Arithmetic processor

Figure 1. Block diagram of the functional units of the Am9511A arithmetic processor integrated...

Figure 3. Schematic logic diagram of a hardware interface between a Zilog Z80A microprocessor and the Am9511A arithmetic processor unit.

Either of the remaining two options to be discussed would provide at least an order of magnitude increase in speed on the Am9511A and are much coveted by the author, but unfortunately they would require financial resources on a very large scale for development into useable floating point arithmetic processors. 

Figure 4. Logic diagram of an Am9511A arithmetic processor unit to S-100 bus hardware interface.

A natural progression from the use of a single Am9511A arithmetic processor chip, which can enhance the performance of F80 functions/subroutines by a factor of up to 25x, is to use a number of such chips in a VP or AP architecture. 

Well, the MVP-9500 would need to be unloaded after the VSUBI and reloaded with the same numbers before the VMUL operation, which is obviously a waste of time. A general rule when programming the MVP-9500, and most other arithmetic processors, is to minimize the transfer of data to and from the arithmetic elements. This is the reason for routines like VMMA which performs the operations E = (A B) + (C D) where A to E are n element vectors the intermediate results A B and C D can be stored on the Am9511A stack and added together when the products are complete. 

The differential equations were solved by numerical integration on a personal computer OLIVETTI M24 with arithmetic processor support. 

Although many relatively complex logic systems are now available, we wiU only consider one arithmetic processor—the half-adder—in this review. Many interesting molecular-scale arithmetic systems have come to light in recent years. Since nnknown molecular processes in the brain have allowed hnmans to be numerate, it was important to demonstrate nnmeracy with synthetic molecules. The simplest semiconductor-based arithmetic device is the half-adder, which has two inputs and two ouqrut channels. 

Brock, B.C., Hunt, W.A. Formally Specifying and Mechanically Verifying Programs for the Motorola Complex Arithmetic Processor DSP. In Proceedings of the 1997 IEEE International Conference on Computer Design VLSI in Computers and Processors, ICCD 1997, October 12-15, pp. 31-36 (1997)... 

Arc-length continuation, steady states of a model premixed laminar flame, 410 Architecture, between parallel machines, 348 Arithmetic control processor, ST-100, 125 Arithmetic floating point operations,... 

Lee [Lee, 1988][Lee, 1989] surveyed processor architecture circa 1988. He pointed out that principal differences between the arithmetic sections in integer microprocessors and DSPs are ... 

The memory organization of DSPs are also different from ordinary processors because (1) Memory is typical static RAM and virtual memory support is totally absent (2) Several machines separate data and instruction streams (Harvard Architecture) (at the cost of extra pins). Additionally, modular arithmetic address modes have been added to most processors. This mode finds particular utility in filter coefficient pointers, ring buffer pointers and, with bit reversed addressing, FFTs. One further difference is the use of loop buffers for filtering. Although often called instruction caches by the chip manufacturers, they are typically very small (for example, the AT T DSP-16 has 16 instructions) and furthermore, the buffer is not directly interposed between memory and the processor. 

Floating Point. Integrated floating point units first arrived as separate coprocessors under the direct control of the microprocessor. However, these processors performed arithmetic with numerous sequential operations, resulting in performance too slow for real-time signal processing. 

Other manufacturers have introduced 16 bit fixed point processors. IBM s Mwave is supposed to be for Multimedia operations but can only address 32K (15 bits) of data memory. The product register is also only 32 bits (data is assumed to be fractional form of only 15 bits) so constant rescaling is necessary. Perhaps its most noteworthy addition is the wide use of saturation arithmetic in the instruction set and the large number of DMA channels. 

Analog Devices also has a 16 bit DSP (the ADSP-2100 series [Roesgen, 1986]) that has found some limited use in audio applications. The 2100 series has limited on chip memory and a limited number of pins (14) for the external memory. Use of a common bus for arithmetic results limits the amount of processor parallelism. However, unlike... 

Sohie and Kloker, 1988] Sohie, G. R. and Kloker, K. L. (1988). A Digital Signal Processor with IEEE Floating-Point Arithmetic. IEEE Micro, 8(6) 49-67. 

Peripheral processors which are capable of performing floating point arithmetic operations at high speed are used to enhance the poor performance of popular general purpose minicomputers in this area. These devices are described in various ways but the following nomenclature will be used in this paper. 

The Intel 8086 or 8088 microprocessors could be used in conjunction with the Intel 8087 floating point processor chip (4) which is probably twice as fast as the Am9511A for on-chip operations and includes extended precision arithmetic in its instruction set. Unfortunately the 8087 was only laid down on paper, not silicon, when this work started. The 8087 is now (January 1981) available in sample quantities at a price far in excess of the Am9511A. In addition to the price and availability problem the instruction set of the 8087 is less suited to chemical computations than the Am9511A in that many transcen-... 

In the ring system, the bandwidth between adjacent processors is fixed howevet, utilizing the special characteristics of a ring connected architecture provides system intercommunication bandwidths which tend towards the arithmetic product of this fixed interprocedure bandwidth and the number of processors in the system. Thus, a proportionate increase in supporting intercommunications bandwidth is available as processors are added to the system Expansion within a ring connected system is, of course, virtually unlimited and has a very low cost impact. 

---