and manufacturable flip chip infrastructure continues to develop worldwide. Significant advances in equipment, processes for flip chip interconnect, and long term reliability of the flip chip assemblies are causing a shift from chip and wire interconnect to non-packaged direct chip attach.
Miniaturized packages, higher density electronics and higher speed are the motivating forces for the true chip size, low inductance electrical interconnection that flip chip offers. As shown in Table 1, the ability to form a high input-output (I/O) packaging concept with low contact resistance, low capacitance, and low lead inductance will drive the microelectronics industry conversion from chip and wire to flip chip.
Flip chip interconnect technology will become the ultimate surface mount (SMT) technique in the 21st century, replacing BGA, μBGA, and CSP, which are best categorized as transition packages. All of these will use flip chip for electrically attaching the integrated circuit (IC) to the package substrate, until cost and space needs require eliminating the package altogether.

The three basic technologies underlying most of the hundreds of flip chip interconnect techniques are anisotropic materials, metallic bump technology, and isotropic conductive polymers. The process and reliability information which follows here focuses on the isotropic conductive polymer approach, or PFC® process. This process uses silver (Ag) filled thermoset and thermoplastic polymers, in combination with stencil printing processes, to form polymer bump interconnects for flip chip integrated circuit (IC) devices.
The following discussion of under-bump metallization (UBM) over aluminum, bump formation processes, and overall reliability of flip chip devices compares the relative performance of the thermoset and thermoplastic polymers which form the primary electrical interconnection.

Sputtered UBM/Electroplated Solder

Electroplating of solder was developed as a less costly and more flexible method than evaporation. The UBM is typically an adhesion layer of titanium tungsten (TiW), a copper wetting layer, and a gold protective layer. The UBM is sputtered or evaporated over the entire surface of the wafer, providing a good conduction path for the electroplating currents.
Bumping begins with photopatterning and plating a copper minibump on the bump sites. This thick copper allows the use of high-tin eutectic solders without consuming the thin copper UBM layer. A second photopatterning and plating of the solder alloy over the minibump forms the solder bump. The photoresist is then removed from the wafer and the bump is reflowed to form a sphere.
Electroplated bumping processes generally are less costly than evaporated bumping. Electroplating in general has a long history and processes are well characterized. The UBM adheres well to the bond pads and passivation, protecting the aluminum pads. Plating can allow closer bump spacing (35 to 50 microns) than other methods of bump formation. Electroplating has become more popular for high bump count (>3,000) chips becasue of its small feature size and precision.
Plating bath solutions and current densities must be carefully controlled to avoid variations in alloy composition and bump height across the wafer. Plating generally is limited to binary alloys.

Solder Bump Flip Chip

This is the second in a series of flip chip tutorials intended for flip chip users and potential users. Tutorial #2 presents an overview of solder bump flip chip bumping and assembly processes. Concurrently, FlipChips Dot Com’s Technology News Updates present industry experts describing the newest developments in their fields; our Literature and Photo pages give supplemental material.

GENERAL
Flip chip assembly by means of solder connection to the bond pads was the first commercial use of flip chip, dating to IBM's introduction of flip chip in the 1960's. Solder bump has the longest production history, the highest current and cumulative production volumes, and the most extensive reliability data of any flip chip technology. Delco developed their solder bump processes in the 1970's; Delco Delphi now assembles over 300,000 solder bumped die per day for automotive electronics.
More recent solder bump flip chip process variations have lowered the manufacturing cost, widened flip chip applicability, and made solder bumped die and wafers available from several suppliers to the commercial market. This introductory survey discusses the operations performed in solder bumping and assembly, and describes several of the solder bump processes now commercially available. The references listed at the end of the tutorial provide details.

PROCESS OVERVIEW
The solder bump flip chip process may be considered as four sequential steps: preparing the wafer for solder bumping, forming or placing the solder bumps, attaching the bumped die to the board, substrate, or carrier, and completing the assembly with an adhesive underfill.

