Talk:AMD K8L

From Wikipedia, the free encyclopedia

↓    Skip to table of contents    ↓

This is the talk page for discussing improvements to the AMD K8L article.
Please sign and date your posts by typing four tildes (~~~~). Put new text under old text. Click here to start a new topic. New to Wikipedia? Welcome! Ask questions, get answers. This is not a forum for general discussion about the article's subject.	Be polite Assume good faith No personal attacks Be welcoming	Article policies No original research Neutral point of view Verifiability

To-do list for AMD K8L:	edit · history · watch · refresh
Here are some tasks you can do: Cleanup: General clean-up and expansion. Other: Incorporate recent news regarding Barcelona

Archives

Archive 1 - 2006

Contents 1 Wording 2 Sources 3 Oops, there goes the K8L codename 4 Requested move 4.1 Survey 4.1.1 Survey - in support of the move 4.1.2 Survey - in opposition to the move 4.2 Discussion 5 Merge from Agena (processor), Kuma (processor), Rana (processor) 6 Page out of date 7 Barcelona 7.1 Opteron (65 nm SOI) 8 Real codename of the "AMD Next Generation Processor Technology" is really K10. 9 Montreal 10 Montreal or April fish ? 11 Report 12 Requested move 12.1 Survey 12.2 Discussion

[edit] Wording

Much of this article seems to be full of marketspeak (which is often misleading and/or nonsensical/non-objective in nature). An addition to the simple changes I made earlier, I've changed:

which will provide better opportunity for diversification at the platform level.

to

making differences in design between different platforms easier.

I believe that's what the writer was intending to say. If the original contributor believes it different, please correct it. --Tene 11:51, 25 January 2007 (UTC)

[edit] Sources

Err, we are using the inq as a factual news source?

More truth comes from a politician than from the inq.

[edit] Oops, there goes the K8L codename

http://theinquirer.net/default.aspx?article=37444

Let's forget what the Inq report is factually correct or not at this moment...

In this report, it reads

“	While Intel and the rest of the industry was using the codename K8L for AMD's next-gen architecture revamp, K8L existed only in AMD's expired internal roadmaps a long, long time ago. The "L" suffix in K8L actually meant Low-Power, a description for the Turion 64, which was introduced years ago. K9 was actually a codename for the dual-core family, only introduced after Turion's debut in Las Vegas.	”
	— Giuseppe Amato, technical director of marketing and sales for the EMEA region, interview report

And What the heck is going on here!!?

P.S. move tag anyone? --202.71.240.18 10:33, 6 February 2007 (UTC)

K9 was the AMD X2 dual core processors. Barcelona is of the K10 design. Selectodude 03:00, 15 April 2007 (UTC)

[edit] Requested move

[edit] Merge from Agena (processor), Kuma (processor), Rana (processor)

Please merge any relevant content from Agena (processor), Kuma (processor), and Rana (processor) per Wikipedia:Articles for deletion/Agena (processor). Thanks. —Quarl ^(talk) 2007-02-11 04:55Z

[edit] Page out of date

First, according to Tom Yager at InfoWorld, Torrenza is code name for the Next Gen platform, not K8L. Further, Barcelona is said to be the first chip from this intiative. [1]. K8L was Turion 64, as already mentioned in the article lead. Assuming this is true, most of the article needs transplanted to Torrenza since K8L is likely to have been the Turion 64.

Further, saying that we're going to stick to calling K8L and Barcelona the same thing, there is a lot of current information that exists about the first chip coming soon, Barcelona, but is not in this article. According to reports, it's to be 40% than Cloverton, and will be released mid-2007 [2]. Further, it's supposed to have 128 bit wide SSE, be 80% faster in floating point over Opteron, offer new VM and power management techniques. Dedicated L2 cache per core, and a L3 cache [3]. So, again, this is the biggest revamp since 2003. The first L3, and 65nm process for AMD [4]. Nja247 (talk • contribs) 12:37, 11 February 2007 (UTC)

Torrenza is not the name for the next-gen microarchitecture. As noted at Torrenza, it is a socket-compatibility initiative. Infoworld has it wrong. The name "K8L" has never been officially acknowledged, and is probably not the actual codename, but sorting out what the actual codename is for the next-gen microarchitecture has been problematic. There have been suggestions that the codename might just be Barcelona. Official AMD literature has usually just said "Revision H". Also, please note that the performance comparison to Clovertown originates with AMD, and cannot yet be independently verified. — Aluvus t/c 19:54, 11 February 2007 (UTC)

