Print 43 comment(s) - last by KristopherKubi.. on Dec 20 at 3:36 PM

A new Intel roadmap claims the company will miss its January 20, 2008 launch cycle -- Taiwanese media points the finger at conspiracy

DigiTimes ran an article earlier this week detailing that Intel will miss its January 20 launch date for all three of its 45nm quad-core desktop processors.

Monica Chen claims, "Intel has already notified its partners that it will push back the launch of the three CPUs to February or March next year, depending on AMD's schedule for triple-core and the upcoming Phenom CPUs."

AMD's newest corporate roadmap released last week claims its tri-core Phenom processors, codenamed Toliman, will launch at 2.3 GHz and 2.4 GHz frequencies in "late Q1 2008."

These Intel processors -- dubbed the Core 2 quad Q9550, Q9450 and Q9300 -- are expected to be Intel's first mainstream 45nm desktop offerings. The company launched its ultra-enthusiast QX9650 3.0 GHz quad-core processor on the November 11, 2007.   Between September 2007 and December 2007, Intel changed the launch date of these processors from "early" January to January 20, 2008.

Chen also adds, "Launching the CPUs now will not benefit Intel much in its battle with AMD, while they could cause damage to Intel's 65nm quad-core CPUs, therefore the company has decided it is in no rush to release new products until AMD is able to present more of a threat."

Intel insiders, speaking off the record, would not comment on the specific January 20th launch parts.  One employee added, "It’s a gigantic stretch and almost humorous to think [Intel] would delay products based on a rumored ongoing delays of a competitive product line announced just six days ago."

January 20th, 2008, would also mark the launch of Intel's dual-core 45nm offerings.  These processors, the E8190, E8200, E8400 and E8500 sell in bulk quantities at $163 to $266; with core frequencies ranging from 2.66 GHz to 3.16 GHz. 

Intel's 45nm mobile processors are confirmed for a January 6, 2008, launch.  Several notebook manufacturers tell DailyTech they already received processors shipments for volume.

These five mobile offerings will range from 2.1 GHz to 2.8 GHz, though Intel guidance suggests these are only dual-core offerings.  Quad-core mobile processors are scheduled on the company roadmap, though not until the Montevina Centrino refresh in late 2008.

January 6, 2008, also marks the launch of the Consumer Electronics Show in Las Vegas.  Dell, HP and Gateway already notified media about embargoed Centrino products set to debut on that same day.

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

By Mitch101 on 12/19/2007 4:01:27 PM , Rating: 2
The name calling tin foil hat comment serves what purpose?

Its a very rare bug that will be fixed in only a few weeks time. Its no big deal that is being blown out of proportion. Were not sure it will effect any known application. It will probably stay off the errata sheet unless its discovered in the wild. I take it you weren't around for the P60 issue?

If you noted Intel wont say the release date they instead now quote the chip will be out in Q1 as expected. 6 weeks past Jan 20th is still Q1. They will not confirm or deny the bug. Why do this if there is no bug? We believe Intel is still looking into what applications may be effected b the bug. Not much is known.

Intel would be quick to squash the rumor if it weren't true.

Do you honestly think Intel cares about what AMD does to delay their own products?

How much profit do you think AMD would make with just 6 weeks from Intel's generosity. Laughable.

By JackBeQuick on 12/19/2007 4:12:39 PM , Rating: 2
I was documenting the f00f bug at my company before most people in this forum were in high school.

"They will not confirm or deny the bug."

See this is what gets me annoyed. You want to believe there is a bug, I understand. The errata has been published for months, and the real integrators (HP, Dell, whathave you) would have leaked it all over to the media if they could find a corner case. How do you think the TLB bug made it public?

"Intel would be quick to squash the rumor if it weren't true."

Not likely. NASA doesn't grant interviews to "journalists" looking to confirm if they landed on the moon either.

All I'm saying is that it totally sounds like something Fudo or Charlie or Rahul got "tipped off" to and did zero research to see if it was true. My money is on that it isn't true.

