Flash memory - NAND and NOR - play an integral role in all types of embedded systems. However, to properly implement such an element in the system, it is good to know how it works and what its characteristics are, especially when selecting memory and choosing between NAND and NOR Flash memory.
Embedded system designers must consider many factors when choosing Flash memory: what type of Flash architecture to use, whether to choose a serial or parallel interface, whether it needs error correction code (ECC), and so on. If the processor or controller only supports one type of interface, this limits your options so you can easily choose memory. However, this is often not the case. For example, some FPGAs support serial NOR Flash, parallel NOR Flash, and NAND Flash for storing configuration data. The same memory can be used to store user data, making it difficult to choose the right memory.
To help designers of embedded systems and other devices that use Flash memory, in the following article we will take a closer look at the differences between NAND and NOR Flash memory.
Fig.1. NOR Flash (left) has an architecture
resembling a NOR gate, while NAND
Flash (right) resembles a NAND gate. Memory architecture
Flash memories store information in memory cells made of floating transistors. The names of the technologies explain how memory cells are organized. In NOR Flash, one end of each memory cell is connected to the source line and the other end directly to the bit line, creating a structure similar to a NOR gate. In NAND Flash, several memory cells (typically eight cells) are connected in series, similar to what is done in a NAND gate structure (see figure 1 ).
The NOR Flash memory architecture provides enough address lines to map the entire memory range. This has the advantage of fast random access and short read times, making it ideal for saving executing code for layouts. Another advantage is 100% known good bits throughout the life of the part - this is important in systems where reliability is a key aspect of the device. Disadvantages of NOR Flash include larger memory cell size, which results in higher cost per bit and lower memory write and erase speeds.
In contrast, NAND Flash has a much smaller cell size and much higher write and erase speeds compared to NOR Flash. Disadvantages include lower read speed and the need to implement an intermediary interface due to the need for I/O mapping. This interface does not allow random memory access, which makes memory handling slightly more difficult. It should be noted that executing code from NAND Flash is accomplished by copying the contents of the memory to RAM, which is different from executing code directly from NOR Flash. Another serious disadvantage is the presence of damaged blocks. NAND Flash typically has 98% good bits when shipped out of the box. Additionally, over the entire life of the element, additional damage occurs to memory blocks. Therefore, it is necessary to use the error correction function (ECC) in the device, as well as a dedicated controller that eliminates damaged blocks from further operation.
Memory capacity
NAND Flash memories are available in much higher storage densities than NOR Flash memories, mainly due to their lower cost per bit and smaller cell dimensions. NAND flash memories typically range from 1 Gb to 16 Gb. NOR Flash memories, in turn, range in size from 64 Mb to 2 Gb. Due to its higher density, NAND Flash memory is mainly used for data storage.
Erase, read and write
In NOR and NAND Flash chips, memory is organized into erase blocks. This architecture helps keep costs lower while maintaining high performance. Smaller block size allows for faster erase cycles. However, the disadvantage of using smaller blocks is the increase in the area and costs of the system. Due to its lower cost per bit, NAND Flash can more cost-effectively support smaller erase blocks compared to NOR Flash. The typical erase block size available today is 8KB to 32KB for NAND Flash and 64KB to 256KB for NOR Flash.
Data deletion operations in NAND Flash are simple, while in NOR Flash each byte must be written with the value "0" before it can be deleted. This causes the delete operation for NOR Flash to be much slower than for NAND Flash. For example, Cypress S34ML04G2 NAND Flash requires 3.5 ms to erase a 128 KB block, while Cypress S70GL02GT NOR Flash takes approximately 520 ms (also for a 128 KB sector).
NOR Flash memory has enough addresses and data lines to map an entire memory region, similar to how SRAM memory works. For example, a 2-Gbit (256 MB) NOR flash memory with a 16-bit data bus will have 27 address lines, allowing read access anywhere in the memory. In NAND Flash memory, memory access is performed using a multiplexed address and data bus. Typical NAND Flash memories use an 8-bit or 16-bit multiplexed address and data bus with additional signals such as Chip Enable, Write Enable, Read Enable, Address Latch Enable, Command Latch Enable, and Ready/Busy. The NAND flash must first receive a command (read, write, or erase), then an address and data. The need to perform these additional operations makes reading the random bit for NAND Flash memory much slower. For example, S34ML04G2 NAND Flash requires 30 µS compared to 120ns for S70GL02GT NOR Flash. So NAND is as much as 250 times slower in this case.
