摘要:
A loopback test to test a communication link for a layered interface where in a master agent programs the electrical parameters for the slave agent, such as, the offset, timing, and current compensation with a loopback control register. The slave and master agent to support an entry into the loopback test based on detection of a header within a packet. The slave and master agent to support exit out of the loopback test based on whether the loop count is finite.
摘要:
A technique for promoting determinism among bus agents within a point-to-point (PtP) network. More particularly, embodiments of the invention relate to techniques to compensate for link latency, data skew, and clock shift within a PtP network of common system interface (CSI) bus agents.
摘要:
A software self test engine is executed from a cache of a processor. The software self test engine is executed using an execution engine of the processor to perform a physical layer self test. The physical layer self test is performed by transmitting a test vector from the execution engine under control of the self test engine to an input/output (“I/O”) unit of the processor along a datapath coupling the execution engine to the I/O unit. The test vector is transmitted along a loop back path including the I/O unit and the datapath to test a hardware device along the loop back path.
摘要:
A method and apparatus for retraining skew compensation in an interface is presented. In one embodiment, a retraining interval is determined, and counters in the transmitting agent and receiving agent count up until the retraining interval is reached. A tracking unit used to select one of several interpolated clocks may then be powered up, and a special retraining phit may be sent across the interface. During the retraining process, the transfer of flits into and out of the flow-control mechanism may be inhibited. When the retraining process is finished, the tracking unit may be powered down.
摘要:
The current method and apparatus provides a novel approach to manage the power consumption of a high speed I/O interface by selectively turning off non-essential portions of the interface. Here only part of the interface is powered off as compared to the whole interface being turned off. From the upper layers (protocol/system) perspective, the interface is always “on”. Thus, this mechanism reduces link power by selectively turning off portions of the link, yet allowing for fast wake up in an interface power management architecture.
摘要:
A method and apparatus for a linearized output driver and terminator is described. In one embodiment the method includes forming a gate electrode on a substrate, the portion of the substrate covered by the gate electrode defining a channel. The method further includes forming a first source/drain doped region on laterally opposed sides of the gate electrode in the substrate. The method also includes forming a spacer on laterally opposed sides of the gate electrode on the substrate. The method also includes forming a linearized drain contact region at a location within the first source/drain doped region sufficiently distant from the gate electrode to define a series resistor in the first source/drain doped region disposed between the gate electrode and the linearized drain contact area based on an expected resistivity of the source/drain doped region, the series resistor coupled electrically to the channel.
摘要:
A multi layer printed circuit board with a 3-load topology is disclosed. First, second, and third integrated circuit (IC) printed wiring board packages having first, second, and third sets of terminals respectively are mounted on opposite sides of the board so that the second set of terminals are directly opposite the third set of terminals. Each package contains an IC die coupled to the respective set of terminals. The IC die in the first package is substantially identical to the one contained in the second package, and different than the one contained in the third package. For improved fanout of the metal lines that interconnect the first package to the second and third packages, each of the first, second, and third sets of terminals in the packages is arranged in substantially a U-shape. Each set of terminals has the same set of signal assignments of a parallel bus implemented by metal lines in the board. The 3-load topology is particularly useful for personal computer motherboard units having twin processors and a bridge chip set, yielding a motherboard having significantly lower number of metal layers, a faster bus and significantly improved noise margin, all with high density IC packages on a wide parallel bus.