Under-Bump Metallization
The first step in solder bumping is to prepare the semiconductor wafer bumping sites on the bond pads of the IC's. This preparation may include cleaning, removing insulating oxides, and providing a pad metallurgy that will protect the IC while making a good mechanical and electrical connection to the solder bump and the board.
This under-bump metallization (UBM) generally consists of successive layers of metal with functions described by their names. The "adhesion layer" must adhere well to both the bond pad metal and the surrounding passivation, providing a strong, low-stress mechanical and electrical connection. The "diffusion barrier" layer limits the diffusion of solder into the underlying material

The Coming of Copper UBM

Several developments in the past few years have solved the numerous problems associated with using copper metal in place of aluminum as the IC interconnect metal. Copper is about three times more conductive than aluminum, and allows higher frequencies to be used with smaller line widths. Many fabricators are converting to copper not just for speed, but also for cost reduction. Thinner conductors allow closer spacing and smaller chips. The switch to copper allows many times more dies per wafer, and this is where the savings come from.
Since copper is much more compatible with bump metals than aluminum, the transition to copper is expected to boost flip chip technology, possibly by eliminating the UBM step. It remains to be seen what the final finish will be for copper ICs. The industry will continue to use wire bonding as the main interconnect method, so the final pad must be compatible with gold wire bonding.
IBM has indicated that its copper chips will have aluminum as the final pad layer to accommodate wire bonding, and this may become standard practice. However, aluminum can be removed easily, without affecting the copper underneath. Aluminum is an amphoteric metal that can be dissolved in both acid and base. Dilute caustic (sodium hydroxide) quickly removes aluminum. Many other reagents also can be used.
Thus, even if new copper chips come with an aluminum finish, a simple aqueous washing step will unveil the desired copper layer. In fact, the aluminum over copper would serve to protect the copper from oxidation. Gold over nickel could also be used on copper pads similar to PWB common finishes. This too would be a very good surface for most bumps.
Conductive adhesives would receive a real boost in the switch to copper because none are compatible with aluminum. Some of the conductive adhesives form reasonably stable junctions with bare copper, especially those using an oxide-penetrating mechanism. Even for those adhesives which are not suitable for bare copper, simple UBM methods could be used. Silver and other finishes can be applied to copper by electroless, maskless plating.
The advent of copper-based chip interconnection metallurgy undoubtedly will simplify FC fabrication in the near future. It will probably be possible to directly bond the copper pads with conductive adhesives. This simple processing ability would have a great impact on cost and infrastructure issues by eliminating UBM and maybe the bumping step. Assemblers could run the entire FC preconditioning and bonding process.

Critical Issues of Wafer Level Chip Scale Package (WLCSP)

ABSTRACT:Some of the critical issues of wafer level chip scale package (WLCSP) are mentioned and discussed in this investigation. Emphasis is placed on the cost analysis of WLCSP through the important parameters such as wafer-level redistribution, wafer-bumping, and wafer-level underfilling. Useful and simple equations in terms of these parameters are also provided. Only solder-bumped with pad-redistribution WLCSPs are considered in this study.

INTRODUCTION:There are at least two major reasons why directly attaching the solder bumped flip chip on organic substrates is not popular yet [1, 2]. Because of the thermal expansion mismatch between the silicon chip and the epoxy PCB, underfill encapsulant is usually needed for solder joint reliability. However, due to the underfill operation, the manufacturing cost is increased and the manufacturing throughput is reduced. In addition, the rework of an underfilled flip chip on PCB is very difficult, if it is not impossible.
The other reason is because the pitch and size of the pads on the peripheral-arrayed chips are very small and pose great demands on the supporting PCB. The high-density PCBs with sequential build-up circuits connected through microvias are not commonly available at reasonable cost yet.
Meantime, a new class of packaging called wafer level chip scale package (WLCSP) provides a solution to these problems [1 – 15]. There are many different kinds of WLCSP, for examples, eight different (ChipScale, EPIC, FCT, Fujitsu, Mitsubishi, National Semiconductor, Sandia National Laboratories, and ShellCase) companies’ WLCSP are reported in [2] and six different (EPS/APTOS, Amkor/Anam, Hyundai, FormFactor, Tessera, and Oxford) companies’ WLCSP are reported in [1]. Just like many other new technologies,

