Talk:JEDEC memory standards

IEC modification

"Their rationale for including these definitions is explained and contrasted to the IEC standard prefixes: - footnote text"

You have elevated a footnote to a main feature. The footnote appears to quote an IEC document. JEDEC was stating that they heard of the IEC binary standard but were ignoring it. None of the JEDEC standards use the KiB or MiB notation. All of the semiconductor memory producers use the JEDEC terms for memory size. Also the JEDEC standards do not use terms like megabyte and gigabyte so there is not point of confusion

I am removing this WP:POV modification. - SWTPC6800 21:49, 8 April 2007 (UTC)

Where does the JEDEC state that they are ignoring the IEC prefixes? They note them as "an alternative", note that their definitions document common usage, and proceed from there. How you can interpret their not using IEC prefixes as outright rejection of them is beyond my comprehension. I find it ironic that you would counsel others about POV when you've created this article for, as far as I can tell, the sole purpose of making a big fuss about the fact that JEDEC uses the common interpretation of "megabyte" in their standards. No, I think NPOV mandates that we show both sides of the coin if this article is going to continue to exist solely to talk about which two letters the standard uses for capacity notation rather than the contents of the actual standards. -- mattb @ 2007-04-11T04:00Z

You can look at JEDEC documents here and while you will find the IEC binary footnote in Terms and Definitions documents you will not find the in any other standards. [[1]]

You can also search all of the JEDEC documents here. [[2]]

You will not find megabyte (etc) used in any working standard. This does not mean the someday in the future JEDEC may change but the latest standards do not use them. JEDEC has an agreement to work with IEC. -- SWTPC6800 22:12, 11 April 2007 (UTC)

Omission != commission. -- mattb @ 2007-04-11T22:33Z

In the case of documents that comply with a standard then it can be said that using a specifc term means that term complies with the standard. Fnagaton 15:12, 23 April 2007 (UTC)

Why semiconductor companies use JEDEC

JEDEC is the standards organization of choice for semiconductor companies because for the flexibility and speed of the process. JEDEC Standard 21 for Configurations for Solid State Memories is published in a 3-ring binder that can be updated monthly. Each semiconductor company (and other members) gets a vote. In IEC each country gets a vote and the process takes years.

Solid State Times OCTOBER, 1999 Volume 1, Issue 4, Page 2 [3]

JEDEC has entered into an important new Memorandum of Understanding with the International Electrotechnical Commission (IEC) to expedite the process of moving semiconductor standards into the international marketplace. The agreement, which became effective September 1, 1999 is part of the IEC-PAS (Publicly Available Specifications) program. The program’s objective is to speed up the international standards process in areas of rapidly developing technology. It is intended to deliver information to the marketplace quickly since PAS are de facto standards. If PAS standards are approved and accepted by the international community, they ultimately will become IEC international standards.

Under the recently signed agreement, JEDEC will identify the specific standards it wishes to submit for PAS consideration. If the IEC accepts the JEDEC standard as a PAS to be entered into their Program of Work (PW), the standard will be published under the joint logos of JEDEC and the IEC, and royalties on the sale of the PAS will be divided equally between the two organizations. If the IEC later adopts the PAS as an IEC international standard, the standard will be published under the IEC’s sole logo and the division of royalties will end.

A significant feature of the new agreement is the fact that JEDEC will continue to have the right to continue developing the relevant technology, to revise and modify the underlying JEDEC standard, and to continue publishing the underlying JEDEC standard on the World Wide Web.

Solid State Times JANUARY, 2000 Volume 2, Issue 1, Page 10 [4]

The IEC works differently than the JEDEC committees where it is one country one vote instead of the JEDEC one company one vote. This voting practice does not favor the larger Semiconductor Countries over the smaller ones. Many smaller countries like the IEC because they can voice their opinion without being overpowered by the larger countries. The IEC normally meets twice per year to discuss new proposals and to review the present balloting results.

The IEC has 4 basic stages of balloting which include New Work Item Proposal (NP), Committee Draft (CD), Committee Draft for Vote (CDV), and Final Draft International Standard (FDIS). Since the IEC meetings are International and their meeting schedule is every 6-months, the balloting cycle can easily be 2-3 years in length. The IEC management group is working on ways to reduce this cycle time in 1/2 in order to meet the needs for the Semiconductor Market.

SWTPC6800 19:05, 12 April 2007 (UTC)

JEDEC Standard 100B.01 and "tera"

Hi, I searched the 100B.01 but could not find a definition of "tera" in terms of "A multiplier equal to 1,099,511,627,776 (2^40 or K^4, where K = 1024)". In fact, 100B.01 seems to only define kilo, mega and giga in a binary meaning. It uses tera in the sense of 10^12 in the table but nowhere else. Thus, I deleted the corresponding line. If I were wrong and there is other evidence, it might be reinsertet. --GlaMax (talk) 09:53, 1 October 2009 (UTC)

Avoiding WP:UNDUE

Regarding this change. If the minor point about quoting parts of a footnote (out of context I might add) is relevant enough to be included in the article then to avoid giving WP:UNDUE weight to to those sections it is necessary to clarify exactly what is included in the standard. Without the clarification that section of the article gives too much undue weight to a footnote (that is quoted out of context). Fnagaton 06:59, 29 January 2010 (UTC)

Nothing is quoted out of context, and the relevant part is reproduced in its entirety. Do you want to reproduce the table of IEC prefixes as well that the specification endorses to be used as an alternative by also referring to their original place of definition? Stop injecting your POINT OF VIEW and stick to what the specification clearly says. The quotation is not a footnote, but labeled as a NOTE in the original, which is the term used for all explanations to the definitions in the document. You are trying to push an argument that the JEDEC specification does not permit or sanction the use of binary prefixes and requires the traditional ones, which is clearly wrong based on the very document in question. Kbrose (talk) 17:07, 29 January 2010 (UTC)

It is about making the article neutral and balanced. The way the article reads without my changes makes it appear to advance a certain position that is different to the real world position. The other more recent JEDEC standard linked in the article does not mention IEC prefixes at all when being used for memory, remember that fact. Fnagaton 00:01, 30 January 2010 (UTC)

The scores of current JEDEC memory standards use KB, MB and GB to specify a semiconductor memory's capacity. JESD79-3D, the DDR3 SDRAM Standard, revised in September 2009, list memory sizes of 512MB, 1GB, 2GB, 4GB and 8GB. JEDEC Standard 100B.01, updated in 2002, has a footnote that mentions the IEC has proposed an "alternative system" for specifying memory size. In the decade after the IEC "alternative system" was introduced, it has absolutely failed to gain acceptance in the computer industry. Any suggestion that the IEC prefixes are used in the real world is undue weight. -- SWTPC6800 (talk) 04:21, 30 January 2010 (UTC)

Call for editors' consensus of the accurate reflection of JEDEC Standard 100B.01

The content of THIS VERSION of the article in section JEDEC Standard 100B.01 reflects properly this original source document for the status of the kilo, mega, and giga prefixes and the role of the binary prefixes. Kbrose (talk) 17:07, 29 January 2010 (UTC)

• Agree. Kbrose (talk) 17:07, 29 January 2010 (UTC)
• Disagree. The article needs to be more neutral and not advance one minority view over the actual real world situation.Fnagaton 00:13, 30 January 2010 (UTC)