2 25, 2005

第16回:「RISCとCISC」

コンピュータに処理を命令する方式として新しく RISC が考え出されてから、RISC 推進派がそれまでの方式に CISC という呼称を与えたということです。CISC をわざわざ日本語にする必要はないでしょうが、説明の目的で一般に「複合命令セットコンピュータ」とされています。

しかし、命名した RISC 陣営では従来方式の複雑すぎるアーキテクチャや過剰な機能に多少の揶揄をこめて complex と形容したのですが、「縮小」命令セットに対する「複合」はちょっと変かもしれません。複合にはふつう compound や composite が対応していますから、素人にとっていたずらに難解です。

同じ複雑でも、complicated や intricate は入り組んでおり、complicate や tangled は込み入っており、involved はややこしく、elaborate は精巧な、という具合に、さまざまな状態や事情があるものです。CISC の場合は、complicated が否定的な形容詞なので皮肉るほうでも敢えてこれを避けたのでしょう。

Electra や Oedipus などのコンプレックスはさておき、技術英語での complex は単に複雑なだけでなく、その複雑さが「高度」であることの随伴現象でもあるということは、長い歴史のある技術英語研究会の初年度で学習しています。


課題文
RISCとCISCの違い

POWER PCには68000シリーズのCISC (Complex Instruction Set Computer) アーキテクチャに代わり、RISC (Reduced Instruction Set Computing) と呼ばれるアーキテクチャが採用されているのが最大の特徴だ。

CPUは一般的にきわめて単純な命令しか持っていないため、複雑な処理は単純な命令を大量に組み合わせて行っている。しかし、複雑な処理を行わせるたびに膨大な量の単純命令をプログラミングしていては効率が悪い。そこで、CISCのCPUの中には複雑な処理命令を単純な命令の集まりに変換するためのマイクロコードと呼ばれる辞書が内蔵されている。

しかしこの仕組みではプログラミングは容易になるものの、命令が発行されるたびに逐次マイクロコードにアクセスする必要があるため実行速度が遅くなってしまう。また、個々の命令の長さが異なるためそれを監視する回路が必要になるという問題もあった。つまり、機能強化するたびにプロセッサがどんどん複雑になってしまうのだ。当然、構造が複雑になればなるほど内部的な信号のやり取りが増えるため、消費電力と発熱量も増加してしまう。

しかも、クロックが高速になればなるほど発熱量が増え、CPUの耐熱性には限界があるため、クロックスピードが上げられないという壁に突き当たってしまったのだ。

そこで、RISCではCISCで約300ほど用意されていた命令の数を使用頻度が高い100個前後まで一気に減らし、マイクロコードの代わりにそれらの命令を高速に実行する専用回路を用意した。しかも、すべての命令を一定の長さ(32bit)に統一することで監視回路をなくして、CPUの構造を簡潔にしたため、より高度な半導体技術を使って集積率が上げられるようにもなった。ただし、簡単な命令セットだけに絞られてしまったので命令の最適化が必要になり、コンパイラなどの開発ツールの負荷は大きく増えている。

POWER PCの構造

POWER PCは「バスインターフェイスユニット」「分岐処理ユニット」「ロード / ストアユニット」「整数演算ユニット」「浮動小数点演算ユニット」の5つのユニットで構成されている。


[テキスト1]

RISCとCISCの違い

POWER PCには68000シリーズのCISC (Complex Instruction Set Computer) アーキテクチャに代わり、RISC (Reduced Instruction Set Computing) と呼ばれるアーキテクチャが採用されているのが最大の特徴だ。

CPUは一般的にきわめて単純な命令しか持っていないため、複雑な処理は単純な命令を大量に組み合わせて行っている。しかし、複雑な処理を行わせるたびに膨大な量の単純命令をプログラミングしていては効率が悪い。そこで、CISCのCPUの中には複雑な処理命令を単純な命令の集まりに変換するためのマイクロコードと呼ばれる辞書が内蔵されている。


[訳例1]

CISC vs. RISC: What is the difference?

The most distinctive feature of the Power PC is that it employs a RISC (Reduced Instruction Set Computer) architecture CPU, replacing the CISC (Complex Instruction Set Computer) architecture used in 68000-series CPUs.