To overcome or reduce the limitations of slower lower reads, NAND Flash is often read as pages, with each page being a smaller division of erase blocks. The contents of one page are read sequentially with the address and command cycle only at the beginning of each read cycle. The sequential access time for NAND Flash is usually shorter than the random access time in NOR Flash devices due to the fact that in NAND Flash memory read sequentially, between individual pages there is no need to provide a command or address in the memory. As the read data block size increases, the cumulative latency in NOR Flash becomes larger than that in NAND Flash. Therefore, NAND Flash can be faster for sequential reads. However, due to the much longer initial read access time for NAND Flash, the performance difference is only visible when transferring large blocks of data, often for sizes above 1 KB.
In both Flash technologies, data can only be written to a block if the block is empty. The slow erase operation of NOR Flash makes the write operation even slower. In NAND Flash, as with reading, data is often written or programmed into pages (usually 2 KB each). For example, writing one page on the S34ML04G2 NAND Flash chip typically takes about 300 µs.
To speed up write operations, modern NOR Flash memories use a programmable buffer, similar to the write page. For example, the S70GL02GT memory supports buffer programming, which enables multi-byte programming with similar write times for a single word. For example, programming a buffer for 512 bytes of data can provide a throughput of 1.14 MB/s.
Energy consumption
NOR Flash memories typically require more current when first powered up than NAND Flash. However, the standby current for NOR Flash is much lower than NAND Flash. The instantaneous active power is comparable for both Flash memories. The power consumed by the system depends on the time during which the memory is active. NOR Flash has the advantage here when it comes to random reads, while NAND Flash consumes relatively much less power for erase, write and sequential read operations.
Reliability
Data recording reliability is an important aspect of any memory device. Flash memory has a phenomenon called bit flipping, whereby some bits can be flipped spontaneously. This phenomenon occurs more often in NAND Flash than in NOR Flash. NAND memories are sold with damaged blocks scattered randomly throughout the chip, mainly due to the efficiency of the production process. More memory cells deteriorate as they are erased, and erase cycles continue throughout the life of the NAND Flash, each time the cells are reprogrammed. Bad block handling is therefore a mandatory feature of NAND Flash. On the other hand, NOR Flash memory is delivered without any bad blocks, and the number of memory blocks that get corrupted over the lifetime of the memory is small. So when it comes to the reliability of stored data, NOR Flash is much better than NAND Flash.
Another aspect of reliability is data retention, where NOR Flash memory again has the advantage. S70GL02GT NOR Flash offers 20 years of data retention for up to 1,000 program/erase cycles. In turn, for example, S34ML04G2 (NAND Flash) memory offers a typical data retention time of 10 years.
An equally important feature to consider is the maximum number of programming and erasing cycles for a given system. Initially, NAND Flash offered 10 times the number of program and erase cycles compared to NOR Flash. With current technological advances, this is no longer true as both memories are now comparable in this respect. For example, both the S70GL02GT (NOR) and S34ML04G2 (NAND) offer up to 100,000 program erase cycles. However, due to the smaller block size used in NAND Flash, less area is erased for each operation. This means that NAND Flash still has a longer overall lifespan.
Table 1 , below, summarizes the main aspects discussed in this article.
Essentially, NOR Flash memory is an excellent choice for applications requiring lower capacity but faster random read access and higher data reliability. Therefore, NOR Flash chips are ideal for storing CPU code, etc. NAND Flash memories, on the other hand, are ideal for applications such as data storage where larger memory capacity and faster write and erase operations are required.