By Mitch101 on 12/19/07, Rating: 0
By KristopherKubicki on 12/19/2007 6:07:03 PM , Rating: 3
Here's what Intel sent me regarding this:

Not saying anything beyond these parts are on track for Q1 like we've always said. Errata are a part of life with CPUs and we are not confirming or denying potential undisclosed errata. All errata are disclosed via the normal eratta process...

If you've got something a little more concrete I can try to get a response

By Mitch101 on 12/20/2007 10:51:12 AM , Rating: 2
You have basically the same information. We don't have a denial of a bug and strangely they even tell you what it is "Errata are a part of life with CPU's". Its only something that will cause the Jan 20th date to be pushed back. Still Q1 but not Jan 20th unless its a paper launch.

We didn't think it was big news just a minor setback. I know I am looking for a quad Yorkfield personally and for work we need them for VM systems. Demand alone we think it would take 6 weeks for the chips to catch up anyhow.

I think everyone was taking to to the level like AMD's recent issue and we don't believe this is the case. If it was then it would be big news. 6 weeks is minor especially since AMD doesnt have a competing product nor do we believe Phenom will overclock enough to compete at that level.

We did have one hit that it might be something when you combine an X48 chipset and the Yorkfield core but getting someone to come forward well good luck. We did a search and their are some indications that memory performance decreases from he X38 to the X48 to support the possibility. DDR1600 mhz and Yorkfield? Only possible nibbles nothing concrete so far.

By Phynaz on 12/19/2007 4:12:40 PM , Rating: 2
You can't fix a cpu bug in "a few weeks time". It takes months.

Plese show one - just one - reliable source that has picked up this news of a bug delaying Penryn.

Nah, didn't think you could.

By Mitch101 on 12/19/2007 4:16:36 PM , Rating: 2
Yes it can be resolved in weeks not months its called Micro Code or uCode. Apparently they don't need a silicon respin only to test the update properly but again no one is exactly sure what the bug is.

By Phynaz on 12/19/2007 4:40:31 PM , Rating: 2
So nobody knows what the bug is, but you know how it's going to be fixed.

And you still can't provide a reputable link.

Does that about sum it up?

By Mitch101 on 12/19/2007 5:06:09 PM , Rating: 1
For the final arguments sake I will say that we don't know what the bug is but we know one exists that will cause the chips to be delayed a short period of time as of now. Most probably for testing to confirm the fix. Who knows it could turn into something else. As of now its only a few weeks. Its not official yet and only Intel can make that statement.

Not sure what the big deal is about over 6 weeks time period. Bugs happen all the time. This one might be a little more and if it is everyone all over the web will be posting the information. 6 weeks wont make any difference to either Intel nor AMD's bottom line sales numbers.

Can you provide a reputable link on the web to prove it doesn't exist? Have any official Intel statement to the contrary?

The 20th will be here soon enough let it go.

By JackBeQuick on 12/19/2007 5:08:29 PM , Rating: 2
Can you provide a reputable link on the web to prove it doesn't exist? Have any official Intel statement to the contrary?

I'm Paul Otelini. Do you have any official Intel statement or repubtable link on the web to prove I'm not?

By Mitch101 on 12/19/2007 5:25:12 PM , Rating: 1
That is funny I might as well be Hoctor Ruiz since were role playing.

The main reason for making 45nm chips is to increase profits. The moment they begin making 45nm parts is the moment their 65nm parts are not as profitable an item. To say that they wont release quad cores because of a lack of competition is laughable especially because consumers don't pay the premium like those who are looking to build servers in their infrastructure.

Its like saying Intel isn't interested in maximizing profits or serving the best most profitable products to the server community especially when budgets are usually renewed at the beginning of the year for most IT departments.

If you feel the need to get in the last word go ahead I am moving on. If we get more information I will post it.

By DallasTexas on 12/19/2007 5:33:01 PM , Rating: 2
Welcome to the forum.

By Samus on 12/20/2007 4:58:41 AM , Rating: 2
As far as I know, microcode updates have patched all cpu errata for years, except for the P60 floating point bug and the PIII-1.13GHz cache overflow bug. Both errata were squashed with a new stepping, which did take a few months.

"It looks like the iPhone 4 might be their Vista, and I'm okay with that." -- Microsoft COO Kevin Turner
Related Articles

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki