Sign in with
Sign up | Sign in

Level 2 And Level 3 Cache

Upgrading And Repairing PCs 21st Edition: Processor Specifications
By

Level 2 Cache

To mitigate the dramatic slowdown every time an L1 cache miss occurs, a secondary (L2) cache is employed.

Using the restaurant analogy I used to explain L1 cache in the previous section, I’ll equate the L2 cache to a cart of additional food items placed strategically in the restaurant such that the waiter can retrieve food from the cart in only 15 seconds (versus 60 seconds from the kitchen). In an actual Pentium class (Socket 7) system, the L2 cache is mounted on the motherboard, which means it runs at motherboard speed (66 MHz, or 15 ns in this example). Now, if you ask for an item the waiter did not bring in advance to your table, instead of making the long trek back to the kitchen to retrieve the food and bring it back to you 60 seconds later, he can first check the cart where he has placed additional items. If the rejust quested item is there, he will return with it in only 15 seconds. The net effect in the real system is that instead of slowing down from 233 MHz to 16 MHz waiting for the data to come from the 60 ns main memory, the system can instead retrieve the data from the 15 ns (66 MHz) L2 cache. The effect is that the system slows down from 233 MHz to 66 MHz.

All modern processors have integrated L2 cache that runs at the same speed as the processor core, which is also the same speed as the L1 cache. For the analogy to describe these newer chips, the waiter would simply place the cart right next to the table you were seated at in the restaurant. Then, if the food you desired wasn’t on the table (L1 cache miss), it would merely take a longer reach over to the adjacent L2 cache (the cart, in this analogy) rather than a 15-second walk to the cart as with the older designs.

The screenshot below illustrates the cache types and sizes in the AMD A10-5800K processor, as reported by CPU-Z.

The AMD A10-5800K processor is a quad-core processor with L1 and L2 cache.The AMD A10-5800K processor is a quad-core processor with L1 and L2 cache.

Level 3 Cache

Most late-model mid-range and high-performance processors also contain a third level of cache known as L3 cache. In the past, relatively few processors had L3 cache, but it is becoming more and more common in newer and faster multicore processors such as the Intel Core i7 and AMD Phenom II and FX processors.

Extending the restaurant analogy I used to explain L1 and L2 caches, I’ll equate L3 cache to another cart of additional food items placed in the restaurant next to the cart used to symbolize L2 cache. If the food item needed was not on the table (L1 cache miss) or on the first food cart (L2 cache miss), the waiter could then reach over to the second food cart to retrieve a necessary item.

L3 cache proves especially useful in multicore processors, where the L3 is generally shared among all the cores. Both Intel and AMD use L3 cache in most of their current processors because of the benefits to multicore designs.

Cache Information for the Intel Core i5-2500 (Sandy Bridge)Cache Information for the Intel Core i5-2500 (Sandy Bridge)

These screenshots illustrate two examples of six-core processors with L1, L2, and L3 cache from both Intel (above) and AMD (below):

Cache information for the AMDPhenom II X6 1055TCache information for the AMDPhenom II X6 1055T

