Sign in with
Sign up | Sign in
Your question

New Intel 45nm Quads Delayed .. Errata Prone ??

Last response: in CPUs
Share
a b à CPUs
January 21, 2008 11:13:46 AM

http://www.dailytech.com/Intel+Explains+45nm+Delays+Err...

Lets talk about why the new 45nm Intel quads are really being delayed ... or is that FUD??

There is plenty there in the thread to discuss ... any ideas?

Sounds like the non-EE parts are the problem.



January 21, 2008 11:19:46 AM

I like how reactionary you are.

As far as I remember the problem is with P35 boards. The higher end X38s with more layers for the PCB don't have an issue.

But.. here I'll give you what you are looking for.

Intel sucks and is horrible and AMD should be the best selling processor in the world. AMD is always on top of things and is the leader in innovation with the IMC and Native Quad Core Design. They are also on top of the high end server market because the FSB is crap.

Edit: Btw, atleast Intel is fixing the bug and not releasing the processor. What's worse? Releasing it knowing there is an issue and trying to cover it up or just not releasing it while trying to fix the problem?
January 21, 2008 11:22:22 AM

cnumartyr said:
I like how reactionary you are.

As far as I remember the problem is with P35 boards. The higher end X38s with more layers for the PCB don't have an issue.

But.. here I'll give you what you are looking for.

Intel sucks and is horrible and AMD should be the best selling processor in the world. AMD is always on top of things and is the leader in innovation with the IMC and Native Quad Core Design. They are also on top of the high end server market because the FSB is crap.

Edit: Btw, atleast Intel is fixing the bug and not releasing the processor. What's worse? Releasing it knowing there is an issue and trying to cover it up or just not releasing it while trying to fix the problem?



Don't forget that non native quad AKA. double cheeseburger sucks...
Related resources
January 21, 2008 11:26:29 AM

Don't forget that Intel is Evil for Selling Chips to Expensive while at the same time pricing their Chips too low to try and Drive AMD out of business.

Intel needs to raise prices to a lower level so AMD is not driven out of business while people can afford to buy the cheaper chips.

(Note: I'm not sure I fully understand the issue, but I believe this is the issue I have frequently heard reported........)
January 21, 2008 11:26:55 AM

I have a hard time believing an Intel employee would use the word "showstopper".

Interesting (but old) information.
January 21, 2008 11:28:41 AM

zenmaster said:
Don't forget that Intel is Evil for Selling Chips to Expensive while at the same time pricing their Chips too low to try and Drive AMD out of business.

Intel needs to raise prices to a lower level so AMD is not driven out of business while people can afford to buy the cheaper chips.

(Note: I'm not sure I fully understand the issue, but I believe this is the issue I have frequently heard reported........)


I said a long time ago Intel should outsource it's pricing and roadmaps to AMD.

AMD should just be buying up all the Intel 45nm they can and reselling them. Even at a loss they will be in a better fiscal situation in 2008.
a b à CPUs
January 21, 2008 11:28:54 AM

Hey ... you stole my line !!!

My Q6600 cometh soon.

Plus my Firewall server (P3) has just racked up 100 days up-time.

I like a balances worksheet ... just getting a bit sick of the pro-Intel views ... so this seemed interesting enough to post.

All hail thunderman !!!



January 21, 2008 11:30:26 AM

Reynod said:
Hey ... you stole my line !!!

My Q6600 cometh soon.

Plus my Firewall server (P3) has just racked up 100 days up-time.

I like a balances worksheet ... just getting a bit sick of the pro-Intel views ... so this seemed interesting enough to post.

All hail thunderman !!!


When/If AMD comes around and offers better performance than those of us (like myself) who follow performance will be labeled as AMD fanboys. Don't worry.
a b à CPUs
January 21, 2008 11:49:27 AM

Thanks TC ... any comments on the FSB noise ??

Do you think Intel will try to blame the mobo makers ??

This story is starting to sound familiar .... LOL.

Maybe they can stick the RAM on the opposite side of the mobo right next to the cpu to reduce the circuit path??