As a rule, the CPU (Central Processing Unit) is able to use only very simple instructions, and must combine huge sets of instructions in order to execute complicated tasks. However, programming vast numbers of simple instructions each time a complex task needs to be executed is horribly inefficient. A dictionary resident in CISC-based CPUs called microcode converts complex instructions into sets of simple instructions inside the processor.


[訳例2]

Differences between RISC and CISC

A major feature of the PowerPC is that it incorporates an architecture called RISC (Reduced Instruction Set Computing), (1) while 68000 series incorporate an architecture called CISC (Complex Instruction Set Computer).

CPUs generally contain extremely simple instructions only. Therefore, when a CPU performs a complex task, it has to combine a large number of the simple instructions. However, it would be inefficient if the CPU combines and (2) programs enormous amount of simple instructions each time (3) when it performs complex tasks. This was why CISC-based CPUs contain a dictionary, called microcode, for (4) translating a complex instruction into a series of simple instructions.

  1. I think you need to mention something about RISC being a replacement for or a later version of CISC. This paragraph does not establish a link between the RISC and CISC.
  2. Is this the right word? Does the sentence talk about programming something or executing or processing instructions?
  3. Since you already used "each time" you don"t need to say "when." This is a bit redundant.
  4. This is probably okay, but I think "converting" may be a better word.

[訳例3]

RISC vs. CISC: How do they compare (1) for performance?

PowerPC, a microprocessor chip (2) incorporated in Macintosh computers, mainly features an architecture called Reduced Instruction Set Computing (RISC) instead of a microprocessor called Complex Instruction Set Computing (CISC) used by (3) Mac's 68000 series.

Because (4) general CPUs understand only simple instructions, they execute large sets of simple instructions to process a more complex one. Executing a large (5) amount of instructions for every complex one, however, reduces efficiency. A CISC-based CPU has a built-in dictionary called microcode that converts a complex instruction into a set of simple instructions.

  1. I think this is all right. But the Japanese original doesn't limit the title to only performance. It may also want to (later on) talk about price or other matters. So, it may be safer just to say "How do they compare?"
  2. Can we be sure that POWER PC is only installed in Macintosh computer? That's probably true, but does the Japanese text say this?
  3. If you mentioned Macintosh computers earlier in the sentence, you probably don't need to mention it again here.
  4. Is this word necessary? What kind of CPU is not general? You may want to say "Because CPUs normally…"
  5. The meaning is correct but the choice of words is a bit off. Rather than "amount" you may want to use "number" or "volume."

[訳例4]

RISC vs. CISC

What makes a PowerPC microprocessor (1)different is that it employs a RISC (Reduced Instruction Set Computing) architecture, while the 68000-series (2) are designed based on the (3) operational scheme called CISC (Complex Instruction Set Computing).

The CPU (Central Processing Unit) normally has a set of basic instructions, which are combined in large numbers to perform complicated tasks. It would be a long and (4) tedious task, however, to program numerous instructions for each and every complex operation. To avoid this (5) problem, CISC chips adopt a (6) technique known as (7) microprogramming to implement the instruction set as a (8) sequence of (9) low level, (10) built-in instructions called microcode.

  1. You may want to say what it is different from. Just saying it is different may be too vague.
  2. Should this also be in the singular form; the rest of the sentence is written around the singular form.
  3. Are RISC and CISC operational schemes? In what way should they be referred to?

    I think they should be referred to as designs or (instruction set) architectures. I just hated to use "architecture" twice in the same paragraph.

  4. In terms of vocabulary, this is fine. The only question I have is one of logic; can a computer understand tedium? Normally, this word is used to express the human feeling of boredom.
  5. Does the previous sentence express a nuance of a problem? It may be better to simply not talk about avoiding problems or such things and directly talk about what improvements were made.
  6. Is this really a technique? Can a chip or processor make use of a technique?

    You can personify a processor, can't you?
  7. Because I am not an expert, I cannot say whether or not this term is correct. I like this word, but I have not seen it before. Does the Japanese text make mention of this?