Source: https://www.embedded.com/flash-101-nand-flash-vs-nor-flash/" target="_blank" rel="nofollow noopener ugc" class="postlink inline" title="" > https://www.embedded.com/flash-101-nand-flash-vs-nor-flash/
Embedded system designers must consider many factors when choosing Flash memory: what type of Flash architecture to use, whether to choose a serial or parallel interface, whether it needs error correction code (ECC), and so on. If the processor or controller only supports one type of interface, this limits your options so you can easily choose memory. However, this is often not the case. For example, some FPGAs support serial NOR Flash, parallel NOR Flash, and NAND Flash for storing configuration data. The same memory can be used to store user data, making it difficult to choose the right memory.
To help designers of embedded systems and other devices that use Flash memory, in the following article we will take a closer look at the differences between NAND and NOR Flash memory.
Fig.1. NOR Flash (left) has an architecture
resembling a NOR gate, while NAND
Flash (right) resembles a NAND gate.
Flash memories store information in memory cells made of floating transistors. The names of the technologies explain how memory cells are organized. In NOR Flash, one end of each memory cell is connected to the source line and the other end directly to the bit line, creating a structure similar to a NOR gate. In NAND Flash, several memory cells (typically eight cells) are connected in series, similar to what is done in a NAND gate structure (see figure 1 ).
The NOR Flash memory architecture provides enough address lines to map the entire memory range. This has the advantage of fast random access and short read times, making it ideal for saving executing code for layouts. Another advantage is 100% known good bits throughout the life of the part - this is important in systems where reliability is a key aspect of the device. Disadvantages of NOR Flash include larger memory cell size, which results in higher cost per bit and lower memory write and erase speeds.
In contrast, NAND Flash has a much smaller cell size and much higher write and erase speeds compared to NOR Flash. Disadvantages include lower read speed and the need to implement an intermediary interface due to the need for I/O mapping. This interface does not allow random memory access, which makes memory handling slightly more difficult. It should be noted that executing code from NAND Flash is accomplished by copying the contents of the memory to RAM, which is different from executing code directly from NOR Flash. Another serious disadvantage is the presence of damaged blocks. NAND Flash typically has 98% good bits when shipped out of the box. Additionally, over the entire life of the element, additional damage occurs to memory blocks. Therefore, it is necessary to use the error correction function (ECC) in the device, as well as a dedicated controller that eliminates damaged blocks from further operation.
Memory capacity
NAND Flash memories are available in much higher storage densities than NOR Flash memories, mainly due to their lower cost per bit and smaller cell dimensions. NAND flash memories typically range from 1 Gb to 16 Gb. NOR Flash memories, in turn, range in size from 64 Mb to 2 Gb. Due to its higher density, NAND Flash memory is mainly used for data storage.
Erase, read and write
In NOR and NAND Flash chips, memory is organized into erase blocks. This architecture helps keep costs lower while maintaining high performance. Smaller block size allows for faster erase cycles. However, the disadvantage of using smaller blocks is the increase in the area and costs of the system. Due to its lower cost per bit, NAND Flash can more cost-effectively support smaller erase blocks compared to NOR Flash. The typical erase block size available today is 8KB to 32KB for NAND Flash and 64KB to 256KB for NOR Flash.
Data deletion operations in NAND Flash are simple, while in NOR Flash each byte must be written with the value "0" before it can be deleted. This causes the delete operation for NOR Flash to be much slower than for NAND Flash. For example, Cypress S34ML04G2 NAND Flash requires 3.5 ms to erase a 128 KB block, while Cypress S70GL02GT NOR Flash takes approximately 520 ms (also for a 128 KB sector).
NOR Flash memory has enough addresses and data lines to map an entire memory region, similar to how SRAM memory works. For example, a 2-Gbit (256 MB) NOR flash memory with a 16-bit data bus will have 27 address lines, allowing read access anywhere in the memory. In NAND Flash memory, memory access is performed using a multiplexed address and data bus. Typical NAND Flash memories use an 8-bit or 16-bit multiplexed address and data bus with additional signals such as Chip Enable, Write Enable, Read Enable, Address Latch Enable, Command Latch Enable, and Ready/Busy. The NAND flash must first receive a command (read, write, or erase), then an address and data. The need to perform these additional operations makes reading the random bit for NAND Flash memory much slower. For example, S34ML04G2 NAND Flash requires 30 µS compared to 120ns for S70GL02GT NOR Flash. So NAND is as much as 250 times slower in this case.
