摘要:
In a computer system comprising a plurality of subsystems, interconnected by a bus comprising bit drivers and bit receivers, data words are transmitted on the bus in the form of code words. The code words are formulated such that the number of bits of the bus which changes with the transmission of successive code words is minimized. A switching code, comprising one or more bits, defines a plurality of napping codes and a data word to be transmitted is mapped by use of the mapping codes to a plurality of code words. One of the plurality of code words differing from a previously transmitted code word in the least number of bit positions is selected. The selected code words is transmitted, together with a switching code, identifying the mapping from which the transmitted code word was generated. At the receiving end of the bus, the switching code is decoded to identify the mapping used in creating the code word. Using the identified mapping, the original data word is recovered.
摘要:
A programmable, timing signal generator propagates a digital wave along a delay chain comprised of series connected inverters that has sufficient stages that the edge of the wave will not propagate to the end during one system clock cycle time. The delay chain is sampled once each clock cycle, and the point to which the wave has progressed is encoded. Programmable, fine leading and fine trailing edge registers store, as a fraction of the clock cycle time, the desired locations of the timing signal leading and trailing edge relative to the clock edges.
摘要:
A non-buffered, noise-immune transmission gate latch for high performance applications is disclosed. The latch data input circuit contains an additional PFET for pulling up the gate of a transmission PFET and an additional NFET to pull down the gate of the transmission NFET to prevent the transmission gate from inadvertently opening when noise is coupled into the data input node. When data input voltage rises to Vdd+Vtp, the additional PFET begins to turn ON, and an inverted clock node is pulled above Vdd. The higher inverted clock node voltage is coupled to the gate of the transmission PFET and thereby prevents the transmission PFET from inadvertently turning ON. When data input voltage drops to -Vtn, the additional NFET begins to turn ON, and a clock node Is pulled below ground. The lower clock node voltage is coupled to the gate of the transmission NFET and thereby prevents the transmission NFET from inadvertently turning ON. Thus two additional transistors provide the necessary noise immunity.
摘要:
A programmable, timing signal generator propagates a digital wave along a delay chain comprised of series connected inverters that has sufficient stages that the edge of the wave will not propagate to the end during one system clock cycle time. The delay chain is sampled once each clock cycle, and the point to which the wave has progressed is encoded. Programmable, fine leading and fine trailing edge registers store, as a fraction of the clock cycle time, the desired locations of the timing signal leading and trailing edge relative to the clock edges.
摘要:
Low-conductance and high-conductance IV characteristics (models) are created using the low and high end of their body voltage ranges, respectively. The body voltage of the device (FET) is initialized to the low end of range at time zero, and then a transient, two dimensional sweep of gate and drain voltages is performed. Drain currents are measured in this two dimensional region and are used to create a piecewise, linear IV model of device. The process is repeated for the highest body voltage. This process differs significantly from prior art bulk device characterization techniques, which did not have to initialize body voltage or perform a transient analysis. The body voltage is modulating during the switching event due to the gate-to-body and diffusion-to-body coupling; and thus only a transient analysis can properly model these coupling effects.
摘要:
An improved clocking circuit is provided for generating a half swing clock. Previous circuit operations required an additional supply voltage rail (Vdd/2), but the preferred embodiment exploits charge sharing to generate a half swing clock with less power and without the additional supply voltage rail. To drive clock nodes to Vdd/2, a shunt transistor is opened, and the fully charged clock node shares its charge with the fully discharged clock node. When capacitances have been properly matched, both nodes will settle at Vdd/2.
摘要:
A clock generation circuit globally distributes a half-frequency clock and doubles the clock frequency locally in a local clock block circuit. The circuit contains several subcircuits which detect the global clock edges (transitions), double the clock frequency and generate two shaped local clocks. A rising edge detection circuit generates a pulse in response to a rising edge of the global clock. A falling edge detection circuit generates a pulse in response to a falling edge of the global clock. A master clock SR (set/reset) latch is reset in response to either pulse and a slave clock SR latch is set in response to either pulse. A delay circuit generates a delayed signal in response to the setting of the master clock SR latch. This delayed signal sets the master clock SR latch and resets the slave clock SR latch. The master clock latch output is repowered to drive the master latches and the slave clock latch output is repowered to drive the slave latches.
摘要:
An improved pull-down latch circuit is provided for better latch performance. Previous pull-down latch circuit performance is compromised during pull-up operation since weak PFETs are employed to pull up latch nodes. A pull up assist circuit is incorporated to assist weak PFET when latch node is being pulled up. The assist circuit is isolated from latch circuit when latch node is being pull down to guarantee that pull down circuit can overcome pull-up circuit (for correct latch operation).
摘要:
A voltage-compensated, constant delay clock buffer for a chip clock distribution circuit employs a variable gain circuit to dynamically control the delay through the first inverter stage. In the presence of no voltage rail collapse the first stage gain is set high which results in a nominal delay through the first stage. As voltage rail collapse occurs local to the clock buffer circuit, the gain on the first stage is reduced to yield a smaller than nominal delay through the first stage in such a way as to compensate for the increased delay in the subsequent stage or stages. The control circuit is responsive to a first voltage rail and a second voltage rail to provide dynamic control of the delay through the first inverter stage. The circuit can compensate the circuit to handle a plurality of second inverter stages with the control circuit adjusting the delay of said first inverter stage, and with the control circuit remaining responsive to a first voltage rail and a second voltage rail.
摘要:
A programmable memory controller generates address and control signal information associated with a word of data which is desired to be transferred first from dynamic random access memory (DRAM) modules. The generated information specifically assists memory support circuitry interfacing with page mode DRAMs. This information is normally provided to the memory support circuitry just before selection of the staring word from a fetch line data buffer. Memory latency, gaps in data transfer, can be reduced when this information is available to the support circuitry as it drives column address and/or column address strobe (CAS) signals to the DRAMs.