    Me too neither. According to a Website, the concept of microprogramming, along with microprogram and microcode, was developed by the same person, Cambridge researcher Maurice Wilkes. But with your comment above, and knowing that it was as early as 1951 for general-purpose computers, now I doubt if it is applicable to today's microprocessors as well.

  8. Is "sequence" the same as "set"?
  9. Is "low-level" the same as "simple"?
  10. Is this set of built-in instructions the same as the dictionary? Does this sentence clearly express the process of converting one type of instruction into another type of instruction?

    I don't think microcode is a dictionary.

[訳例5]

What's the difference between RISC and CISC?

The (1) newest model of (2) Macintosh, POWER PC, has (3) a number of advantageous features over its predecessors. Chief among them is the RISC (Reduced Instruction Set Computer) architecture, replacing CISC (Complex Instruction Set Computer) used for 68000 processors.

In general, a CPU is only equipped with very simple instructions, so that it performs complex processing with (4) intensive computation of combined simple instructions. However, it is not efficient (5) to perform programming significant amounts of combinations of simple instructions every time the computer performs complex processing. This problem is solved by incorporating an (6) interpreter, called "(7) microcord", into (8) a CPU with a (9) CISC-type instruction set for converting complex processing into a set of simple instructions.

  1. Does the original article say that the machine is the newest model?
  2. Is it safe to say directly that the machine is a Macintosh? Does the original article mention the company"s name?
  3. Is it mentioned that there are a number of advantages or does the article merely say that the RISC is a distinctive feature of the PowerPC?
  4. Rather than "intensive" you may want to say "complicated" tasks.
  5. This is a bit hard to understand. You need to add a preposition (of) after "perform", but you may also want to think about what you want to say. Do you want to talk about performing of instructions or execution of instructions? This isn't clear.
  6. This is probably either a dictionary or converter, not an interpreter.
  7. Check spelling.
  8. This should probably be in the plural form.
  9. This is a bit long – how about "CISC-based"?

○ PowerPC
Performance Optimization With Enhanced RISC for Personal Computer

○ マイクロコードと呼ばれる辞書


専門家のご意見
microcodeの本質が「きわめて低レベルのプロセッサ動作定義コード」であることを考えると、「辞書」とまでは表現しない方が無難と思われます。

問題の個所は、かみ砕いた文面であるため、砕きすぎて誤認を誘っているのかなと思われます。

「そこで、 CISCのCPUの中には複雑な処理命令を単純な命令の集まりに変換するためのマイクロコードと呼ばれる辞書が内蔵されている。」の個所は、「・・・マイクロコードと呼ばれるきわめて低レベルのプロセッサ動作定義コードが内蔵されている。」とすれば、誤認の誘発はなくなります。

[テキスト2]

しかしこの仕組みではプログラミングは容易になるものの、命令が発行されるたびに逐次マイクロコードにアクセスする必要があるため実行速度が遅くなってしまう。また、個々の命令の長さが異なるためそれを監視する回路が必要になるという問題もあった。つまり、機能強化するたびにプロセッサがどんどん複雑になってしまうのだ。当然、構造が複雑になればなるほど内部的な信号のやり取りが増えるため、消費電力と発熱量も増加してしまう。

しかも、クロックが高速になればなるほど発熱量が増え、CPUの耐熱性には限界があるため、クロックスピードが上げられないという壁に突き当たってしまったのだ。

[訳例1]

While this process may make programming easier, the execution of instructions is slowed down because the CPU has to access the microcode every time an instruction is issued. And because the lengths of the instructions are varied, it is necessary to have a circuit to supervise the execution of the instructions. In other words, each enhancement of the functions necessitates increasingly complex processors. Naturally, the more complicated the architecture becomes, the greater the number of internal signals, and the greater the consumption of electricity and generation of heat.

An added complication is the faster the clock speed, the greater the amount of heat generated. And because there is a limit to how much heat a CPU can withstand, the clock speed cannot be raised above a certain level.

