Logical Memory Organization, Types of RAM, and
Cache Memory
1. Logical Memory Organization
Logical memory organization refers to the way a computer system divides and manages
main memory (RAM) for different purposes. In early computer systems, especially DOS-
based systems, memory was divided into fixed regions such as
Conventional Memory,
Upper Memory,
High Memory Area,
Expanded Memory,
Extended Memory.
1.1 Conventional Memory (Base RAM)
Conventional memory (also called Base RAM) is the first part of system memory. The first
640 KB of system memory is known as conventional memory.
Size: 640 KB (0 KB to 640 KB).
Used by standard DOS programs, device drivers, and memory-resident programs.
This memory was the most important part for old DOS-based computers.
In the DOS era, most applications could only use conventional memory. This created the
famous limitation called the '640 KB barrier'.
1.2 Upper Memory Area (UMA)
Upper Memory Area (UMA) refers to the memory region above conventional memory. It is
384 KB located above the 640 KB region.
Size: 384 KB (640 KB to 1024 KB).
Reserved for system use such as ROM shadowing, BIOS ROM, and device drivers.
Used for system devices and special use.
1.3 High Memory Area (HMA)
High Memory Area is the first 64 KB of the second megabyte of memory., it is the first 64 KB
of the 2nd MB.
Size: 64 KB.
Used in real mode.
Used by DOS to load some system components above 1 MB to free conventional
memory.
1.4 Expanded Memory (EMS)
Expanded Memory is a memory management technique developed to overcome the 1 MB
addressing limitation of the 8088 and 8086 processors. EMS was introduced around 1984.
Introduced to expand conventional memory beyond 1 MB in IBM XT compatible
computers.
Technique used: Bank switching (switching between memory banks).
Designed mainly for DOS software programs requiring additional memory.
Expanded memory cannot be accessed directly by processor.
It is accessed through a 64 KB segment in High Memory Area using mapping hardware.
In bank switching, only a portion of expanded memory is mapped into the CPU address
space at one time. The system switches between different banks when required.
1.5 Extended Memory (XMS)
Extended memory refers to all memory above the first 1 MB. All memory above 1 MB is
called extended memory.
Memory above 1 MB is called extended memory.
This memory extends from the end of HMA to the end of installed RAM.
Supported by processors that allow protected mode access.
Best suited for Windows and multitasking operating systems.
1.6 Comparison: Conventional vs Expanded vs Extended Memory
Parameter Conventional Expanded (EMS) Extended (XMS)
Memory Range 0 KB to 640 KB Above 1 MB Above 1 MB directly
(mapped)
Used By DOS DOS applications Windows &
programs/drivers needing extra multitasking OS
memory
Access Method Direct Bank switching via Direct in protected
64 KB window mode
Main Limitation Small size Complex mapping Requires advanced
CPU/OS
1.7 Real Life Interpretation of Logical Memory Organization
Logical memory can be compared to a building:
• Conventional memory is the main working floor.
• Upper memory is reserved for management/system offices.
• Expanded memory is like an external storage room accessed through a small gate.
• Extended memory is like extra floors which modern users can access directly.
2. Types of RAM
RAM (Random Access Memory) is the primary working memory of a computer. It stores
data and instructions temporarily which are needed by the CPU during processing. RAM is
volatile, meaning its contents are lost when power is turned off.
2.1 Main Classification of RAM
Static RAM (SRAM)
Dynamic RAM (DRAM)
2.2 Static RAM (SRAM)
Static RAM is a type of RAM that holds data in a static form as long as the memory has
power. According to the PDF, SRAM does not need refreshing and stores one bit of data
using four transistors.
Does not need refreshing.
Faster than DRAM.
More expensive and consumes more power.
Used in cache memory (L1, L2, L3).
2.3 Dynamic RAM (DRAM)
Dynamic RAM stores each bit of data as an electrical charge in a capacitor. The capacitor
charge leaks, so DRAM requires refreshing repeatedly.
