We all are familiar with RAM. But have you heard about Intelligent RAM? The goal of Intelligent RAM (IRAM) is to design an economical computer by designing a processor during a memory fabrication process; instead of in a typical logic fabrication method, and include memory on-chip. To design a processor in a DRAM process one should study the business and culture of the DRAMs that is quite different from microprocessors.
IRAM is attractive nowadays for many reasons. First, the gap between the performance of processors and DRAMs has been widening at 500th per year for ten years. So that despite heroic efforts by applications developers, compiler writers, and architects, more applications are restricted by memory speed nowadays than in the past.
Second, since the actual processor occupies only about one third of the die, the approaching gigabit DRAM has enough capacity that whole programs and information sets will match on one chip. Within the past, so very little memory may fit onchip with the CPU that IRAMs were mainly considered as building blocks for multiprocessors.
Third, DRAM dies have grown about 500th each generation; DRAMs are being created with more metal layers to accelerate the longer lines of those larger chips. And also, the high speed interface of synchronous DRAM would require quick transistors on the DRAM chip. These 2 DRAM trends should build logic on DRAM closer to the speed of logic on logic fabs than in the past.
Now let us see some of the advantages of Intelligent RAM or IRAM:
RAM and Processor on single chip therefore less BUS length and because of that higher data transfer.
To reduce latency, the wire length should be kept as short as possible. This implies the fewer bits per block the better. A lot of lower latency may be obtained by utilizing quicker circuit topologies. Also intelligent floor designing, and redesigning the address/data bussing schemes. Applications with unpredictable memory accesses and very massive memory “footprints”. Like information bases, could benefit of the potential 5X to 10X decrease in IRAM latency.
Memory size and width:
Another advantage of IRAM over typical designs is that the ability to regulate both the size. And width of the on-chip DRAM. Instead of being restricted by powers of two in length or width, as is typical DRAM, IRAM designers will specify exactly the number of words and their width. This flexibility will improve the cost of IRAM solutions versus memories made of typical DRAMs.
Efficiency in energy:
Integrating a microprocessor and DRAM memory on the same die offers the potential for improving energy consumption of the memory system. An IRAM has the potential for higher performance than a traditional approach. Since higher performance for a few fixed energy consumption are often translated into equal performance at a lower amount of energy. The performance benefits of IRAM are often translated into lower energy consumption.
Finally, IRAM is also attractive in applications where board space is precious. Such as cellular phones or portable computers; since it integrates many chips into one.