[訳例2]

With the microcode, programming itself became easier; however, execution speed became slower. (1) Because the microcode needed to be accessed each time an instruction was issued. Another problem was that a circuit was required for monitoring the length of instructions because each of the instructions had different (2) length. In short, the more features (3) were enforced, the more complicated the processor became. There is no doubt that the more internal (4) components are, the more signals (5) to be transmitted between them, resulting in (6) increase of power consumption as well as heat generation.

(7) What was worse, CISC-based CPUs faced a barrier that their clock speed (8) cannot be higher. This was because the heat resistance (9) of CPU was limited (10) although the heat continued to increase as the clock speed became higher.

  1. This is an incomplete sentence. It should be attached to another sentence or rewritten so it becomes a complete sentence.
  2. This needs to be the plural form.
  3. It's not clear what you mean to say here. Why "enforced"?
  4. This parallel construction is not written correctly – it should be "…the more internal components there are…"
  5. To complete the parallel construction, this should be "…the more signals are transmitted…"
  6. This should have an article in front of this word.
  7. You are saying something is worse, but you haven't really explained that anything was "bad."
  8. A verb tense problem here. If you use the past tense earlier in this sentence, this should be "could not be..."
  9. This needs an article.
  10. The construction of this sentence is a bit odd. What you are trying to say doesn't quite fit into this construction. The sentence should talk about the limit in raising the clock speed being due to the limit in the CPU's ability to withstand heat.

[訳例3]

While this (1) mechanism facilitates processing, it slows down execution because a CPU needs to access the microcode for each complex instruction. Another problem is that a CPU needs a circuit to supervise instructions that have different (2) length. In other words, functional enhancements make processors more complex. More complex architecture naturally requires more signal exchanges, resulting in more power consumption and heat (3) release.

Higher clock speeds (4) release more heat, but a clock can not resist heat above a certain level. This barrier keeps a clock (5) from speeding up.

  1. Is the sentence referring to a "mechanism"? Or maybe a process? I'm not sure about "mechanism." I think the sentence may be talking about a process that facilitates programming.
  2. This should be in the plural form.
  3. It's probably the same thing, but why refer to heat release rather than heat generation?
  4. I think the question here is which comes first; the release of heat or the generation of heat? And which is probably the focus of the sentence? The creation of heat by the higher clock speed or the release of heat that has been generated?
  5. This sounds like it is intentional that the speed is not faster. A different phrase may express this idea better. You want to say something about the heat barrier holding down the clock speed.

[訳例4]

This (1) technique facilitates easy programming, but its downside is that the CPU takes extra time to run as microcode (2) need be accessed each time (3) an instruction is issued. Also, an additional circuit is required to (4) verify the execution of (5) variable-length instructions. This is one example that functional enhancement could result in more complicated electronics, which in turn leads to more power consumption and heat (6) dissipation as a result of an increase in the traffic of internal signals. Another reality is that the faster the clock speed is, the more heat the CPU generates. Because CPU chips are likely to suffer from excess heat, heat (7) dissipation is a major barrier to higher clock (8) frequencies.

  1. See earlier note about this word.
  2. This should probably be "…needs to be accessed…"
  3. Do you think this should clearly state what kind of instruction you mean? Are you talking about all instructions or only complex instructions?
  4. Does this circuit only verify or does it also supervise?

    My understanding is that one of the purposes of using variable-length instructions is to limit the amount of wasted space. Variable-length instructions use special decoding circuits that count the number of bytes of each instruction and frame it (the instruction) according to its byte length. So I see no implication of checking, verifying, supervising, validating, etc., though the author uses 監視 to explain this process.

  5. Do you mean that the circuit verifies or checks only variable-length instructions? Has the article established whether or not there are both uniform and variable-length instructions?

    CISC architecture handles variable-length instructions; RISC uses fixed-length instructions.

  6. I think it may be a good idea to logically decide what order heat is handled. Is heat generated first and then this heat is dissipated? Does this sentence primarily wish to express the generation of heat or the actual dissipation of heat. I think either one will probably work, but logically I feel that it is the generation of heat that causes problems, not the dissipation of it.
  7. Is the dissipation of heat a barrier to higher clock speeds? Does this sound logical?
  8. Is this usually referred to as higher clock speeds or higher clock frequencies?