Needs refresh cycles.
Slower than SRAM but cheaper.
Used as main memory in computers.
2.4 SDRAM (Synchronous DRAM)
SDRAM is a type of DRAM that works in synchronization with the system clock. This
improves speed compared to older asynchronous DRAM.
Synchronized with system clock.
Faster than standard DRAM.
Used in older systems.
2.5 DDR (Double Data Rate SDRAM)
DDR RAM is a form of SDRAM that transfers data twice per clock cycle (both rising and
falling edges). This doubles the data rate without increasing the clock frequency.
Transfers data twice per clock cycle.
Higher performance than SDRAM.
2.6 DDR2 RAM
DDR2 is an improved version of DDR RAM. It provides higher speed and lower power
consumption.
Higher bandwidth than DDR.
Lower voltage and better efficiency.
Improved internal architecture.
2.7 DDR3 RAM
DDR3 is a faster and more power-efficient version of DDR2. It is widely used in many
systems.
Higher speed and bandwidth than DDR2.
Lower power consumption.
Better performance for multitasking applications.
2.8 Modern RAM: DDR4 and DDR5 (Extra Information)
Modern computers now use DDR4 and DDR5 memory. These provide extremely high speed
and efficiency required for gaming, AI, and high-performance computing.
DDR4: higher bandwidth and efficiency than DDR3.
DDR5: very high bandwidth, used in advanced systems.
2.9 Real Life Interpretation of RAM Types
RAM can be compared to a student's working table:
• DRAM is like a large table where many books can be kept (cheap and large).
• SRAM is like a small side desk where the most important books are kept (fast but
expensive).
3. Cache Memory
Cache memory is a small and very high-speed memory located near the CPU. It stores
frequently used data and instructions so that the CPU does not need to access slower RAM
repeatedly.
3.1 Need for Cache Memory
CPU speed is much faster than RAM. If CPU directly depends on RAM, it wastes time waiting
for data. Cache memory reduces this waiting time and improves system performance.
Reduces memory access time.
Improves processing speed.
Bridges the speed gap between CPU and RAM.
3.2 Working of Cache Memory
Whenever CPU needs data, it first searches the cache. If data is found, it is called a Cache Hit.
If data is not found, it is called a Cache Miss, and CPU fetches the data from RAM and stores
it in cache.
3.3 Cache Hit and Cache Miss
Cache Hit: Data found in cache → CPU gets data quickly.
Cache Miss: Data not found → CPU accesses RAM (slower).
3.4 Levels of Cache Memory
L1 Cache
L1 cache is the fastest and smallest cache located inside each CPU core.
Very fast.
Very small size.
Stores most frequently used data.
L2 Cache
L2 cache is larger than L1 but slower. It acts as a backup cache for L1.
Larger than L1.
Moderate speed.
May be inside CPU or near CPU.
L3 Cache
L3 cache is larger than L1 and L2 and is often shared among multiple CPU cores.
Largest cache.
Shared among CPU cores.
Slower than L1/L2 but faster than RAM.
3.5 Cache Memory Technology
Cache memory is usually built using SRAM technology.
SRAM-based, hence very fast.
Costly compared to DRAM.
Limited in size.
3.6 Advantages of Cache Memory
Increases overall CPU performance.
Reduces execution time of programs.
Improves multitasking and gaming performance.
3.7 Disadvantages of Cache Memory
Very expensive compared to RAM.
Limited storage size.
Increases processor cost.
3.8 Real Life Interpretation of Cache Memory
Cache memory is like keeping frequently used books on your study table. Instead of going to
the library (RAM) again and again, you keep important books nearby for quick access.
4. Conclusion
Logical memory organization was essential in early computer systems to manage limited
memory efficiently. RAM provides temporary working space for applications, while cache
memory improves CPU speed by storing frequently used instructions and data closer to the
processor.