Ask a Category Expert

Create a new thread in the Reviews comments forum about this subject

Example: Notebook, Android, SSD hard drive

Display all 36 comments.
This thread is closed for comments
Top Comments
  • 11 Hide
    DelightfulDucklings , October 14, 2013 10:22 PM
    Very interesting article, I quite enjoyed the part about Cache memory
Other Comments
  • 5 Hide
    xkm1948 , October 14, 2013 9:13 PM
    Really nice intro article!
  • 11 Hide
    DelightfulDucklings , October 14, 2013 10:22 PM
    Very interesting article, I quite enjoyed the part about Cache memory
  • 2 Hide
    kindiana , October 14, 2013 10:27 PM
    nice article
  • 9 Hide
    burnley14 , October 14, 2013 10:29 PM
    One of the most interesting and informative articles I've ever read on the site. Great job!
  • 2 Hide
    aredflyingbird , October 14, 2013 10:41 PM
    Agreed, excellent article.
  • 2 Hide
    palladin9479 , October 14, 2013 10:55 PM
    Really good article, actually was spot on with how caching works.
  • 3 Hide
    AndrewJacksonZA , October 15, 2013 12:22 AM
    "Forward From The Editor"
    Shouldn't that be "Foreword?"
  • 3 Hide
    iam2thecrowe , October 15, 2013 2:23 AM
    I need more cache in my kitchen.
  • 0 Hide
    LalitMotagi , October 15, 2013 2:25 AM
    Great Article.
  • -1 Hide
    Rex Romero , October 15, 2013 2:30 AM
    Andrew. It's so advanced so it's forward. lol
  • 1 Hide
    groundrat , October 15, 2013 4:04 AM
    Excellent article.
  • 0 Hide
    ojas , October 15, 2013 5:22 AM
    Quote:
    For example, if you live on a street in which the address is limited to a two-digit (base 10) number, no more than 100 distinct addresses (00–99) can exist for that street (102). Add another digit, and the number of available addresses increases to 1,000 (000–999), or 103.

    Should be 10^2, 10^3, and in the next para, 2^x.
  • 0 Hide
    ojas , October 15, 2013 6:11 AM
    Quote:
    Note: Early versions of EM64T-equipped processors from Intel lacked support for the LAHF and SAHF instructions used in the AMD64 instruction set.

    This was very interesting, considering both instructions were supported even by the humble 8086.
  • 3 Hide
    spookyman , October 15, 2013 6:20 AM
    I have one still from 18 years ago. Still one of the best tech books I own. Though mine was just starting to touch the Intel Pentium processor. It even covered IBM's PS/2 computers and technology. Its amazing how much more hardware intensive PC's were back then they are now.
  • 0 Hide
    ojas , October 15, 2013 6:50 AM
    I ended up buying the 19th edition after last year's excerpts on Tom's Hardware, the 20th wasn't available in India then.

    These sections seem more or less unchanged, except for the mention of Ivy and Vishera, and i think the CPU-z screenshots are new as well.
  • 0 Hide
    AndrewJacksonZA , October 15, 2013 6:54 AM
    Quote:
    Quote:
    Note: Early versions of EM64T-equipped processors from Intel lacked support for the LAHF and SAHF instructions used in the AMD64 instruction set.

    This was very interesting, considering both instructions were supported even by the humble 8086.
    Apparently they were missing from the early 64bit CPUs from AMD and Intel. They appeared in March 2005 for AMD CPUs and June 2005 for Intel CPUs
    https://en.wikipedia.org/wiki/X86-64#Older_implementations

    Yet at the very least the 80386 supported them:
    http://css.csail.mit.edu/6.858/2011/readings/i386/LAHF.htm

    So it appears that it was an early-64 bit CPU issue only.
  • 0 Hide
    ta152h , October 15, 2013 9:04 AM
    I could only get to page three before being thoroughly disgusted by the lack of knowledge of the writer. How can he be writing books, without actually knowing the material?

    The Prescott introduced 64-bit to the Intel world, not the Core 2. Kind of common knowledge. The Athlon XP had a 36-bit address bus? I don't remember ever seeing that.

    Then we go to the misinformation about the 8086/8088 to 386.

    In actuality, there were four modes in the 80386. Real, Virtual 86, Protected 286, and Protected 386. Yup, four. And no, Windows 3.0 was not expected to run on an 8088 or 80286, because it DID use Virtual 86, which those processors could not support. You know, the part where they let you go from one DOS task to another. That was in the hardware. And that hardware started with the 80386.

    Moreover, the 80286 did NOT have the same instruction set as the 8086. Only in real mode did it. And why do you suppose it was called real mode? Maybe because the addresses were not virtualized? The 80286, as mentioned above, did have virtual addresses in what was called the 80286 Protected Mode. It not only ran Real Mode apps much faster, but when in Protected Mode was very capable of running multitasking Operating Systems, something that could not be done well on the 8086. It also increased the memory bus to 24-bits, albeit still using 64K bit segments.

