Latency oriented processor architecture: Difference between revisions

Revision as of 19:01, 5 January 2017

Revision as of 04:32, 17 March 2017

Line 3:

'''Latency oriented processor architecture''' is the [[microarchitecture]] of a [[microprocessor]] designed to serve a serial computing [[Thread (computing)|thread]] with a low latency. This is typical of most [[Central Processing Unit]]s (CPU) being developed since the 1970s. These architectures, in general, aim to execute as many instructions as possible belonging to a single serial thread, in a given window of time; however, the time to execute a single instruction completely from fetch to retire stages may vary from a few cycles to even a few hundred cycles in some cases.<ref>{{cite book| author1=John Paul Shen |author2=Mikko H. Lipasti |year=2013 |title=Modern Processor Design |publisher=McGraw-Hill Professional |isbn=1478607831}}</ref>{{page needed|date=November 2016}} Latency oriented processor architectures are the opposite of throughput-oriented processors which concern themselves more with the total [[throughput]] of the system, rather than the service [[Latency (engineering)|latencies]] for all individual threads that they work on.<ref name=YanSohilin2016>{{cite book|author=Yan Solihin |year=2016 |title=Fundamentals of Parallel Multicore Architecture |publisher=Chapman & Hall/CRC Computational Science |isbn=978-1482211184}}</ref>{{page needed|date=November 2016}}<ref name=GarlandKirk>{{cite journal|title=Understanding Throughput-Oriented Architectures |author1=Michael Garland |author2=David B. Kirk |journal=Communications of the ACM |volume=53 |number=11 |pages=58–66}}</ref>

==[[Flynn's taxonomy]]==

{{Main|Flynn's taxonomy}}

Latency oriented processor architectures would normally fall into the category of [[SISD]] classification under flynn's taxonomy. This implies a typical characteristic of latency oriented processor architectures is to execute a single task operating on a single data stream. Some [[SIMD]] style multimedia extensions of popular instruction sets, such as Intel [[MMX (instruction set)|MMX]] and [[Streaming SIMD Extensions|SSE]] instructions, should also fall under the category of latency oriented processor architectures;<ref name=YanSohilin2016/> because, although they operate on a large data set, their primary goal is also to reduce overall latency for the entire task at hand.

Line 15 ⟶ 16:

Most architectures today use shorter and simpler instructions, like the [[load/store architecture]], which help in optimizing the instruction pipeline for faster execution. Instructions are usually all of the same size which also helps in optimizing the instruction fetch logic. Such an ISA is called a [[Reduced instruction set computing|RISC]] architecture.<ref>{{cite conference|last1=Bhandarkar|first1=Dileep|last2=Clark|first2=Douglas W.|title=Performance from Architecture: Comparing a RISC and a CISC with Similar Hardware Organization|journal=Proceedings of the Fourth International Conference on Architectural Support for Programming Languages and Operating Systems|date=1 January 1991|pages=310–319|doi=10.1145/106972.107003|url=http://dl.acm.org/citation.cfm?id=107003&CFID=860927590&CFTOKEN=39315780|publisher=ACM}}</ref>

===Instruction Pipeliningpipelining===

{{Main article|instruction pipeline}}

Pipelining overlaps execution of multiple instructions from the same executing thread in order to increase clock frequency or to increase the number of instructions that complete per unit time; thereby reducing the overall execution time for a thread. Instead of waiting for a single instruction to complete all its execution stages, multiple instructions are processed simultaneously, at their respective stages inside the pipeline. {{efn|''Computer Organization and Design: The Hardware/software Interface'', Chapter 4<ref name="interface"/>}}

Line 21 ⟶ 22:

===Register-renaming===

{{Main article|Register renaming}}

This technique is used to effectively increase the total register file size than that specified in the ISA to programmers, and to eliminate false dependencies. Suppose we have two consecutive instructions which reference the same register. The first reads the register while the second writes to it. To maintain correctness of the program, it is essential to make sure that the second instruction does not write to the register before the first can read its original value. This is an example of a [[WARWrite after read|Write-After-Read (WAR)]] dependency. To eliminate this dependency, the pipeline would 'rename' the instruction internally by assigning it to an internal register. The instruction is therefore allowed to execute and results produced by it will now be immediately available to all subsequent instructions, even though the actual destination register intended by the program will be written to later. Similarly if both the instructions simply meant to write to the same register [[data dependency|Write-After-Write (WAW)]], the pipeline would rename them and ensure that their results are available to future instructions without the need to serialize their execution. {{efn|''Computer Architecture: A Quantitative Approach'', Section 3.1<ref name="quant"/>}}