To overcome or reduce the limitations of slower lower reads, NAND Flash is often read as pages, with each page being a smaller division of erase blocks. The contents of one page are read sequentially with the address and command cycle only at the beginning of each read cycle. The sequential access time for NAND Flash is usually shorter than the random access time in NOR Flash devices due to the fact that in NAND Flash memory read sequentially, between individual pages there is no need to provide a command or address in the memory. As the read data block size increases, the cumulative latency in NOR Flash becomes larger than that in NAND Flash. Therefore, NAND Flash can be faster for sequential reads. However, due to the much longer initial read access time for NAND Flash, the performance difference is only visible when transferring large blocks of data, often for sizes above 1 KB.
In both Flash technologies, data can only be written to a block if the block is empty. The slow erase operation of NOR Flash makes the write operation even slower. In NAND Flash, as with reading, data is often written or programmed into pages (usually 2 KB each). For example, writing one page on the S34ML04G2 NAND Flash chip typically takes about 300 µs.
To speed up write operations, modern NOR Flash memories use a programmable buffer, similar to the write page. For example, the S70GL02GT memory supports buffer programming, which enables multi-byte programming with similar write times for a single word. For example, programming a buffer for 512 bytes of data can provide a throughput of 1.14 MB/s.
Energy consumption
NOR Flash memories typically require more current when first powered up than NAND Flash. However, the standby current for NOR Flash is much lower than NAND Flash. The instantaneous active power is comparable for both Flash memories. The power consumed by the system depends on the time during which the memory is active. NOR Flash has the advantage here when it comes to random reads, while NAND Flash consumes relatively much less power for erase, write and sequential read operations.
Reliability
Data recording reliability is an important aspect of any memory device. Flash memory has a phenomenon called bit flipping, whereby some bits can be flipped spontaneously. This phenomenon occurs more often in NAND Flash than in NOR Flash. NAND memories are sold with damaged blocks scattered randomly throughout the chip, mainly due to the efficiency of the production process. More memory cells deteriorate as they are erased, and erase cycles continue throughout the life of the NAND Flash, each time the cells are reprogrammed. Bad block handling is therefore a mandatory feature of NAND Flash. On the other hand, NOR Flash memory is delivered without any bad blocks, and the number of memory blocks that get corrupted over the lifetime of the memory is small. So when it comes to the reliability of stored data, NOR Flash is much better than NAND Flash.
Another aspect of reliability is data retention, where NOR Flash memory again has the advantage. S70GL02GT NOR Flash offers 20 years of data retention for up to 1,000 program/erase cycles. In turn, for example, S34ML04G2 (NAND Flash) memory offers a typical data retention time of 10 years.
An equally important feature to consider is the maximum number of programming and erasing cycles for a given system. Initially, NAND Flash offered 10 times the number of program and erase cycles compared to NOR Flash. With current technological advances, this is no longer true as both memories are now comparable in this respect. For example, both the S70GL02GT (NOR) and S34ML04G2 (NAND) offer up to 100,000 program erase cycles. However, due to the smaller block size used in NAND Flash, less area is erased for each operation. This means that NAND Flash still has a longer overall lifespan.
Table 1 , below, summarizes the main aspects discussed in this article.
Essentially, NOR Flash memory is an excellent choice for applications requiring lower capacity but faster random read access and higher data reliability. Therefore, NOR Flash chips are ideal for storing CPU code, etc. NAND Flash memories, on the other hand, are ideal for applications such as data storage where larger memory capacity and faster write and erase operations are required.
Source: https://www.embedded.com/flash-101-nand-flash-vs-nor-flash/" target="_blank" rel="nofollow noopener ugc" class="postlink inline" title="" > https://www.embedded.com/flash-101-nand-flash-vs-nor-flash/
