1 / 60

Chapter 7 Memory Systems

Chapter 7 Memory Systems. Secondary Memory (Disk) M D. Secondary Memory (Archival) M A. Cache Memory M C. Main Memory M M. Memory Hierarchy. CPU ---- Registers. Memory Content: M C  M M  M D  M A. Memory Parameters: Access Time: increase with distance from CPU

wei
Télécharger la présentation

Chapter 7 Memory Systems

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Chapter 7Memory Systems

  2. Secondary Memory (Disk) MD Secondary Memory (Archival) MA Cache Memory MC Main Memory MM Memory Hierarchy CPU ---- Registers Memory Content: MC MM MD MA • Memory Parameters: • Access Time: increase with distance from CPU • Cost/Bit: decrease with distance from CPU • Capacity: increase with distance from CPU ELEC 5200/6200 From Patterson/Hennessey Slides

  3. Exploiting Memory Hierarchy • Users want large and fast memories! SRAM access times are .5 – 5ns at cost of $4000 to $10,000 per GB.DRAM access times are 50-70ns at cost of $100 to $200 per GB.Disk access times are 5 to 20 million ns at cost of $.50 to $2 per GB. • Try and give it to them anyway • build a memory hierarchy 2004 ELEC 5200/6200 From Patterson/Hennessey Slides

  4. Locality of Reference • A principle that makes having a memory hierarchy a good idea • If an item is referenced:temporal locality: it will tend to be referenced again soon spatial locality: nearby items will tend to be referenced soon Why does code have locality? • Our initial focus: two levels (upper, lower) • block: minimum unit of data • hit: data requested is in the upper level • miss: data requested is not in the upper level ELEC 5200/6200 From Patterson/Hennessey Slides

  5. Memory Performance • Access Time (latency) – from initiation of memory read to valid data returned to CPU • Cache: TA typically = CPU cycle time • Main: TA typically = multiple CPU cycles • Disk: TA typically several orders of magnitude greater than CPU cycle time (software controlled) • Bandwidth (throughput) = number of bytes transferred per second BW = (bytes/transfer) x (transfers/second) Ex. Synchronous Dynamic RAM “burst transfers” ELEC 5200/6200 From Patterson/Hennessey Slides

  6. Memory Access Modes • Random Access – locations accessed in any order with access time independent of location • SRAM • ROM/EPROM/EEPROM/Flash technologies • DRAM (some special modes available) • Cache • Serial Access – locations must be accessed in a predetermined sequence (ex. tape) • Disk – position read element over a track, and then serial access within the track • CDROM – data recorded in a “spiral” ELEC 5200/6200 From Patterson/Hennessey Slides

  7. Memory Write Strategies • Writing a new value of an item in a hierarchical memory which has copies in multiple memory levels • “Write-through” strategy • Update all levels of memory on each write • All levels “consistent” or “coherent” at all times • “Write-back” strategy • Update closest (fastest) copy during write operation • Copy the value to other levels at “convenient time” • Memory state temporarily inconsistent ELEC 5200/6200 From Patterson/Hennessey Slides

  8. Cache • Two issues: • How do we know if a data item is in the cache? • If it is, how do we find it? • Three basic cache organizations: • Fully associative • Direct mapped • Set associative • Assume block size is one word of data ELEC 5200/6200 From Patterson/Hennessey Slides

  9. = = = = Fully Associative Cache • Associative memory – access by “content” rather than address Concurrently compare CPU address to all cache address fields 1 2 3 4 5 6 7 8 9 10 Valid entry Address from CPU Cache Memory Main Memory ELEC 5200/6200 From Patterson/Hennessey Slides

  10. Hits vs. Misses • Read hits • this is what we want! • Read misses • stall the CPU, fetch block from memory, deliver to cache, restart • Write hits: • can replace data in cache and memory (write-through) • write the data only into the cache (write-back the cache later) • Write misses: • read the entire block into the cache, then write the word ELEC 5200/6200 From Patterson/Hennessey Slides

  11. = = = = Fully Associative Cache • CPU memory read from address 5 1 2 3 4 5 6 7 8 9 10 Address from CPU 5 Cache “Hit” Data to CPU Main Memory ELEC 5200/6200 From Patterson/Hennessey Slides

  12. = = = = Fully Associative Cache • CPU memory read from address 8 1 2 3 4 5 6 7 8 9 10 Address from CPU 8 Cache CPU reads Main Memory after Cache miss “Miss” (no “hit”) Main Memory ELEC 5200/6200 From Patterson/Hennessey Slides

  13. = = = = Cache Memory Structures • Update cache with data from address 8 1 2 3 4 5 6 7 8 9 10 Cache Use “open” cache line Main Memory ELEC 5200/6200 From Patterson/Hennessey Slides

  14. = = = = Cache Memory Structures • Where should the CPU place data from address 2?? • no open cache lines • must replace data in some line 1 2 3 4 5 6 7 8 9 10 Address 2 Cache “Miss” (no “hit”) Main Memory ELEC 5200/6200 From Patterson/Hennessey Slides