    Either way. The higher the clock frequency, the faster the clock speed.

[訳例5]

Although (1) this mechanism makes the programming easier, longer execution time is required because (2) for each instruction issued the CPU needs to access a corresponding microcode. Another problem is that an (3) additional circuit is required to monitor each instruction for (4) difference in length, which means that the microprocessor becomes more complex in structure with each functional enhancement. The more complex the structure becomes, the more power consumption and heat generation results because of the increased internal signal occurrences.

Making the clock speed of the CPU faster (5) cannot be a solution to compensate for the longer execution time because faster clock speed generates more heat and CPUs (6) cannot be heat resistant accordingly. The CISC technology is (7) finally running out of steam.

  1. Is this a mechanism? A device? A component? What is it?
  2. What kind of instruction are you referring to? Any kind of instruction or for each complex instruction?
  3. I think the original Japanese only says "circuit" not "additional circuit."
  4. Is the circuit there to look for whether or not there is a difference in length?
  5. I think "is not" fits better here.
  6. Check into the difference between "resisting heat" and "withstanding heat."
  7. Why did you choose "finally" here?

○ 発熱量
heat dissipation、heat release だと放熱、排熱(getting rid of heat)。
熱が発生してから放熱されるという順序を考えると heat generation が妥当。

Heat occurs, then undergoes radiation, conduction or convection before dissipation.
適切なヒートシンクや放熱ファンがないとCPU は極端に遅くなる、または壊れてしまう。

○ 熱量

amount of heat、quantity of heat

amount (「総量」のニュアンス)
1 The total of two or more quantities; the aggregate.
2 A number; a sum.
3 A principal plus its interest, as in a loan.
4 The full effect or meaning; import.
5 Quantity: a great amount of intelligence.

quantity(「測定対象」のニュアンス)
1 a. A specified or indefinite number or amount. b. A considerable amount or number: sells drugs wholesale and in quantity. c. An exact amount or number.
2 The measurable, countable, or comparable property or aspect of a thing.
3 Mathematics Something that serves as the object of an operation.
4 a. Linguistics The relative amount of time needed to pronounce a vowel, consonant, or syllable. b. The duration of a syllable in quantitative verse.
5 Logic The exact character of a proposition in reference to its universality, singularity, or particularity.
yourdictionary.com (http://www.yourdictionary.com/)

○ 問題

問題 is not always a problem.
matter (to mention)
issue (to discuss)
problem (to solve)

話題、課題、問題などがすべて簡単に「問題」とされていることについて、Gregさんに限らず英語ネイティブは首をかしげる。

○ 壁

barrier (to break), deadlock (to break), wall (to break down), obstacle (to overcome)…

[テキスト3]

そこで、RISCではCISCで約300ほど用意されていた命令の数を使用頻度が高い100個前後まで一気に減らし、マイクロコードの代わりにそれらの命令を高速に実行する専用回路を用意した。しかも、すべての命令を一定の長さ(32bit)に統一することで監視回路をなくして、CPUの構造を簡潔にしたため、より高度な半導体技術を使って集積率が上げられるようにもなった。ただし、簡単な命令セットだけに絞られてしまったので命令の最適化が必要になり、コンパイラなどの開発ツールの負荷は大きく増えている。

[訳例1]

To overcome this obstacle, RISC slashed the number of instructions from the approximately 300 used by CISC to around 100 of the most frequently used instructions. And in place of the CISC microcode, RISC has a dedicated circuit to execute those instructions at high speed. With RISC, because all of the instructions are of uniform length (32 bits), a supervisory circuit is no longer necessary, allowing the CPU structure to be simple and streamlined. The simplified design has made it possible to achieve higher degrees of integration of the CPU through advanced semiconductor technology. However, the reduction to solely simple instruction sets demands the optimization of instructions, imposing a greater burden on the compiler and other development tools..

