摘要:
An exemplary embodiment of the invention is a method and apparatus for delaying the start of an array built-in self-test (ABIST) until after the ABIST memory arrays have been started. The length of the delay is determined by the value in a programmable delay located on the integrated circuit. The initiation of the ABIST test is delayed by the time specified in the programmable delay.
摘要:
An array built-in self test (ABIST) system includes a first latch having a first data input, a first scan input and first output and a second latch having a second data input, a second scan input and a second output. The system also includes a first ABIST logic block coupled to the first output that compares a first expected value with a first data value received at the first data input and provided to the first ABIST logic block after a first clock is applied to the first latch. The system also includes a second ABIST logic block coupled to the second output that compares a second expected value with a second data value received at the second data input and provided to the second ABIST logic block after a second clock is applied to the second latch.
摘要:
A computing system is provided which includes a processor having a cache memory. The cache memory includes a plurality of independently configurable subdivisions, each subdivision including a memory array. A service element (SE) of the computing system is operable to cause a built-in-self-test (BIST) to be executed to test the cache memory, the BIST being operable to determine whether any of the subdivisions is defective. When it is determined that one of the subdivisions of the cache memory determined defective by the BIST is non-repairable, the SE logically deletes the defective subdivision from the system configuration, and the SE is operable to permit the processor to operate without the logically deleted subdivision. The SE is further operable to determine that the processor is defective when a number of the defective subdivisions exceeds a threshold.
摘要:
A soft-fust test algorithm is distributed on-chip from an ABSIT engine through an LSSD shift register chain to dynamically evaluate a plurality of arrays with redundancy compensation for bad elements and repair those that are fixable. Using single-bit MISR error evaluation an ABSIT test sequence is executed concurrently on all arrays through the shift register chain. If any arrays are in error, redundancy compensation is employed and the ABIST test is repeated for all possible array redundant combinations until a functional configuration for each array is identified or all possible redundant combinations have been tried. Once functioning array configurations are verified, the associated soft-fuse states can be used to blow fuses and/or extracted for further system setup, permanent fuse-blowing and yield analysis. Multiple shift register chains driven by separate ABIST engines may be required to test all arrays on a chip.
摘要:
An apparatus and method for protecting a computer system from array reliability failures uses Array Built-In Self-Test logic along with code and hardware to delete cache lines or sets that are defective, identify corresponding fuse repair values, proactively call home if spare fuses are not available, schedule soft fuse repairs for the next system restart, schedule line deletes at the next restart, store delete and fuse repairs in a table (tagged with electronic serial id, timestamp of delete or ABIST fail event, address, and type of failure) and proactively call home if there were any missed deletes that were not logged. Fuse information can also be more permanently stored into hardware electronic fuses and/or EPROMs. During a restart, previous repairs are able to be applied to the machine so that ABIST will run successfully and previous deletes to be maintained with checking to allow some ABIST failures which are protected by the line deletes to pass.
摘要:
An array built-in self test (ABIST) system includes a first latch having a first data input, a first scan input and first output and a second latch having a second data input, a second scan input and a second output. The system also includes a first ABIST logic block coupled to the first output that compares a first expected value with a first data value received at the first data input and provided to the first ABIST logic block after a first clock is applied to the first latch. The system also includes a second ABIST logic block coupled to the second output that compares a second expected value with a second data value received at the second data input and provided to the second ABIST logic block after a second clock is applied to the second latch.
摘要:
A computing system is provided which includes a processor having a cache memory. The cache memory includes a plurality of independently configurable subdivisions, each subdivision including a memory array. A service element (SE) of the computing system is operable to cause a built-in-self-test (BIST) to be executed to test the cache memory, the BIST being operable to determine whether any of the subdivisions is defective. When it is determined that one of the subdivisions of the cache memory determined defective by the BIST is non-repairable, the SE logically deletes the defective subdivision from the system configuration, and the SE is operable to permit the processor to operate without the logically deleted subdivision. The SE is further operable to determine that the processor is defective when a number of the defective subdivisions exceeds a threshold.
摘要:
An array built-in self test system has a scannable memory elements and a controller which, in combination, allow self test functions (e.g. test patterns, read/write access, and test sequences) to be modified without hardware changes to the test logic. Test sequence is controlled by logical test vectors, which can be changed, making the task of developing complex testing sequences relatively easy.
摘要:
The output register of an array and the Multiple Input Signature Register (MISR) logic is implemented with one set of L1/L2 master/slave latches and single additional slave latch. This new combined logic uses less critical area on a chip without a performance impact on the array access time or circuit testing.
摘要:
A computing system is provided which includes a processor having a cache memory. The cache memory includes a plurality of independently configurable subdivisions, each subdivision including a memory array. A service element (SE) of the computing system is operable to cause a built-in-self-test (BIST) to be executed to test the cache memory, the BIST being operable to determine whether any of the subdivisions is defective. When it is determined that one of the subdivisions of the cache memory determined defective by the BIST is non-repairable, the SE logically deletes the defective subdivision from the system configuration, and the SE is operable to permit the processor to operate without the logically deleted subdivision. The SE is further operable to determine that the processor is defective when a number of the defective subdivisions exceeds a threshold.