[edit] Barcelona

Why no information on this up and coming product? There's a lot of info about it on the web. 82.45.240.51 17:54, 18 February 2007 (UTC)

Yeah right, "Barcelona is the upcoming server chip, features 65 nm SOI process, with the highest clock speed for 2-way or above is 2.3 GHz, the chip implements HT version 2.0 at 1.0 GHz, providing 8 GB/s bandwidth, supports registered DDR2 DIMM, and is the first batch of chips under the microarchitecture...", okay now, do you think this is enough?
BTW, information of the "Barcelona" chip will eventually put up in this article or even the Opteron article, unless it is "notable" enough to be starting a new article, the information of "Barcelona" and the infomation of this page will ultimately be merged into some articles (e.g. Athlon 64 X2 for dual-core desktop "mainstream" parts, Opteron for server parts, and Athlon 64 FX for "enthusiasts" Quad FX platform parts) and/or split into new articles (e.g. the quad-core desktop parts are rumoured to be named as Athlon 64 X4), so the information on this page is just temporary, and will not stay here forever! Also as the {{ future products }} tag said, nothing about the product is clear yet, and *everything* may be subjected to change at the every end moment, and as you say, "There's a lot of info about it on the web", there're also speculations and may not be the latest and up-to-date information about the final product you mentioned (e.g. Radeon R600 - I know it's me to pump everything the'Inq says into the article, but I just cannot resist to pump those speculations in, please BAN me forever!!), and yeah, then it's time for edits and reverts, edit wars, and semi-protect perhaps? It just cannot help anything by putting the "specs" of "Barcelona" here... well, if you think the specs of "Barcelona" is as important and notable than the microarchitecture itself, then I think the specs should be written like the ones in Athlon 64 FX models section in the K8 article. --202.71.240.18 07:51, 24 February 2007 (UTC)

Who died and left you King? There is verifiable information available on the branding and models available at launch. The "Crystal Ball" argument does not apply. Observ 23:54, 17 March 2007 (UTC)

Oh, yeah!
"Bacelona is the name of microarchitecture!!" AND
"Bacelona is the brand name of the product!"

Do tell me what is the importance of the "Bacelona" CPU you keep yelling about? "The FIRST CPU implemented the micro-architecture?" What about the "Agena"? What about the other "Stars" codenamed Desktop CPU products coming in the near future? and what about the brand name "Athlon 64"? Do they have less importance than the "Bacelona" SERVER CPU itself? The Agena CPU has even more new technologies, such as HyperTransport 3.0, why is that less importance than Bacelona? Thus, why insist on putting only the information about Bacelona in this article?

If you do put the information about "Bacelona" Server products, please as well put the Desktop parts and the future products (aka "Shanghai" and Desktop parts as "Agena"), AND do not use a single paragraph (or a table) to "briefly" describe the product you're yepping about. The bolded paragraph of my last reply is an example of what I mean a worthless paragraph. For your reading pleasure, I quoted that thing here, it goes like:-

"Barcelona is the upcoming server chip, features 65 nm Sillcon-on-Insulator (SOI) process, with the highest clock speed for 2-way or above is 2.3 GHz, the chip implements HyperTransport (HT) version 2.0 at 1.0 GHz, providing 8 GB/s bandwidth, supports registered DDR2 DIMM, and is the first batch of chips under the microarchitecture and features several SSE improvements..."
And then the repeat of the architectrural improvements. Is that important?

I cannot see anything important brought by "Bacelona" itself as IT LACKS IMPORTANCE by itself or as a whole, so as the codenames, what for starting the pages with the codenames and copy what this page contains?? (e.g. Agena (processor), Agena FX (processor), Kuma (processor), Rana (processor) etc) and that idea to add information of Bacelona is almost the same as starting a page for each of the codenames of the procesors.

I do not care if the sources are verifiable, cause it doesn't matter to me (and {{ cleanup }} templates as well as the AfD procedures per WP:xxx will take care of the not verifiable sources).