[訳例2]

To overcome these problems caused in CISC, RISC architecture was developed. In the RISC architecture, the number of instructions, which were about 300 in the CISC, was reduced down to about 100 that occurred at a higher frequency (1) among others. Also prepared in the RISC architecture was a dedicated circuit, in place of microcode, for executing these instructions at high speed. In addition, to eliminate the need for the monitoring circuit, the length of the instruction was made to a (2) predetermined length, 32 bits. This simplified the configuration (3) of CPU, and thus it became possible to increase (4) package density by using (5) a much more highly advanced IC technology. However, RISC also has its drawbacks. Since the instructions are limited only to simple instruction sets, (6) the optimization of instruction is required, and thus the load on developing tools such as compilers greatly increased.

  1. This segment is probably not necessary.
  2. I don't think the sentence is talking about "pre-determined" length, but rather "uniform length."
  3. This should be in the plural form or use an article.
  4. I am not an expert, so I don't know if this is the right term, but it sounds a bit odd to me.
  5. This is probably not necessary.
  6. I would probably leave this out.

[訳例3]

To clear this hurdle, (1) engineers selected about (2) 100 most-used instructions out of about 300 instructions prepared for CISC. The engineers also introduced a (3) Microprocessor Supervisory Circuit that can execute these instructions at a high speed instead of microcode. Standardizing instruction length (32 bits), they successfully eliminated the need (4) of a Microprocessor Supervisory Circuit and streamlined (5) a CPU's architecture. As (6) the result, (7) higher packaging densities can be achieved by up-to-date chip-making technology. The (8) drastic reduction in the number of instructions, however, demands optimization of the instructions, which puts a heavy load on development tools including compilers.

  1. Is there a reason that engineers are specifically mentioned? Are engineers the only people who wish to speed up the clock speed? And did the engineers SELECT these instructions?
  2. It might be a good idea to mention that these 100 instructions are for RISC. This is not mentioned in the sentence.
  3. This term is in capital letters. Is this the official name for this unit?
  4. The preposition should probably be "for."
  5. The article should be "the" here.
  6. The article should be "a" here.
  7. Since I am not an expert in this field, I cannot say whether or not this is correct, but is this a recognized term?
  8. By saying "drastic", you are implying that this was not a good thing. Usually, this word is used to describe something that had a negative effect. I think you want to emphasize that the reduction of instructions was a good thing.

[訳例4]

While the CISC architecture contains a set of about 300 instructions, the RISC design reduces the number of instructions to some 100 that are used most often. For faster execution, these RISC instructions are hardwired to dedicated circuits to (1) replace microcode. The CPU structure is simplified with the instruction length fixed (to 32 bits), eliminating the need of a (2) verification circuit. A streamlined CPU design allows for higher integration (3) with further advanced semiconductor technology. However, a simple, small instruction set means an additional burden on the development tools and utilities, including (4) a compiler structured to optimize individual instructions.

  1. I am not too clear on this. Are you saying that RISC instructions replace microcode or do the dedicated circuits replace microcode?

    RISC architecture eliminates the level of microcode so that machine instructions directly control the processor.

  2. See earlier note about the word "verification"
  3. This sentence construction might give the impression that higher integration must wait for further advanced technology.
  4. Is a compiler a development tool or utility? Which of these devices does a compiler belong to? I wonder if it would make for a better sentence if we understood more clearly which category the compiler belonged to.

    In software development, programs called tools and utilities are used. The development tools include compilers, assemblers and debuggers. Utility programs perform very specific tasks, usually related to managing system resources including memory and disk space.

[訳例5]

RISC takes a different approach to overcoming the above described (1) problem. It drastically reduces the number of the instructions from approximately 300 for CISC to around (2) frequently-used 100 instructions. It also provides a dedicated high-speed circuit for performing the selected instructions instead of the (3) legacy microcode. In addition (4) to that, the RISC instructions (5) having the same code length ((6) 32 bits) allows the designers to eliminate the monitoring circuit and make the CPU structure simpler, resulting in higher degree of (7) component integration by more advanced semiconductor manufacturing technologies. A drawback is, however, that the reduced set of instructions requires each instruction to be optimized, placing significantly (8) more burden upon development tools such as a compiler to perform the optimization.

