ADVANCED COMPUTER ARCHITECTUREAND PARALLEL PROCESSING

ADVANCED COMPUTER ARCHITECTURE
AND PARALLEL PROCESSING

Shared memory architecture
1- Shared memory systems form a major category of multiprocessors.
2- In this category all processors share a global memory.
3- Communication between tasks running on different processors is performed
through writing to and reading from the global memory.
4- All interprocessor coordination and synchronization is also accomplished via
the global memory.
5- A shared memory computer system consists of :
- a set of independent processors.
- a set of memory modules, and an interconnection network as shown in
following Figure 4.1
6- Two main problems need to be addressed when designing a shared memory
system:
• performance degradation due to contention, and
• coherence problems
2

Shared memory architecture cont.
3

Shared memory architecture cont
• Performance degradation might happen when multiple processors are trying to
access the shared memory simultaneously. ?
• Caches : having multiple copies of data, spread throughout the caches might lead
to a coherence problem
- The copies in the caches are coherent if they are all equal to the same value.
- If one of the processors writes over the value of one of the copies, then the copy
becomes inconsistent because it no longer equals the value of the other copies.
4

4.1 CLASSIFICATION OF SHARED
MEMORY SYSTEMS
• The simplest shared memory system consists of one memory module (M)
that can be accessed from two processors (P1 and P2).
• An arbitration unit within the memory module passes requests through to a
memory controller. If the memory module is not busy and a single request
arrives, then the arbitration unit passes that request to the memory
controller and the request is satisfied. The module is placed in the busy
state while a request is being serviced. If a new request arrives while the
memory is busy servicing a previous request, the memory module sends a
wait signal, through the memory controller, to the processor making the new
request. 5

4.1 CLASSIFICATION OF SHARED
MEMORY SYSTEMS cont.
• In response, the requesting processor may hold its request on the line until the
memory becomes free or it may repeat its request some time later. If the
arbitration unit receives two requests, it selects one of them and passes it to the
memory controller. Again, the denied request can be either held to be served
next or it may be repeated some time later.
• Categories of shared memory systems Based on the interconnection network:
 Uniform Memory Access (UMA) .
 Non-uniform Memory Access (NUMA)
 Cache Only Memory Architecture (COMA)
6

Uniform Memory Access (UMA)
 In the UMA system a shared memory is accessible by all processors through an
interconnection network in the same way a single processor accesses its
memory.
 All processors have equal access time to any memory location.
 The interconnection network used: a single bus, multiple buses, or a crossbar
switch(diagonal,straight).
 Called it SMP (symmetric multiprocessor) systems?
 access to shared memory is balanced (Each processor has equal opportunity to
read/write to memory, including equal access speed).
7

Nonuniform Memory Access
(NUMA)
• each processor has part of the shared memory attached. The memory has a
single address space. Therefore, any processor could access any memory
location directly using its real address. However, the access time to modules
depends on the distance to the processor.
• the tree and the hierarchical bus networks are architectures used to interconnect
processors.
8

Cache-Only Memory Architecture
(COMA)
• Similar to the NUMA, each processor has part of the shared memory in the
COMA. However, in this case the shared memory consists of cache memory.
• requires that data be migrated to the processor requesting it.
• There is no memory hierarchy and the address space is made of all the caches.
There is a cache directory (D) that helps in remote cache access.
9

4.2 BUS-BASED SYMMETRIC
MULTIPROCESSORS
 Shared memory systems can be designed using :
• bus-based (simplest network for shared memory systems).
• switch-based interconnection networks (message passing).
 The bus/cache architecture alleviates the need for expensive multi-ported
memories and interface circuitry as well as the need to adopt a message-
passing paradigm when developing application software.
 the bus may get saturated if multiple processors are trying to access the shared
memory (via the bus) simultaneously. (contention problem) solve uses Caches
 High speed caches connected to each processor on one side and the bus on the
other side mean that local copies of instructions and data can be supplied at the
highest possible rate.
 If the local processor finds all of its instructions and data in the local cache hit
rate is 100%, otherwise miss rate of a cache, so must be copied from the
global memory, across the bus, into the cache, and then passed on to the local
processor. 10

