Talk:Arithmetic logic unit

Suggest 1 possible wiki link and 2 possible backlinks for "arithmetic and logical unit"

This comment has been moved here from Talk:arithmetic and logical unit.

An automated Wikipedia link suggester has some possible wiki link suggestions for the "arithmetic and logical unit" article:

• Can link carry-out: ...es, which can be used to indicate cases such as carry-in or [[Take-out|carry-out]], overflow, or other statuses....

Additionally, there are some other articles which may be able to linked to this one (also known as "backlinks"):

• In Combinatorial logic, can backlink arithmetic logic unit: ...natorial and sequential logic. For example, the part of an [[arithmetic and logical unit|arithmetic logic unit]], or [[ALU]], that does mathematical calculations is made fr...
• In List of 7400 series integrated circuits, can backlink Arithmetic Logic Unit: ...: 9-bit Odd/Even Parity Generator and Checker *74181: 4-bit [[arithmetic and logical unit|Arithmetic Logic Unit]] and Function Generator...

Notes: The article text has not been changed in any way; Some of these suggestions may be wrong, some may be right.
Feedback: I like it, I hate it, Please don't link to. --LinkBot 11:29, 1 Dec 2004 (UTC)

Rationale

Doesn't say why many operations are grouped in an ALU.
AI says it's because up to 70% of circuits are shared between operations, but I find no proper source. Musaran (talk) 13:44, 10 November 2023 (UTC)

In what way are operations grouped? Who is AI? What do you mean when you say operations “share circuits”? Lambtron talk 18:41, 10 November 2023 (UTC)

I think I understand what you mean. Are you asking why an ALU has the ability to perform a variety of operations, and what purpose does that serve? Lambtron talk 01:02, 11 November 2023 (UTC)

First paragraph needs rewording: change “integer binary numbers” to “integer represented as binary numbers”.

Correction required to first paragraph: First paragraph needs rewording: change from “integer binary numbers” to “integers represented as binary numbers”. 2A02:C7E:5A76:8100:F0AC:B684:F79C:4102 (talk) 16:30, 25 March 2024 (UTC)

Ones' complement, two's complement and shift and subtract are unclear and misleading

The existing text referring to Ones' complement and two's complement is unclear. Specifically,

Two's complement: A (or B) is subtracted from zero and the difference appears at Y.

The text is essentially circular, in that it assumes that subtract means a twos' complement subtract.

Increment: A (or B) is increased by one and the resulting value appears at Y.

The text doesn't explain what increased means, i.e., is there an end-around carry?

Decrement: A (or B) is decreased by one and the resulting value appears at Y.

The text doesn't explain what decreased means, i.e., is there an end-around borrow?

Arithmetic shift: the operand is treated as a two's complement integer, meaning that the most significant bit is a "sign" bit and is preserved.

This doesn't address what happens to low order bits in a left shift, and an arithmetic right shift must propogate the sign bit for both ones' complement and two's complement machines.

I'm not sure of the best way to fix this. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 18:38, 3 July 2024 (UTC)

The ALU article seems to assume two's complement arithmetic which does not do end-around carries. Maybe that assumption needs to be stated. On your individual questions:

Two's complement: This is trying to explain the urinary function, not the number system. So to get the two's compliment of a number, A is subtracted from zero with no end-around carries. I removed the (negation) in parenthesis as it did not match the Wikipedia negation article, though some processors have a NEG instruction.

Increment/Decrement: Should probably say "1 is added to/subtracted from A and the resulting value appears on Y"

Arithmetic shift: Both arithmetic shift and logical shift have problems. Maybe they should be defined like this:

Left arithmetic shift, zero is shifted in from right (LSB).

Right arithmetic Shift: Sign bit is duplicated from left (MSB). This is used to shift signed integers.

Left logical shift: same as left arithmetic shift

Right logical shift: a logic zero is shifted into the operand from the left. This is used to shift unsigned integers.

It's important that we don't over-describe these functions in the text. Article is about ALUs. If the reader is curious, she can click on the links. RastaKins (talk) 20:27, 3 July 2024 (UTC)

The vocabulary in your response also assumes 2's complement; "subtracted" means three different things on 1s' complement, 2's complement and sign-magnitude machine. Similarly, an arithmetic left shift behaves differently. What is need is a table showing the operation on one axis and the representation on the other. For example, an arithmetic left shift on a 1s' complement machine is a circular left shif; the right side is filled with sign bits rater than zero bits. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 20:23, 4 July 2024 (UTC)

Intel naming conventions for GPU elements

An unregistered editor has repeatedly added the statement "Some GPU or SoC providers, such as Apple and Intel, use ALU term to equal to the GPU shader, especially unified shader", citing a PDF overview of Intel's Xe architecture. The article lede already says ALUs are used in GPUs, so it's not clear to me why this statement is relevant, other than the fact that Intel happens to include "ALU" in the names of various architectural elements. Also, since there's no mention of Apple in the ref, Apple here appears to be OR. Finally, the statement is not a properly structured sentence and therefore, if deemed worthy of keeping, should be rewritten. I've reverted a couple times now, but the statement keeps popping back up and I'm not looking for an edit war. Am I missing something here? Lambtron talk 23:14, 9 January 2025 (UTC)