  1. Is this really a problem or just an issue or matter?
  2. Change the word order to “100 frequently-used instructions.”
  3. It's not clear what you mean by "legacy" here.
  4. To what? What does this refer to?
  5. Are there RISC instructions that have varied lengths?
  6. Is there a reason for using "code length"? Is it the same as "instruction length"?
  7. Does this refer to packing higher densities?
  8. I think a better word is "greater burden."

○ 集積率
packing (packaging) density
集積率は「単位面積あたりの素子の数」と書かれているほうがわかりやすい。
integration rate などはコンピュータに親しみのない人にはわかりにくい。経済では統合とか集中の意味で使われている。

40年ほど前にテキサスインスツルメンツやフェアチャイルドが先駆けてコンピュータの第二世代を画したICが集積回路と訳されたから集積率はきっと integration でよかろうと思うわけだが、これはけっこうビッグワードのようだ。有名なMoore の法則ですら、The number of transistors per chip is doubling every 18 months. と平たく説明されている。集積率に限らず、アチラの文明を輸入すると簡単なものでもむやみに難しい日本語になってしまう。たとえば、レコードプレーヤーが電気蓄音機(電蓄)、ハイファイが高忠実度。さすがに、ステレオあたりから翻訳をあきらめたようだ。

○ command、instruction
「マイクロソフト単語帳」によると、

command:
人がコンピュータを操作する目的で、コンピュータに対して入力する指示命令文のこと、
instruction:
コマンドのこと。コンピュータやそのソフトウェア コンポーネントに動作を指示するための命令文のこと、
ということで両者はほぼ同じもののようです。ただし捕捉説明として、
command:
現在のコンピュータでは、あらかじめシステムに組み込まれているプログラム(アプリケーション ソフトやバッチ プログラム、あるいは、スクリプト)を実行するためには、その動作を開始させるための何らかの指示となるもの(つまり、何らかのトリガとなるもの)を始めに入力する必要がある。そのトリガとなるものを、簡単な単語や語句を組み合わせで表現したものを『コマンド』と呼ぶ。ほとんどの場合、コマンドは人にとってその処理の内容が容易に理解できるような単純なキーワード(例えば、copy、delete 等)を組み合わせた構文に基づいて作られている。人は、その定められた構文にしたがって必要なコマンドを組み立ててキーボードから入力したり、あるいは、画面上でクリックしたりしてコンピュータを操作する。
instruction:
CPU の動作を規定するために用いられる 1 バイト~数バイトの長さの符号 (ビットのパターン)のことを『命令』と呼ぶ。そして、CPU が実行する単純な動作を 1つずつ体系的に定義した『命令』の集合(命令一式)のことを、 『命令セット』 と言う。つまり、 CPU は命令セットを構成する命令毎に定義された符号を主記憶から順次取り出して(fetch して) 解釈し、それに対応した動作を次々に実行する装置だと言える。機械語やアセンブリ言語のプログラムは、その命令セットの命令の中から、処理に必要な命令を実行する順番に1つ1つ並べて記述して、CPU の動作を連続した処理手続きとして表現したもの。

また、「コンピュータ用語辞典」では、「commandは、コンピュータの中央処理装置(CPU)の「命令」であるinstructionとほぼ同じ意味をもっているが、入出力チャネル制御装置(input/output channel unit)が入出力操作を実行するための「命令」のことをコマンド(command)と呼び区別している。」としており、区別することもあるようです。ここで、「入出力チャネル」とは、「内部記憶装置と周辺装置の間のデータ転送を扱う機能単位」のことです。

六本木にあった日本IBMのライブラリから20 年ほど前に手に入れた(図書館から手に入れるというとなにか違法行為をはたらいたような感じがするが、購入)IBM Data Processing Glossary (GC 20-1699-4) を原典とする「情報処理用語」という骨董品的語彙集によれば、command language は指令言語で、「実行すべき機能を開始させる働きをもつ手続き的演算子を主構成要素とする原始言語」、computer instruction は計算機命令で、「計算機用の機械語命令」とそっけない。

要するに、コマンドは、ユーザー(オペレータ)がコマンドラインなどを介して直接マシンに与える動作指示、そしてインストラクションは、ユーザー(プログラマ)がプログラムにプログラミングして間接的にマシンに与える動作指示、というふうに理解している。

[テキスト4]

POWER PCの構造

POWER PCは「バスインターフェイスユニット」「分岐処理ユニット」「ロード / ストアユニット」「整数演算ユニット」「浮動小数点演算ユニット」の5つのユニットで構成されている。

[訳例1]

Configuration of Power PC

The Power PC is made up of five units: a bus interface unit, a branch processing unit, a load/store unit, and a floating point unit.

