CSC 456 Fall 2013/1c wa: Difference between revisions

From Expertiza_Wiki
Jump to navigation Jump to search
No edit summary
No edit summary
Line 38: Line 38:


miss rate reported, spec benchmarks
miss rate reported, spec benchmarks




Line 46: Line 43:
'''L1 & L2 cache'''
'''L1 & L2 cache'''
  {|border=1
  {|border=1
    | System
 
    | Year
| System
    | L1 Size (cache)
| Year
    | L1 Speed (cc)
| L1 Size (cache)
    | L1 Associativity
| L1 Speed (cc)
    | L2 Size
| L1 Associativity
    | L2 Speed
| L2 Size
    | L2 Associativity
| L2 Speed
    | L2 On-Chip?
| L2 Associativity
    | L3 Size
| L2 On-Chip?
    | L3 Speed
| L3 Size
    | L3 Associativity
| L3 Speed
    | L3 On-Chip?
| L3 Associativity
    | Main Mem Penalty (cc)
| L3 On-Chip?
    | Notes:
| Main Mem Penalty (cc)
    |-
| Notes:
    | Pentium 4
|-
    | 10/2000
    | 8 KB (trace)
    | 2
    | 4 Way
    | 256KB
    | full speed
    | 8 Way
    | Yes
    | None
    |
    |
    |
    | 4 (Mem controller)
    |
    |-
    | Alpha 21064(DEC)
    | 09/1992
    | (8+8) KB
    | 1
    | Direct
    | 128kb - 16mb
    | 3-16 cc
    | Direct
    | No
    | 8 cc
    |
    |-
    | UltraSPARC III
    | ?/2001
    | (32+64) KB
    | 1
    | 4 Way
    | 2-8MB
    | 2-3
    | ?
    | No
    | 4 (Mem controller)
    |
    |-
    | Alpha 21064(DEC)
    | 09/1992
    | (8+8) KB
    | 1
    | Direct
    | 128kb - 16mb
    | 3-16 cc
    | Direct
    | No
    | 8 cc
    |
    |-
    | UltraSPARC
    | 06/1995
    | (16+16) KB
    | 1
    | 2-Way & Direct
    | 512KB - 4MB
    | 1, pipe=3
    | Direct
    | No
    | N/A
    |
    |
    |
    | 2-3
    | 64-bit w/ Core@200MHz
    |-
    | Alpha 21164(DEC)
    | 01/1995
    | (8+8) KB
    | 1
    | Direct
    | 96 Kb
    | 2
    | 3 way set
    | Yes
    | ?
    |-
    | SuperSPARC
    | ?/1992
    | (16+20) KB
    | ?
    | 4 & 5 way set
    | 0 - 2 MB
    | ?
    | ?
    | No
    | N/A
    |
    |
    |
    | ?
    | Used to render Toy Story, Core @ 40MHz,
    |-
| IBM 360/85
| IBM 360/85
    | 1968
| 1968
    | 16 - 32 KB
| 16 - 32 KB
    | ?
| ?
    | Sector
| Sector
    | None
| None
    | N/A
| N/A
    | N/A
| N/A
    | N/A
| N/A
    | None
| None
    | N/A
| N/A
    | N/A
| N/A
    | N/A
| N/A
    | ?
| ?
    | First processor with a cache, clock speed 12.5MHz
| First processor with a cache, clock speed 12.5MHz
    |-
|-
| IBM 486
| 1989
| 8 kb
| ?
| L1 Associativity
| 256 kb
|
|
| no
|
|
|
|
|
|
|-
| Intel 80486
| Intel 80486
    | 1989
| 1989
    | 8 KB
| 8 KB
    | ?
| ?
    | 4-way associative
| 4-way associative
    | None
| None
    | N/A
| N/A
    | N/A
| N/A
    | N/A
| N/A
    | None
| None
    | N/A
| N/A
    | N/A
| N/A
    | N/A
| N/A
    | ?
| ?
    | First processor with a cache, clock speed 12.5MHz
| First processor with a cache, clock speed 12.5MHz
 