And if you want Wiki to become a product database, fine, I won't stop you. I just want to tell you that until this time, this article is about the microarchitecture, not the CPUs. If you want to pump in information about server/desktop CPUs (including the codeanme, clock frequencies and stepping etc.), please add the information in seperate article(s) named under the final branding of the CPUs as in Athlon 64 X4 and Opteron, not in this stupid page! Daamit.--202.71.240.18

P.S.And for your assurance, the information of the microarchitecture will be merged/moved later to those pages, they will not be left alone in this "temporary" article. And oh! I've found a right place for those tables, List of AMD Athlon 64 microprocessors and List of AMD Opteron microprocessors, here you go! :)
P.S.2: I was suggesting a style for others to follow, cause I personally do not like one article having one style and the other have others. The style for information of processors in Athlon 64 and Athlon 64 X2 and Opteron are unified in style, and is good, and I cannot see a reason that only AMD K8L can have an exception on style of how to present those information. and BTW, the information you're talking is limited to clock-speed, cache size and width, as well as the voltages, and steepings, sounds similar to the info in those pages, example see below. Others are the improvements of the architecture and is repeated in the sections in this article, so why bother to copy those again in the same page and why do not put them in the lists I mentioned above or follow the style?

---

Appendix - the example:

[edit] Opteron (65 nm SOI)

Quad-core — Bacelona (22yy, 82yy)

• CPU-Steppings: ?
• L1-Cache: 64 + 64 KiB (Data + Instructions) per core
• L2-Cache: 512 KiB, fullspeed per core
• L3-Cache: 2048 KiB, shared
• MMX, Extended 3DNow!, SSE, SSE2, SSE3, AMD64, SSSE3/SSE4a?
• Socket F+, ? GHz HyperTransport 3.0
• Registered DDR2 SDRAM required, ECC possible
• VCore: ? V
• First Release: Q1 2007
• Clockrate: 2100 - 2500 MHz

[edit] Real codename of the "AMD Next Generation Processor Technology" is really K10.

Arllt3 05:59, 30 March 2007 (UTC)

Giuseppe Amato confirmed that in an video, as i say in the article. He already said this in an news of the Inquirer, mentionned in the article. We took this with precautions, because it was precisely from the Inquirer. And everybody knows the legendary reliability of the Inq...

But now, no doubt is allowed. "AMD Next Generation Processor Technology" is the K10. So, i think the article sould be largely modified by someone more experimented that me in Wikipedia. To modify both K8L and K10 articles. —The preceding unsigned comment was added by Arllt3 (talk • contribs) 05:57, 30 March 2007 (UTC).

Yeah, yeah! Nobody cares, and nobody supports a move (above), and blah blah... but do redirect "AMD K10" to "Athlon 64 X4" for god's sake! --202.71.240.18 07:24, 1 April 2007 (UTC)

[edit] Montreal

“	If you are AMD, you put two Shanghais on an MCM and get to eight cores. How? HT3.	”
	— Charlie Demerjian, The Inquirer report

Ouch! Isn't that a kind of K11? Go! Start a new page for that!! DAAMIT. --202.71.240.18 11:05, 3 April 2007 (UTC)

Update, the Shanghai is AMD K10.5!! Reference:http://www.fudzilla.com/index.php?option=com_content&task=view&id=496&Itemid=35 --202.71.240.18 10:31, 13 April 2007 (UTC)

[edit] Montreal or April fish ?

Arllt3 20:51, 5 April 2007 (UTC) Two things: It comes form the Inq, and it is dated from first April. An April fish ?

Who cares if it real or it's just "April Fish"? ;) --202.40.137.202 03:24, 10 April 2007 (UTC)

oops, this is a follow up, guess the Montreal is better, as http://www.fudzilla.com/index.php?option=com_content&task=view&id=480&Itemid=1, Shanghai is an MCM 8-core processor on 45 nm with more L3 cache. OMG --202.71.240.18 10:39, 11 April 2007 (UTC)

[edit] Report

http://www.fudzilla.com/index.php?option=com_content&task=view&id=456&Itemid=1

Nice going, AMD!

Live demo of AMD K8L/K10 + R600 + RD790 = cute!

Hope can give you more cash in Q3. :p --202.40.137.202 03:28, 10 April 2007 (UTC)

[edit] Requested move

The consensus is for a move to AMD K10. However, that article already exists and covers a similar, if not identical, topic. The two articles should be merged. --Stemonitis 07:13, 19 April 2007 (UTC)