[訳例2]

Configuration of POWER PC

POWER PC comprises five units: Bus Interface Unit, Branch Processing Unit, Load/Store Unit, Integer (1) Arithmetic Logic Unit, and Floating-point Processing Unit.

  1. Is this a computer term? This is probably correct; I am not familiar with it.

[訳例3]

Configuration of PowerPC

PowerPC consists of five units: a bus interface unit, a branch processing unit, a load/ store unit, integer units, and a floating point unit.

[訳例4]

PowerPC Execution Unit

PowerPC contains five execution units: (1) Bus Interface Unit, Branch Processing Unit, Load/Store Unit, Integer Unit, and Floating Point Unit.

  1. Are these trademark names of these components? If not, these can be left in small letters. Capital letters would used if these were actual product names.

[訳例5]

The structure of POWER PC

POWER PC is comprised of five units, a bus interface unit, a branch processing unit, a load/store unit, an integer (1) arithmetic unit, and floating-point arithmetic unit.

  1. I am not an expert, but is "arithmetic" necessary here?

○ 屋内ではインテリア的な利用
a number of は、ふつう 5 ~ 10。

○ Some Asides
一般の人
common people
一般人(名も位もない unsophisticated で rough な感じ)。
ordinary people
(平)凡人で、多少ネガティブな感じ。
average person
普通の人。
general public
一般大衆。
public people は redundant。

現在使用中のパソコンのCPUはCISCのPentium4でクロックが1.69 GHz。自宅にある上位モデルに比べると反応が遅すぎてストレスがたまります。でも、そのうちWindowsに代わるLonghornが出るまで買い替えは賢明でないと思うので、しばらくはOSの再インストールやメモリの増設で凌ぐ覚悟をしており、先日はその件でメーカーの顧客サポートに気軽にフリーダイヤル。

例によって自動音声のアーセーコーセーに言われるとおりにして、そばの小諸蕎麦に大もりを食べに行って戻ってくるとちょうどつながって、間延びしたような気だるいような女性の声で調子が狂いますが、それでも真正面から真剣に話をしているようなので納得しながら我慢して、電話を切るときに「そこはどこですか」と聞いてみたら「(中国の)大連です」というから驚くではありませんか。

専門的なことを尋ねる日本人を相手に日本語をこのように巧みに操って対応する姑娘に感心しながらも、もしたとえば最近増えているオタク的な人がもっと複雑な質問を東北弁でまくしたてたらどうするのだろうなどと、よけいなことを心配したりします。それにひきかえ、なぜかとても平和なこの国の会社では電話を取る女子従業員が「少々お待ちください」を平気で連発してイライラさせられます。なにもかもグローバリゼーションの試練にさらされる時代になり、こんな風習も遠からず消え去るはず。

いずれにせよ、この日本人のパソコンはアメリカのDELLで台湾のメーカーが製造し、中国のヘルプデスクが顧客サービスを行っています。そんな折、IBMがPC事業を中国企業に売却のニュース。Lenovo という会社はIBMが世界中に展開していた拠点もそっくり買い取り一挙にDELL、HPに続く世界第三位。これに勢いを得て中国はハイテク分野でも今後どんどん海外に進出というのが共通の Chinalysis。中国のトラはすぐそこまで来ています。

中国といえば、以前に技術英語研究会のメンバーであったTさんが、あるとき突然中国語を勉強すると言い出して独り西安に旅立ちました。数年経って現地から奥さんまで調達して堂々の帰国を果たし、現在はある大学で非常勤講師。今から考えると、その先見の明はさすがでした。

投稿者 kz : 2005年02月25日 22:49

トラックバック

このエントリーのトラックバックURL:

コメント

コメントしてください




保存しますか?