    OS/2 1.x was the best example of an OS using 286 Protected mode, although any software using "Extended Memory" was taking advantage of the greater addressing of the 286, albeit in an inelegant way.

    I stopped reading after page three, as it's just discouraging to think people are writing books without being accurate. OK, so we have the author that got it wrong, fair enough, but what about the people who are supposed to error check it. I certainly don't know everything, and I know this stuff, and it's pretty basic. No one caught this? Are you kidding me? The 286 stuff might be a bit far away, but not knowing that x86-64 first appeared in the Prescott line is really difficult to understand, and is very basic. This is made more so because of all the rumors that the processor was made to support it, but Intel was hiding it so as to not undercut the Itanium. In time, it was proven true.

    Please, don't spread misinformation. Someone will repeat this stuff, and then someone else will, and it becomes 'fact' despite being wrong. If you publish a book, make a friggin effort! I'm sure I could errors the rest of the way, but it's just too annoying for me to wade through this rubbish.

    By the way, the term CPU bus is an ambiguous one. The CPU has multiple buses, and if you used that term with me, I'd wonder which one you were referring to. Find a more accurate term, like PCI-E bus if that's what you are trying to say.
  • -2 Hide
    ezorb , October 15, 2013 9:22 AM
    I feel that this is bellow the level of this website, even below the level of Maximum PC (which has a great podcast), this is the book my grandfather would buy if he wanted to try his hand at build a PC
  • -1 Hide
    hardrock40 , October 15, 2013 10:08 AM
    Really nice article well worth the read for sure.
  • -2 Hide
    ta152h , October 15, 2013 10:42 AM
    I could only get to page three before being thoroughly disgusted by the lack of knowledge of the writer. How can he be writing books, without actually knowing the material?

    The Prescott introduced 64-bit to the Intel world, not the Core 2. Kind of common knowledge. The Athlon XP had a 36-bit address bus? I don't remember ever seeing that.

    Then we go to the misinformation about the 8086/8088 to 386.

    In actuality, there were four modes in the 80386. Real, Virtual 86, Protected 286, and Protected 386. Yup, four. And no, Windows 3.0 was not expected to run on an 8088 or 80286, because it DID use Virtual 86, which those processors could not support. You know, the part where they let you go from one DOS task to another. That was in the hardware. And that hardware started with the 80386.

    Moreover, the 80286 did NOT have the same instruction set as the 8086. Only in real mode did it. And why do you suppose it was called real mode? Maybe because the addresses were not virtualized? The 80286, as mentioned above, did have virtual addresses in what was called the 80286 Protected Mode. It not only ran Real Mode apps much faster, but when in Protected Mode was very capable of running multitasking Operating Systems, something that could not be done well on the 8086. It also increased the memory bus to 24-bits, albeit still using 64K bit segments.

    OS/2 1.x was the best example of an OS using 286 Protected mode, although any software using "Extended Memory" was taking advantage of the greater addressing of the 286, albeit in an inelegant way.

    I stopped reading after page three, as it's just discouraging to think people are writing books without being accurate. OK, so we have the author that got it wrong, fair enough, but what about the people who are supposed to error check it. I certainly don't know everything, and I know this stuff, and it's pretty basic. No one caught this? Are you kidding me? The 286 stuff might be a bit far away, but not knowing that x86-64 first appeared in the Prescott line is really difficult to understand, and is very basic. This is made more so because of all the rumors that the processor was made to support it, but Intel was hiding it so as to not undercut the Itanium. In time, it was proven true.

    Please, don't spread misinformation. Someone will repeat this stuff, and then someone else will, and it becomes 'fact' despite being wrong. If you publish a book, make a friggin effort! I'm sure I could errors the rest of the way, but it's just too annoying for me to wade through this rubbish.

    By the way, the term CPU bus is an ambiguous one. The CPU has multiple buses, and if you used that term with me, I'd wonder which one you were referring to. Find a more accurate term, like PCI-E bus if that's what you are trying to say.
Display more comments