Then they can stick fans on them ...

Or maybe insulate the traces ... LOL.

Doesn't say much for proposed increase in memory bandwidth for Servers ...

Guess that's why they stuffed soooo much cache on them new chips ??





January 21, 2008 7:33:24 PM

Whatever the issue is seems to have stopped the release of the X48 chipset. I has yet to launch and if memory serves me right it was supposed to launch last year before the holidays.
a c 126 à CPUs
January 21, 2008 8:32:46 PM

Reynod said:
Thanks TC ... any comments on the FSB noise ??

Do you think Intel will try to blame the mobo makers ??

This story is starting to sound familiar .... LOL.

Maybe they can stick the RAM on the opposite side of the mobo right next to the cpu to reduce the circuit path??

Then they can stick fans on them ...

Or maybe insulate the traces ... LOL.

Doesn't say much for proposed increase in memory bandwidth for Servers ...

Guess that's why they stuffed soooo much cache on them new chips ??


They already have memory with fans. Its called the Corsiar Dominator X series. Kinda nice knowing your Ram will get some cool air.

I am glad Intel was willing to lose some sales over trying to make sure the chips are free of bad bugs. And the only reason you see pro-Intel right now is obvious. Intel is doing great now. They release great chips at good prices all around. The Q6600 is a steal in terms of 2 years ago price for a equivalent CPU only at single core. The same would go if AMD's Phenom lived up to its 40% better name. Then it would be pro-AMD.

Think as to why Intel stopped them before users got them. To fix them. AMD knew about the TLB bug which may not affect users but hell in the server area it will really hit them hard since they are high load and use a lot of cache. And they didn't stop them until after and even let people buy them and gave them a patch that hits performance. Yay.

You are right. Adding the extra cache is a way to help alleviate the FSB and RAM usage. But that will change once Nehalem comes out at the end of this year. Then it will be a equal fighting ground all around and whoever performs better will be the top company. Thats if AMD has anything out by then.
January 21, 2008 8:57:32 PM

cnumartyr said:
I like how reactionary you are.

As far as I remember the problem is with P35 boards. The higher end X38s with more layers for the PCB don't have an issue.

But.. here I'll give you what you are looking for.

Intel sucks and is horrible and AMD should be the best selling processor in the world. AMD is always on top of things and is the leader in innovation with the IMC and Native Quad Core Design. They are also on top of the high end server market because the FSB is crap.

Edit: Btw, atleast Intel is fixing the bug and not releasing the processor. What's worse? Releasing it knowing there is an issue and trying to cover it up or just not releasing it while trying to fix the problem?



Intel can take their time with the 45nm because Intel's 65nm is wooping AMD's rear, I don't see much reason to bring Intel's 45nm processors out now anyway. I think its a slap in the face to all us computer users releasing a processor with a bug problem, I would never buy from a company that pulled a stunt like this. AMD got their money from us so why should that bother them. :ouch: 
a b à CPUs
January 22, 2008 6:34:48 AM

FDiv sound familiar ... ??

January 22, 2008 12:15:07 PM

Reynod said:
FDiv sound familiar ... ??


That was a logic calculation error that rarely came up.

Word, Playa.
January 22, 2008 1:45:29 PM

It was the GLUE!!! The GLUE should keep that bug at bay!!!

Where did I put that bottle... I had anti something.. showstopper?

:oops: 
January 23, 2008 12:46:31 AM

Reynod said:
FDiv sound familiar ... ??


Actually it was a lookup table issue. To speed up basic calculations they have look up tables in the processor. unfortunately the table had wrong values in them. That mistake cost Intel over $500 million from what I heard.
a b à CPUs
January 23, 2008 5:01:03 AM

The current errata isn't likely to cost them money because no-one listens to murmurs from the masses.

The stock market crash is shaving more from their net worth than AMD tho ... Perception is everything when your on top !

When your on the bottom tho ...

Shaving 20% off nothing is ... er ... <checks lookup table > 1.2 ??

Woops ... better run that calculation on my Opteron because it is MISSION CRITICAL !!


!