MULTIPROCESSORS cont.
 One of the goals of the cache is to maintain a high hit rate, or low miss rate
under high processor loads. A high hit rate means the processors are not using
the bus as much. Hit rates are determined by a number of factors, ranging from
the application programs being run to the manner in which cache hardware is
implemented.
11

• We define the variables for hit rate, number of processors, processor speed, bus
speed, and processor duty cycle rates as follows:
 N = number of processors;
 h = hit rate of each cache, assumed to be the same for all caches;
 (1 - h) = miss rate of all caches;
 B = bandwidth of the bus, measured in cycles/second;
 I = processor duty cycle, assumed to be identical for all processors, in
fetches/cycle; and
 V = peak processor speed, in fetches/second
12

• The effective bandwidth of the bus is BI fetches/second.
If each processor is running at a speed of V, then misses are being generated
at a rate of V(1 - h). For an N-processor system, misses are simultaneously
being generated at a rate of N(1 - h)V. This leads to saturation of the bus when
N processors simultaneously try to access the bus. That is, 𝑵 𝟏 − 𝒉 𝑽 ≤ 𝑩𝑰.
The maximum number of processors with cache memories that the bus can
support is given by the relation,
𝑁 ≤
𝐵𝐼
1−ℎ 𝑉
13

• Example 1
Suppose a shared memory system is constructed from processors that can
execute V = 107 instructions/s and the processor duty cycle I = 1. The caches
are designed to support a hit rate of 97%, and the bus supports a peak
bandwidth of B = 106 cycles/s. Then, (1 - h) = 0.03, and the maximum number
of processors N Is 𝑁 ≤
106∗1
(0.03∗107)
=3.33.
Thus, the system we have in mind can support only three processors!
We might ask what hit rate is needed to support a 30-processor system. In this
case, h = 1- BI/NV = 1 - (106(1))/((30)(107)) = 1 - 1/300, so for the system we
have in mind, h = 0.9967. Increasing h by 2.8%. results in supporting a factor of
ten more processors.
14

4.3 BASIC CACHE COHERENCY
METHODS
• Multiple copies of data, spread throughout the caches, lead to a coherence
problem among the caches. The copies in the caches are coherent if they all
equal the same value. However, if one of the processors writes over the value of
one of the copies, then the copy becomes inconsistent because it no longer
equals the value of the other copies. If data are allowed to become inconsistent
(incoherent), incorrect results will be propagated through the system, leading to
incorrect final results. Cache coherence algorithms are needed to maintain a
level of consistency throughout the parallel system.
Copies data in the
caches
coherence inconsistent
15

Cache–Memory Coherence
 In a single cache system, coherence between memory and the cache is
maintained using one of two policies:
• write-through:
The memory is updated every time the cache is updated.
• write-back:
The memory is updated only when the block in the cache is being
replaced.
16

Cache–Cache Coherence
• In multiprocessing system, when a task running on processor P requests the data in
global memory location X, for example, the contents of X are copied to processor P’s
local cache, where it is passed on to P. Now, suppose processor Q also accesses X.
What happens if Q wants to write a new value over the old value of X?
 There are two fundamental cache coherence policies:
 write-invalidate
Maintains consistency by reading from local caches until a write occurs.
 write-update
Maintains consistency by immediately updating all copies in all caches.
18

Shared Memory System
Coherence
• The four combinations to maintain coherence among all caches and
global memory are:
 Write-update and write-through
 Write-update and write-back
 Write-invalidate and write-through
 Write-invalidate and write-back
• If we permit a write-update and write-through directly on global memory
location X, the bus would start to get busy and ultimately all processors
would be idle while waiting for writes to complete. In write-update and write-
back, only copies in all caches are updated. On the contrary, if the write is
limited to the copy of X in cache Q, the caches become inconsistent on X.
Setting the dirty bit prevents the spread of inconsistent values of X, but at
some point, the inconsistent copies must be updated.
20

ADVANCED COMPUTER ARCHITECTUREAND PARALLEL PROCESSING

More Related Content

What's hot (20)

Similar to ADVANCED COMPUTER ARCHITECTUREAND PARALLEL PROCESSING (20)

Recently uploaded (20)