quiz image

Flash Memories: NOR and NAND

ManageableSatire avatar
ManageableSatire
·
·
Download

Start Quiz

Study Flashcards

48 Questions

What is the main advantage of NAND flash memories over NOR flash memories?

Higher storage density

What is the primary limitation of flash memories?

Wear-out due to limited number of writings

What is the purpose of the Internet Archive project?

To make the historical record of the Internet over time

What is the typical cost of flash memories in 2016?

$0.3/GiB

What is the primary component of a typical cluster building block?

Servers

What is the name of the project that uses the VME T-80 rack?

Internet Archive

What is the limitation of DRAM in terms of writing speed?

10 to 100 times faster

What is the primary advantage of solid-state disks (SSDs)?

Faster reading and writing speed

What is the total raw capacity of 40 PetaBox GB2000 storage nodes in a standard VME rack?

80 TB

What is the average seek time of the 7,200 RPM PATA drives?

8.5 ms

What is the PATA link speed?

133 MiB/sec

How much overhead does the ATA controller add to perform a disk I/O?

0.1 ms

How many CPU instructions are used by the operating system for a disk I/O?

50,000

What is the average I/O size when collecting a new snapshot?

50 KiB

What is the total power dissipation of the 40 PetaBox GB2000 storage nodes in a standard VME rack?

3 kW

What is the cost of the 48-port 10/100/1000 Ethernet switch and all cables for a rack?

$3,000

What is the primary purpose of magnetic disks in a computer system?

To serve as a non-volatile storage

What is the typical range of sectors per track in a magnetic disk?

100 to 500

What is the relation between the read-write head and the platter in a magnetic disk?

The read-write head is used to access information on the platter

Why are magnetic disks usually considered slow compared to other storage devices?

Because they are based on a rotating disk

What is the typical range of tracks per disk surface in a magnetic disk?

5k to 30k

What is the advantage of using magnetic disks in the lowest level of the memory hierarchy system?

They are relatively cheap

How does the arm assembly move in a magnetic disk to access a given cylinder?

It moves to select a given cylinder

What is the main difference between magnetic disks and flash memory in terms of storage?

Magnetic disks are slower, while flash memory is faster

What is the primary advantage of RAID 6 over RAID 4?

It provides an additional level of redundancy

How does RAID 6 recover data when two disks fail?

Using the row parity and then the diagonal parity sequentially

What is the purpose of the diagonal parity disk in RAID 6?

To store the parity of the blocks in the same diagonal

How many disks are required to implement RAID 6 with p data disks?

p + 1

What is the primary advantage of flash memory over traditional EEPROM?

All of the above

How long does a 256-Byte transfer of flash memory take?

6.5 µs

Why is flash memory faster than traditional disks?

It has faster read access times

What is the primary limitation of flash memory?

It has limited write cycles

What is the primary difference between RAID 5 and RAID 6?

RAID 5 is able to recover from one failure, while RAID 6 is able to recover from double failures.

What is the purpose of the parity information in RAID 5?

To enable recovery from a single failure.

How is the parity calculated in RAID 5?

Using the XOR function.

What is the advantage of using RAID 5?

It enables recovery from a single failure.

What is the purpose of RAID 6?

To enable recovery from double failures.

How does RAID 6 differ from RAID 5?

RAID 6 considers another parity block, while RAID 5 does not.

What is the purpose of the row-diagonal parity in RAID 6?

To enable recovery from double failures.

What is the difference between the parity calculation in RAID 5 and the row-diagonal parity in RAID 6?

The parity calculation in RAID 5 is used for single failures, while the row-diagonal parity in RAID 6 is used for double failures.

What is the primary advantage of RAID 5 over other RAID levels?

It offers better performance through stripping

What is the main difference between RAID 3 and RAID 4?

RAID 3 uses bit-interleaved parity while RAID 4 uses block-interleaved parity

What is the purpose of parity calculation in RAID systems?

To allow for single disk failure recovery

What is the main disadvantage of using RAID 0?

It does not offer redundancy

What is the primary advantage of mirroring in RAID systems?

It increases information availability

What is the main difference between RAID 1 and RAID 2?

RAID 1 uses mirroring while RAID 2 uses ECC

What is the main advantage of using flash memory over magnetic disks?

It has lower latency

What is the primary purpose of redundancy in RAID systems?

