摘要:
A method for providing a through wafer connection to an integrated circuit silicon package. A hole is first created in the silicon package with an inner surface area extending from the bottom surface of the silicon package to the top surface of the silicon package. The hole is created by one of two methods. The first involves mechanical drilling with a diamond bit rotated at a high rate of speed. The second involves ultrasonically milling utilizing a slurry and steel fingers. The inner surface area of the hole is covered with an insulating material to insulate the conductive material which is later deposited and to serve as a diffusion barrier, then a seed material is placed in the hole. Finally, the hole is filled with a conductive material which is utilized to provide large power inputs or signaling connections to the integrated circuit chips.
摘要:
A silicon wafer is etched to form a first and second series of guidance features. The features of the first series are larger than and surround the features of the second series. The second series is clustered into groups and a hole is formed in the center of each group. The wafer is designed to integrate a silicon package having preformed contacts with a plurality of silicon-based chips. The package and each chip has a series of guidance recesses which correspond to the guidance features of the first and second series, respectively. One chip is placed on top of each group of the second series, and the package is placed on top of the first series. The recesses in the package and chips will precisely align with and slidingly engage the upper ends of the features. Since the features of the first series are larger than those of the second series, there is a clearance between the package and the chips. In the final stage of assembly, the package is restrained from movement and a pin is inserted through each hole in the wafer to force the chips into contact with the contacts on the package. Heat is then applied to fuse solder balls on the chips with the contacts to form a complete and finished assembly.
摘要:
A zero/one anticipator having an integrated sign selector is disclosed. A leading zeros string and a leading ones string are generated by examining carry propagates, generates, and kills of two adjacent bits of two input operands to an adder within a floating-point processor. The leading zeros string is for a positive sum, and the leading ones string is for a negative sum. A normalization shift amount is then determined from the leading zeros string and the leading ones string. A sign of a sum of the two input operands is then determined separately but concurrently with the normalization shift amount determination process. The sign is then utilized to select either the positive sum or the negative sum for a proper normalization shift amount.
摘要:
A leading-zero anticipator having an independent sign bit determination module is disclosed. An apparatus for anticipating leading zeros for an adder within a floating-point processor includes a leading-zero anticipator and a sign determination module. The leading-zero anticipator generates a leading zeros string and a leading ones string by examining carry propagates, generates, and kills of two adjacent bits of two input operands of the adder. The leading zeros string is intended for a positive sum, and the leading ones string is intended for a negative sum. Independent of the leading-zero anticipator, the sign determination module determines a sign of the output of the adder in concurrence with the operations within the leading-zero anticipator.
摘要:
A low latency fused multiply-adder for adding a product of a first binary number and a second binary number to a third binary number is disclosed. The low latency fused multiply-adder includes a partial product generation module, a partial product reduction module, and a carry propagate adder. The partial product generation module generates a set of partial products from the first binary number and the second binary number. Coupled to the partial product generation module, the partial product reduction module combines the set of partial products with the third binary number to produce a redundant Sum and a redundant Carry. Finally, the carry propagate adder adds the redundant Sum and the redundant Carry to yield a Sum Total.
摘要:
A processor includes execution resources and condition code logic. The execution resources execute an arithmetic or logical instruction by arithmetically or logically combining at least two operands. Concurrent with the execution of the arithmetic or logical instruction by the execution resources, the condition code logic determines less than, greater than, and equal to condition code bits associated with the result of the arithmetic or logical instruction. In one embodiment, the condition code logic includes a single computation stage that receives as inputs individual bit values of bit positions within first and second operands and logically combines the individual bit values. The single computation stage outputs, for each bit position, propagate, generate, and kill signals that collectively indicate values for the less than, greater than, and equal to condition code bits. One or more merging stages coupled to the computation stage then merge the propagate, generate, and kill signals into output signals that set the condition code bits. The condition code logic is also capable of receiving externally computed condition code bits associated with complex instructions and utilizing such condition code bits to produce the output signals.
摘要:
A 6-to-3 carry-save binary adder is disclosed. The 6-to-3 carry-save adder includes a means for receiving six data inputs and a means for simultaneously adding the six data inputs to generate a first data output, a second data output, and a third data output. The first data output is a SUM output, the second data output is a CARRY—2 output, and the third data output is a CARRY—4 output.
摘要:
A superscalar computer architecture for executing instructions out-of-order, comprising a multiplicity of execution units, a plurality of registers, and a register renaming circuit which generates a list of tags corresponding to specific registers that are not in use during loading of a given instruction. A table is constructed having one bit for each register per instruction in flight. The entries in the table may be combined in a logical OR fashion to create a vector that identifies which registers are in use by instructions that are in flight. Validity bits can also be generated to indicate validity of the generated tags, wherein a generated tag is invalid only if an insufficient number of registers are available during loading of the given instruction. The execution units are preferably pipelined.
摘要:
A method for anticipating leading zeros/ones in a floating-point processor is disclosed. A leading zeros string and a leading ones string is generated by examining carry propagates, generates, and kills of two adjacent bits of two input operands to an adder within a floating-point processor. The leading zeros string is for a positive sum, and the leading ones string is for a negative sum. A normalization shift amount is calculated directly and concurrently from the leading zeros string and the leading ones strings prior to a determination of a sign of an output of the positive sum and the negative sum.
摘要:
A processor includes at least an execution unit that executes an instruction by performing an operation indicated by the instruction utilizing one or more operands and condition code logic that determines less than, greater than, and equal to condition code bits associated with the instruction concurrently with execution of the instruction by the execution unit. In one embodiment, the condition code logic includes a single computation stage that receives as inputs individual bit values of bit positions within first and second operands and logically combines the individual bit values. The single computation stage outputs, for each bit position, propagate, generate, and kill signals that collectively indicate values for the less than, greater than, and equal to condition code bits. One or more merging stages coupled to the computation stage then merge the propagate, generate, and kill signals into output signals that set the condition code bits.