  15. = = = = Multi-Word Blocks • Cache line holds a multi-word block from main memory • Take advantage of spatial locality of reference 0 1 2 3 4 5 6 7 8 9 Block 0 Block # Word offset Blk 1 2 Cache 3 4 Main Memory ELEC 5200/6200 From Patterson/Hennessey Slides

  16. Direct Mapped Cache • Mapping: address is modulo the number of blocks in the cache Main memory address tag index cache line ELEC 5200/6200 From Patterson/Hennessey Slides

  17. Direct Mapped Cache • For MIPS:What kind of locality are we taking advantage of? ELEC 5200/6200 From Patterson/Hennessey Slides

  18. Direct Mapped Cache • Taking advantage of spatial locality: ELEC 5200/6200 From Patterson/Hennessey Slides

  19. Motorola 68020/68030 Cache ELEC 5200/6200 From Patterson/Hennessey Slides

  20. Cache Misses • Compulsory – cache “empty” at startup • Capacity – cache unable to hold entire working set • Conflict – two memory loc’s map to same cache line Conflict (1-way) Conflict (2-way) Compulsory – “flat” but negligible in this example Conflict (8-way) Fig. 7.31 ELEC 5200/6200 From Patterson/Hennessey Slides

  21. Decreasing miss ratio with associativity • Problems: • Fully associative – high cost, misses only when capacity reached since any item can go in any line • Direct-mapped: less expensive, but can hold only one item with a given index, leading to conflicts • Compromise: Set Associative Cache • “N-way” set associative cache has N direct-mapped caches • One “set” = N lines with a particular index • Item with index K can be placed in line K of any of the N direct-mapped caches • Results in fewer misses due to conflicts ELEC 5200/6200 From Patterson/Hennessey Slides

  22. Example: 4-way set-associative cache ELEC 5200/6200 From Patterson/Hennessey Slides

  23. VAX 11/780 Cache ELEC 5200/6200 From Patterson/Hennessey Slides

  24. Intel 80486/Pentium L1 Cache 80486 4-way Set Associative (Pentium 2-way) 2K (486) 4K (Pentium) 128 Lines (sets) LRU Valid* (3) (4) Tags (20 bit) Data line = 16 bytes (486) 32 bytes (Pentium) * Instruction cache: 1 valid bit per line * Data cache: 2-bit MESI state per line 80486 – 8K unified I/D Pentium – 8K/8K I/D Pentium II/III – 16K/16K I/D Write-back/Write-through programmable ELEC 5200/6200 From Patterson/Hennessey Slides

  25. MESI Cache Line State • MESI = 2-bit cache line “state” for “write-back” cache • I = Invalid (line does not contain valid data) • S = Line valid, with shared access – no writes allowed • E = Line valid, with exclusive access – data may be written • M = Line valid & modified since read from main (must be rewritten to main memory) read I S invalidate read invalidate invalidate E M write ELEC 5200/6200 From Patterson/Hennessey Slides

  26. Intel Approximated-LRU Replacement • 3-bit number B2B1B0 assigned to each “set” of 4 lines • Access L0/L1 – set B0=1 • Access L2/L3 – set B0=0 • Access L0 – set B1=1 or L1 – set B1=0 • Access L2 – set B2=1 or L1 – set B2=0 B0 B0B1 = 00 – replace L0 01 – replace L1 B0B2 = 10 – replace L2 11 – replace L3 0 1 B1 B2 0 1 0 1 L0 L1 L2 L3 ELEC 5200/6200 From Patterson/Hennessey Slides

  27. Intel Pentium P4 vs. AMD Opteron ELEC 5200/6200 From Patterson/Hennessey Slides Figure 7.35

  28. Average Memory Access Time • Assume main accessed after cache miss detected: TC ,TM = cache and main memory access times HC = cache hit ratio TAavg = TC*HC + (1-HC)(TC + TMavg) = TC + (1-HC)(TMavg) miss penalty • Extending to 3rd level (disk): TAavg = TC + (1-HC)(TM) + (1-HC)(1-HM)(TDavg ) Note that TM << TDavg ELEC 5200/6200 From Patterson/Hennessey Slides

  29. Average Memory Access Time • Assume main accessed concurrent with cache access • abort main access on cache hit • main access already in progress on cache miss TA = TC*HC + (1-HC)(TM) Called “look-aside” cache (previous “look through”) Problem: main memory busy with aborted accesses (unacceptable if memory shared) ELEC 5200/6200 From Patterson/Hennessey Slides

  30. Performance • Increasing the block size tends to decrease miss rate: • Use split caches because there is more spatial locality in code: ELEC 5200/6200 From Patterson/Hennessey Slides

  31. Performance ELEC 5200/6200 From Patterson/Hennessey Slides

  32. Performance • Simplified model:execution time = (execution cycles + stall cycles) x cycle timestall cycles = # of instructions x miss ratio x miss penalty • Two ways of improving performance: • decreasing the miss ratio • decreasing the miss penalty What happens if we increase block size? ELEC 5200/6200 From Patterson/Hennessey Slides