To allow for single disk failure recovery

Study Notes

Flash Memory

  • NOR flash memories are a direct competitor of traditional EEPROM, randomly addressable, and typically used in BIOS.
  • NAND flash memories offer higher storage density, but can only be read in blocks, making them cheaper per gigabyte and more common than NOR flash.
  • NAND flash has a wear-out limitation of around 100,000 to 1 million recordings, but this can be expanded through uniform distribution of writes across blocks.

Solid-State Disks (SSDs)

  • SSDs have replaced floppy disks and hard drives in mobile systems.
  • They have become a popular choice due to their high I/O performance and low cost.

Clusters (I/O Servers) Evaluation

  • The evaluation is based on a VME T-80 rack, used in the Internet Archive project.
  • The storage node PetaBox GB2000 has 4× 500 GiB PATA disk drives, 512 MiB DDR266 DRAM, and a 1 GHz C3 processor.
  • The 40× GB2000s fit in a standard VME rack, giving a total of 80 TB of raw capacity.
  • The nodes are connected with a 48-port 10/100/1000 switch, dissipating around 3 kW.

Cost and Performance Evaluation

  • The cost of the processor, DRAM, ATA disk controller, and power supply is $500.
  • The cost of 7,200 RPM PATA drives is $375 × 4.
  • The cost of the 48-port 10/100/1000 Ethernet switch is $3,000.
  • The average I/O size is 16 KiB for accesses to the historical record and 50 KiB for collecting a new snapshot.
  • The performance evaluation is based on the cost per I/O per second (IOPS) of the 80 TB rack.

Storage and I/O Systems

  • Non-volatile memory is part of the memory hierarchy system or I/O system, connected to the I/O buses.
  • Main choices for storing data are magnetic disks and flash memory.
  • Magnetic disks serve as non-volatile storage, are big, cheap, and slow, and reside in the lowest level of the memory hierarchy system.
  • Disks have a rotating disk covered with a magnetic surface, using a read-write head per surface to access information.

RAID 6

  • RAID 6 is a scheme that allows for recovery of a block even if two disks fail
  • It uses p - 1 diagonals to protect p disks
  • Each block of the diagonal parity disk contains the parity of the blocks in the same diagonal
  • Recovery procedure involves using the row-diagonal approach to recover one of the four blocks within a failed disk, and then using the diagonal parity to recover another block

How RAID 6 Works

  • If data disks number 1 and 3 fail, standard RAID recovery using the first row (row parity) does not help
  • The recovery procedure starts by recovering one of the four blocks within the failed disk (disk 1), using the diagonal parity
  • Then, the diagonal 2 is used to recover a block from disk 1, since it does not involve disk 3
  • The loop continues until all blocks are finally recovered

Flash Memory

  • Flash memory technology is similar to traditional EEPROM, but with higher memory capacity per chip and low-power consumption
  • Read access time is slower than DRAM, but much faster than disks
  • A 256-Byte transfer of flash would take around 6.5 µs, and 1,000 times more on disks

RAID Levels

RAID 5

  • This level has distributed block-interleaved parity
  • It is able to recover from one failure
  • Illustrated in Fig. 9.4

RAID 6

  • This level is a RAID 5 extension, considering another parity block
  • It is able to recover from double failures
  • Illustrated in Fig. 9.5

RAID Details

  • In RAID 6, there is the row-diagonal parity
  • Each diagonal does not cover one disk, i.e., one is left out
  • Disks should not fail, but all hardware fails at some point in time

Reliability vs. Availability

  • In RAID, reliability becomes a problem considering the system will have less reliability, i.e., more disks then bring a greater fail probability
  • Conversely, availability is increased, i.e., failures do not necessarily lead to unavailability
  • Latency is not necessarily reduced, availability is enhanced through the addition of redundant disks, and lost information can be rebuilt through redundant data

Standard Levels Summary

  • Table 9.1 presents a summary about the RAID levels
  • Fig. 9.4 illustrates the RAID 3, RAID 4, and RAID 5 systems

Learn about the differences between NOR and NAND flash memories, including their history, advantages, and writing processes. Understand how they compare to traditional EEPROM and their uses in computer systems.

Make Your Own Quizzes and Flashcards

Convert your notes into interactive study material.

Get started for free

More Quizzes Like This

Solid-State Drives (SSDs) Quiz
5 questions
Solid-State Drive (SSD) Quiz
5 questions
Solid-State Drives (SSDs)
32 questions
Types of Computer Memory
12 questions

Types of Computer Memory

AffectionateTourmaline avatar
AffectionateTourmaline
Use Quizgecko on...
Browser
Browser