The infrastructure of WLCSP is not well established
The standard of WLCSP is not well established
WLCSP expertise is not commonly available
Bare wafer is not commonly available
Bare wafer handling is delicate
High cost for poor-yield IC wafers
Wafer bumping is still too costly
High cost for low wafer-bumping yield, especially for high-cost dies
Wafer-level redistribution is still too costly
High cost for low wafer-level redistribution yield, especially for high-cost dies
Troubles with System Makers if the die shrinks
Test at speed and burn-in at high temperature on a wafer are difficult
Single-point touch-up on the wafer is difficult
PCB assembly of WLCSP is more difficult
Solder joint reliability is more critical
Alpha particles produce soft errors by penetrating through the lead-bearing solder on WLCSP
Impact of lead-free solder regulations on WLCSP
Who should do the WLCSP? IC Foundries or Bump Houses?
What are the cost-effective and reliable WLCSPs and for what IC devices?
How large is the WLCSP market?
What is the life cycle of WLCSP?

WLCSP COSTS

Since 100% perfect wafers cannot be made at high volume today, the true IC chip yield (YT) plays the most important role in cost analysis. Also, the physical possible number of undamaged chips (Nc) stepped from a wafer is need for cost analysis, since (YTNc) is the number of truly good die on a wafer. Nc is given by [1, 2, 16]
where
A = xy (2)
and
In Equations (1) – (3), x and y are the dimensions of a rectangular chip (in millimeters, mm) with x no less than y; q is the ratio between x and y; f is the wafer diameter (mm); and A is the area of the chip (in square millimeters, mm2). For example, for a 200 mm wafer with A = 10 x 10 = 100 mm2, then Nc ~

Wafer Redistribution Costs

Wafer-level redistribution is the heart of the WLCSPs. The cost of wafer-level redistribution is affected by the true yield (YT) of the IC chip, the wafer-level redistribution yield (YR), and the good die cost (CD). The actual wafer-level redistribution cost per wafer (CR) is:
CR=CWR+(1–YR)YTNCCD (4)
where CWR is the wafer-level redistribution cost per wafer (ranging from $50 to $200), YR is the wafer-level redistribution yield per wafer, CD is the good die cost (not the cost of an individual die on the wafer), Nc is given in Equation (1), and YT is the true IC chip yield after at-speed/burn-in system tests (or individual die yield). Again, it can be seen that the actual wafer-level redistribution cost per wafer depends not only on the wafer-level redistribution cost per wafer but also on the true IC chip yield per wafer, wafer-level redistribution yield per wafer, and good die cost.
Wafer-level redistribution yield (YR) plays a very important role in WLCSP. The wafer-level redistribution yield loss (1-YR) could be due to: (1) more process steps; (2) wafer breakage; (3) wafer warping; (4) process defects such as spots of contamination or irregularities on the wafer surface; (5) mask defects such as spot, hole, inclusion, protrusion, break, and bridge; (6) feature-size distortions; (7) pattern mis-registration; (8) lack of resist adhesion; (9) over etch; (10) undercutting; (11) incomplete etch; and (12) wrong materials. It should be noted that wafer-level redistribution are not reworkable. It has to be right the first time, otherwise, someone has to pay for it!
The uses of Equations (1) and (4) are shown in the following examples. If the die size of a 200 mm wafer is 100 mm2, true IC chip yield per wafer is 80% (since the importance of YT has been shown in [16, 17], only one value of YT will be consider in this study), wafer-level redistribution yield per wafer is 90%, wafer-level redistribution cost per wafer is $100, and the die cost is $100 (e.g., microprocessors), then from Equation (1), Nc = 255, and from Equation (4), the actual wafer-level redistribution cost per wafer is $2140. For the same size of wafer if the die cost is $5 (e.g., memory devices), then the actual wafer-level redistribution cost per wafer is $202. It is noted that for both cases, the actual wafer-level redistribution cost per wafer is much higher than the wafer-level redistribution cost (CWR = $100)!
On the other hand, if the wafer-level redistribution yield is increased from 90% to 99%, then the actual cost for redistributing the microprocessors wafer is reduced from $2140 to $304 and for redistributing the memory wafer is reduced from $202 to $110.2. Thus, wafer-level redistribution yield plays an important role in the cost of wafer-level redistribution and the wafer-level redistribution houses should stride to make YR > 99%, especially for expensive good dies