===Memory Organizationorganization===

{{Main article|Memory hierarchy}}

The different levels of memory, which includes [[Cache (computing)|caches]], [[main memory]] and [[non-volatile storage]] like hard disks (where the program instructions and data reside), are designed to exploit [[Locality of reference|spatial locality]] and [[temporal locality]] to reduce the total [[memory access time]]. The less time the processor spends waiting for data to be fetched from memory, the lower number of instructions consume pipeline resources while just sitting idle and doing no useful work. The instruction pipeline will be completely stalled if all its internal buffers (for example [[reservation station]]s) are filled to their respective capacities. Hence, if instructions consume less number of idle cycles while inside the pipeline, there is a greater chance of exploiting [[Instruction level parallelism]] (ILP) as the fetch logic can pull in greater number of instructions from the cache/memory per unit time. {{efn|''Computer Organization and Design: The Hardware/software Interface'', Chapter 5<ref name="interface"/>}}

===Speculative Executionexecution===

{{Main article|Branch predictor}}

A major cause for pipeline stalls are control flow dependencies, i.e. when the outcome of a branch instruction is not known in advance (which is usually the case). Many architectures today use branch predictor components to guess the outcome of a branch. Execution continues along the predicted path for the program but instructions are tagged as speculative. If the guess turns out to be correct, then the instructions are allowed to complete successfully and to update their results back to register file/memory. If the guess was incorrect, then all speculative instructions are flushed from the pipeline and execution (re)starts along the actual correct path for the program. By maintaining a high prediction accuracy, the pipeline is able to significantly increase throughput for the executing thread. {{efn|''Computer Architecture: A Quantitative Approach'', Section 3.3<ref name="quant"/>}}

Line 35 ⟶ 36:

Not all instructions in a thread take the same amount of time to execute. Superscalar pipelines usually have multiple possible paths for instructions depending upon current state and the instruction type itself. Hence, to increase [[instructions per cycle]] (IPC) the pipeline allows execution of instructions out-of-order so that instructions later in the program are not stalled due to an instruction which will take longer to complete. All instructions are registered in a re-order buffer when they are fetched by the pipeline and allowed to retire (i.e. write back their results) in the order of the original program so as to maintain correctness. {{efn|''Computer Architecture: A Quantitative Approach'', Sections 3.4, 3.5<ref name="quant"/>}}

===Superscalar Executionexecution===

{{Main article|Superscalar}}

A super-scalar instruction pipeline pulls in multiple instructions in every clock cycle, as opposed to a simple scalar pipeline. This increases [[Instruction level parallelism]] (ILP) as many times as the number of instructions fetched in each cycle, except when the pipeline is stalled due to data or control flow dependencies. Even though the retire rate of superscalar pipelines is usually less than their fetch rate, the overall number of instructions executed per unit time (> 1) is generally greater than a scalar pipeline. {{efn|''Computer Architecture: A Quantitative Approach'', Sections 3.6-3.8<ref name="quant"/>}}

==Contrast with Throughputthroughput oriented processor architectures==

In contrast, a [[throughput]] oriented processor architecture is designed to maximize the amount of 'useful work' done in a significant window of time. Useful work refers to large calculations on a significant amount of data. They do this by parallelizing the work load so that many calculations can be performed simultaneously. The calculations may belong to a single task or a limited number of multiple tasks. The total time required to complete 1 execution is significantly larger than that of a latency oriented processor architecture, however, the total time to complete a large set of calculations is significantly reduced. Latency is often sacrificed in order to achieve a higher throughput per cycle.<ref name=GarlandKirk/> As a result, a latency oriented processor may complete a single calculation significantly faster than a throughput-oriented processor; however, the throughput-oriented processor could be partway through hundreds of such computations by the time the latency oriented processor completes 1 calculation.<ref name=YanSohilin2016/>