|
    |
|-
|-
| IBM 486
| SuperSPARC
    | 1989
| ?/1992
    | 8 kb
| (16+20) KB
    | ?
| ?
    | L1 Associativity
| 4 & 5 way set
    | 256 kb
| 0 - 2 MB
    |  
| ?
    |  
| ?
    | no
| No
    |  
| N/A
    |  
|
    |  
|
    |
|
    |  
| ?
    |  
| Used to render Toy Story, Core @ 40MHz,
|-
| Alpha 21064(DEC)
| 09/1992
| (8+8) KB
| 1
| Direct
| 128kb - 16mb
| 3-16 cc
| Direct
| No
| 8 cc
|
|-
| Alpha 21064(DEC)
| 09/1992
| (8+8) KB
| 1
| Direct
| 128kb - 16mb
| 3-16 cc
| Direct
| No
| 8 cc
|
|-
| UltraSPARC
| 06/1995
| (16+16) KB
| 1
| 2-Way & Direct
| 512KB - 4MB
| 1, pipe=3
| Direct
| No
| N/A
|
|
|
| 2-3
| 64-bit w/ Core@200MHz
|-
| Alpha 21164(DEC)
| 01/1995
| (8+8) KB
| 1
| Direct
| 96 Kb
| 2
| 3 way set
| Yes
| ?
|-
| K6-III
| 1999
| 32 kib
| 100 mhz
| 2 way
| 256 kib
| 100 mhz
| 4 way
| yes
| 1 mb
|  
|
| no
|  
|  
|-
|-
| K6-III
| Pentium 4
    | 1999
| 10/2000
    | 32 kib
| 8 KB (trace)
    | 100 mhz
| 2
    | 2 way
| 4 Way
    | 256 kib
| 256KB
    | 100 mhz
| full speed
    | 4 way
| 8 Way
    | yes
| Yes
    | 1 mb
| None
    |  
|
    |
|  
    | no
|  
    |  
| 4 (Mem controller)
    |  
|
|-
| UltraSPARC III
| ?/2001
| (32+64) KB
| 1
| 4 Way
| 2-8MB
| 2-3
| ?
| No
| 4 (Mem controller)
|
|-
|-
| Itanium 2
| Itanium 2
    | 2002
| 2002
    | 16 KB
| 16 KB
    | 1
| 1
    | 4 -way
| 4 -way
    | 256KB
| 256KB
    | 5
| 5
    | 8-way
| 8-way
    | yes
| yes
    | 3 MB
| 3 MB
    | 12
| 12
    | 12 way (4 per MB)
| 12 way (4 per MB)
    | Yes
| Yes
    | ?
| ?
    | Notes:
| Notes:
|}
|}


----
----

Revision as of 02:24, 25 September 2013

Trends in cache size and organization

Task 1c. Trends in cache size and organization Over the years, caches have grown larger--up to a point, and then L1 caches tended for awhile to decrease in size. Why? In the early 1980s, associativity increased; beginning about 1990, it decreased, and then by about 2000, it was increasing again. Why? When was the first machine with an L2 cache? An L3 cache? How fast were the various levels of caches, and how did this speed compare to main memory? There is a wealth of information to bring to bear on this topic.


To-Do:
find first L3 and fill in
order the list by year
analyze: average cc speed for a given level, main mem cc speed evolution


Theory: Cache Associativity decreased as cache size became larger because it became too expensive to have to search the cache each time once the cache was too large. Also, bigger the cache size as a percentage of main memory, the less need for associativity. But while caches and main memory have both grown, main memory size has grown faster in the 2000’s. So when the percentage of cache to main memory goes down associativity needs to increase.


The Pentium/Pentium (1995)pro was the first processor to have the l2 cache on the processor chip. Before this, the l2 cache was an option to add on to the motherboard. [1]

Systems to consider in table

Pentium
amd
Mips
sun-microsystems: sparc
ibm: power pc
DEC: alpha

Penalty <100 when before 2000 after 2000 started to increase to get to main memory
< 20 1 level fine
<=100 2 level
>=200 3 level

miss rate reported, spec benchmarks >=200 3 level

miss rate reported, spec benchmarks


L1 & L2 cache

System Year L1 Size (cache) L1 Speed (cc) L1 Associativity L2 Size L2 Speed L2 Associativity L2 On-Chip? L3 Size L3 Speed L3 Associativity L3 On-Chip? Main Mem Penalty (cc) Notes:
IBM 360/85 1968 16 - 32 KB ? Sector None N/A N/A N/A None N/A N/A N/A ? First processor with a cache, clock speed 12.5MHz
IBM 486 1989 8 kb ? L1 Associativity 256 kb no
Intel 80486 1989 8 KB ? 4-way associative None N/A N/A N/A None N/A N/A N/A ? First processor with a cache, clock speed 12.5MHz
SuperSPARC ?/1992 (16+20) KB ? 4 & 5 way set 0 - 2 MB ? ? No N/A ? Used to render Toy Story, Core @ 40MHz,
Alpha 21064(DEC) 09/1992 (8+8) KB 1 Direct 128kb - 16mb 3-16 cc Direct No 8 cc
Alpha 21064(DEC) 09/1992 (8+8) KB 1 Direct 128kb - 16mb 3-16 cc Direct No 8 cc
UltraSPARC 06/1995 (16+16) KB 1 2-Way & Direct 512KB - 4MB 1, pipe=3 Direct No N/A 2-3 64-bit w/ Core@200MHz
Alpha 21164(DEC) 01/1995 (8+8) KB 1 Direct 96 Kb 2 3 way set Yes ?
K6-III 1999 32 kib 100 mhz 2 way 256 kib 100 mhz 4 way yes 1 mb no
Pentium 4 10/2000 8 KB (trace) 2 4 Way 256KB full speed 8 Way Yes None 4 (Mem controller)
UltraSPARC III ?/2001 (32+64) KB 1 4 Way 2-8MB 2-3 ? No 4 (Mem controller)
Itanium 2 2002 16 KB 1 4 -way 256KB 5 8-way yes 3 MB 12 12 way (4 per MB) Yes ? Notes:

References Itanium Specs(p.20)
Ref1
Intel Processors
First on-board L1
Cache Trend Table‎
Sector Caches

Retrieved from "https://wiki.expertiza.ncsu.edu/index.php?title=CSC_456_Fall_2013/1c_wa&oldid=79104"