  33. Example – “gcc”compiler on MIPS • Instruction count = IC • Instruction cache miss rate = 5% • Data cache miss rate = 10% • Clocks/instruction (CPI) = 4 if no misses/stalls • Miss penalty = 12 cycles (all misses) • Instruction frequencies: • lw = 22% of instructions executed • sw = 11% of instructions executed Question: What is the effect of cache misses on CPU performance (ex, on CPI)? ELEC 5200/6200 From Patterson/Hennessey Slides

  34. Decreasing miss penalty with multilevel caches • Add a second level cache: • often primary cache is on the same chip as the processor • use SRAMs to add another cache above primary memory (DRAM) • miss penalty goes down if data is in 2nd level cache • Example: • CPI of 1.0 on a 5 Ghz machine with a 5% miss rate, 100ns DRAM access • Adding 2nd level cache with 5ns access time decreases miss rate to .5% • Using multilevel caches: • try and optimize the hit time on the 1st level cache • try and optimize the miss rate on the 2nd level cache ELEC 5200/6200 From Patterson/Hennessey Slides

  35. Hardware Issues • Make reading multiple words easier by using banks of memory • It can get a lot more complicated... ELEC 5200/6200 From Patterson/Hennessey Slides

  36. Cache Complexities • Not always easy to understand implications of caches: Theoretical behavior of Radix sort vs. Quicksort Observed behavior of Radix sort vs. Quicksort ELEC 5200/6200 From Patterson/Hennessey Slides

  37. Cache Complexities • Here is why: • Memory system performance is often critical factor • multilevel caches, pipelined processors, make it harder to predict outcomes • Compiler optimizations to increase locality sometimes hurt ILP • Difficult to predict best algorithm: need experimental data ELEC 5200/6200 From Patterson/Hennessey Slides

  38. Virtual Memory • Main memory can act as a cache for the secondary storage (disk) • Advantages: • illusion of having more physical memory • program relocation • protection ELEC 5200/6200 From Patterson/Hennessey Slides

  39. Segmentation vs. Paging • Page: transparent to programmer • Virtual address partitioned into fixed-size pages • Physical memory likewise partitioned • Virtual pages map exactly into physical pages • Segment: defined by programmer/compiler • Compiler partitions program/data into segments of related information (code, data, etc.) • Segments easily protected (read-only, execute-only, etc.) • Segments can be of variable size • Memory manager must find sufficient free memory to hold each segment (fragmentation may occur) • Less common than paging ELEC 5200/6200 From Patterson/Hennessey Slides

  40. Pages: virtual memory blocks • Page faults: the data is not in memory, retrieve it from disk • huge miss penalty, thus pages should be fairly large (e.g., 4KB) • reducing page faults is important (LRU is worth the price) • can handle the faults in software instead of hardware • using write-through is too expensive so we use writeback ELEC 5200/6200 From Patterson/Hennessey Slides

  41. V i r t u a l p a g e n u m b e r P a g e t a b l e P h y s i c a l m e m o r y P h y s i c a l p a g e o r V a l i d d i s k a d d r e s s 1 1 1 1 0 1 1 0 D i s k s t o r a g e 1 1 0 1 Page Tables for Address Translation ELEC 5200/6200 From Patterson/Hennessey Slides

  42. Page Tables Page Descriptor ELEC 5200/6200 From Patterson/Hennessey Slides

  43. Segmented Memory Virtual Memory Physical Memory Memory Management Unit (MMU) S1 x S3 free S2 S3 x S1 Segment Table free S4 Sufficient free memory for segment S2, but fragmented, preventing loading of S2 ELEC 5200/6200 From Patterson/Hennessey Slides

  44. Mapping Segmented Address Virtual Address Physical Memory Segment # Displacement Segment Table + V prot limit base datum Protection Physical address Compare segment Limit to displacement ELEC 5200/6200 From Patterson/Hennessey Slides

  45. Combining Segmentation and Paging Virtual Address Physical Memory Segment # Page # Offset offset from top of page Physical page + V prot limit base V prot page # datum Page Table Segment Table • Partition segments into pages • Combine best features of both Page # Offset* Physical Address * offset same as in virtual address ELEC 5200/6200 From Patterson/Hennessey Slides

  46. Making Address Translation Fast • A cache for address translations: translation lookaside buffer (TLB) Typical values: 16-512 entries, miss-rate: .01% - 1% miss-penalty: 10 – 100 cycles ELEC 5200/6200 From Patterson/Hennessey Slides

  47. TLBs and caches ELEC 5200/6200 From Patterson/Hennessey Slides

  48. TLBs and Caches ELEC 5200/6200 From Patterson/Hennessey Slides

  49. Concurrent TLB & Cache Access • Normally access main memory cache after translating logical to physical address • Can do these concurrently to save time Virtual address Virtual page # Offset translate Physical address Offset Physical page # Cache format Tag Index Byte • Offset same in virtual & physical address, • Index available for cache line access while translation occurs • Check tag after translation and cache line access ELEC 5200/6200 From Patterson/Hennessey Slides

  50. Modern Systems ELEC 5200/6200 From Patterson/Hennessey Slides

More Related