Wafer Bumping Costs

Wafer bumping is the heart of solder-bumped WLCSPs. The cost of wafer bumping is affected by YT, CD, YR and the wafer-bumping yield (YB). The actual wafer bumping cost per wafer (CB) is:
CB=CWB+(1–YB)YRYTNCCD (5)
where CWB is the wafer bumping cost per wafer (ranging from $25 to $250), YB is the wafer-bumping yield per wafer, YR is the wafer-level redistribution yield per wafer, CD is the good die cost, Nc is given in Equation (1), and YT is the true IC chip yield after at-speed/burn-in system tests (or individual die yield). Again, it can be seen that the actual wafer bumping cost per wafer depends not only on the wafer-bumping cost per wafer but also on the true IC chip yield per wafer, wafer-bumping yield per wafer, good die cost, and wafer-level redistribution yield per wafer.
Just like YR, wafer bumping yield (YB) plays a very important role in wafer bumping. The wafer bumping yield loss (1-YB) could be due to: (1) wrong process;, (2) different materials; (3) too tall or short of a bump height; (4) not enough shear strength; (5) un-even shear strength; (6) broken wafers or dies; (7) solder bridging; (8) damaged bumps; (9) missing bumps; and (10) scratch of the wafer.
For the pervious example, if the wafer-bumping yield per wafer is 90% and wafer bumping cost per wafer is $120, then the actual wafer bumping costs per (the microprocessors) wafer are, respectively, $1956 if YR = 90% and $2139.6 if YR = 99%, and the actual wafer bumping costs per (the memory) wafer are, respectively, $211.8 if YR = 90% and $220.98 if YR = 99%. Again, it should be noted that the actual wafer-bumping cost per wafer is much higher than the wafer-bumping cost (CWB = $120).
On the other hand, if the wafer-bumping yield is increased from 90% to 99%, then the actual costs for bumping the microprocessors wafer are, respectively, $303.6 if YR = 90% and $321.96 if YR = 99%, and the actual costs for bumping the memory wafer are, respectively, $129.18 if YR = 90% and $130.1 if YR = 99%. Thus, wafer-bumping yield plays an important role in the cost of wafer bumping and the wafer bumping houses should stride to make YBYR > 99%, especially for expensive good dies. If there is no wafer-level redistribution, then there is no wafer redistribution yield loss, i.e., YR = 1, then Equation (5)

SUMMARY

More than 20 different critical issues of WLCSP have been mentioned. The most important issue (cost) of WLCSP has been analyzed in terms of the true IC chip yield, wafer-level redistribution yield, wafer-bumping yield, wafer-level underfill yield, and die size and cost. Also, useful equations in terms of these parameters have been presented and demonstrated through examples. Some important results are summarized as follows.
IC chip yield (YT) plays the most important role in WLCSP. If YT is low for a particular IC device, then it is not cost-effective to house the IC with WLCSP, unless it is compensated for by performance, density, and form factor.
Wafer-level redistribution yield (YR) plays the second most important role in WLCSP. Since this is the first post wafer processing after the IC FAB, the wafer-level redistribution houses should stride to make YR > 99% (99.9% is preferred). Otherwise, it will make the subsequent steps very expensive by wasting the material and process on the damage dies.
Wafer-bumping yield (YB) plays the third most important role in WLCSP. The wafer bumping house should strive to make YRYB > 99% (99.9% is preferred) to minimize the hidden cost, since they cannot afford to damage the already redistributed good dies.
Based on cost and process points of view, wafer-level underfill is not a good idea for solder-bumped flip chip on low-cost substrates.