Intel to announce shift in microprocessor technology

Associated Press | August 12, 2005

[fantom]

this looks good. we need to keep moving forward. but for goodness sake , how about we don't sell it to our enemies at least until the next upgrade.

[Boundless]

The article mentions AMD's higher IPC, but fails to
mention if Intel expects improvements on that front.

The shift to the Pentium-M core has been predicted by
CPU enthusiasts for what, over a year now? Is there
any real news here?

> "AMD will face tougher competition once Intel moves
> to the new architecture," Brookwood said.

Sure, if they stand still.

[randog]

But Netburst hit a roadblock last year as jumps in frequency failed to produce the expected improvements in overall performance. In addition, the chips required more power and thus generated considerably more heat.

Oh, good--we won't have to buy bottles of liquid nitrogen to cool our PCs.

Looks like AMD made some smart moves and now leads Intel in the proc race. Good for them. I put an Athlon in Mrs. randog's 'puter last year and we're both pleased with the performance.

[CharlesWayneCT]

If they could only switch to the correct byte ordering.

[NoCmpromiz]

Oh, good--we won't have to buy bottles of liquid nitrogen to cool our PCs.

Did you know that if you pour liquid nitrogen on an open text book, and then turn the page, the page cracks off, sort of like breaking a cracker? (Don't ask how I know..)

Not a single Intel chip in any of the computers in this house, all AMD, thank you very much....

[anonymous_user]

Mac ping?

[konaice]

In the past, Intel launched architectures on chips for desktop PCs and then carried the technology to other platforms, such as servers and notebooks.

This statement is basically untrue. The 486, Pentium, Pentium II, Pentium III, and P4 as well as the xeon were ALL initially targeted at the server market, with Intel's announcements and initial advertising in each case virtually poo-pooing the need for these chips on the desktop.

Of course the market thought otherwise, and Intel was glad to take the money.

Only the Pentium (4) M and Celerons were originally clearly intended to go on the desktop machines.

[AZLiberty]

Whew! I was worried that, following Apple's switch to Intel, Intel would switch to PowerPCs.

[konaice]

In the past, Intel launched architectures on chips for desktop PCs and then carried the technology to other platforms, such as servers and notebooks.

Oh gawd, if they did that half the worlds software would need a rewrite, not to mention half the internet standards.

I'm afraid that bone headed move is going to have a LONG legacy, and we will be fighting that for decades.

[Swordmaker]

anonymous_user asked:

Mac Ping?

About this article reporting Intel Switching their processor architecture after Apple switched their Mac architecture to Intel ... and I thought, "what the heck"... so PING!

If you want on or off the Mac Ping List, Freepmail me.

[anonymous_user]

I was worried ... Intel would switch to PowerPCs.

Microsoft did. ;)

[Moonman62]

It also has roots in the Pentium III processor that Intel launched in 1999.

I've stuck with the PIII because it runs cool, even without a fan. There's no need to use one of the newer waffle irons unless you play games.

[SunkenCiv]

It's early. Steve Jobs could decide to stick with the PowerPC family. ;')

[supercat]

If they could only switch to the correct byte ordering.

Although in most situations nowadays there aren't really advantages to either byte ordering, there are some advantages to little-endian when it comes to machines' ability to read and process data efficiently, and no advantages at all that I'm aware of to big-endian except for making hex dumps look nice.

BTW, I wonder if the decision to use mod-65535 checksums in IP packet standards has anything to do with the fact that they are orthogonal to byte ordering (meaning that byte-swapping all the data in a packet will byte-swap the checksum).

[Lauretij2]

early Athlons forever ruined me for AMD. They'd burn to a crisp in seconds if the heat sink wasn't exactly right. The first notice of trouble was that plastic burning smell. The core of the CPU literally liquified and poured out all over the place.

No thanks, Intel for me. Even if that was one AMD line that was bad.. 'Fool me once, you can't fool me again'.

[MediaMole]

The current AMD chips are great, but I will admit that the Intel chips are more tolerant of user idiocy.

When I got my first PIII chip, I couldn't get the heatsink on, so i figured, "I'll turn it on for a minute and make sure I have it in correctly."

The computer booted and immediately shut off with no damage. The chip still powers my mom's computer several years later.

AMD now has the same kind of technology, but they did lose a lot of people to burnt CPUs.

[Lauretij2]

I've installed a lot of CPU's. The amd's burned up real bad. I had the heat sink on, but apparently it wasn't 'just right' for them.

I'm sure they are better now, but I'm ruined forever for the lack of a shutdown circuit that cost me a lot of money. Intel always has had that feature.

I tried AMD, got burned (figuratively and literally)

[NJ_gent]

"I've installed a lot of CPU's. The amd's burned up real bad. I had the heat sink on, but apparently it wasn't 'just right' for them."

I'm not trying to be insulting at all, but if you've installed a lot of CPUs, how could there be any problem with the heat sinks not being on correctly? I've never had a problem with a heat sink before, especially on Socket 370 and Socket A (462). Clip one side on, press down on the other side until it clicks in. I understand some others have had problems with the older style sockets, but I've just never understood how that's possible. Can someone enlighten me?

"I'm ruined forever for the lack of a shutdown circuit that cost me a lot of money. Intel always has had that feature."

Actually, Intel didn't add the feature until the Pentium II and Celeron processors. AMD added it to the AthlonXP and AthlonMP processors, but some mainboards didn't properly support the AMD thermal diodes immediately. All currently produced processors from both companies contain thermal controls. Unfortunately, Intel's Pentium 4 processor will not give any errors if it is overheating. Instead, it will actually just run slower than it normally should to cut down on heat production. What this means is that if your fan goes on your Pentium 4 processor, you're going to end up with a very slow computer for seemingly no good reason. Not an ideal solution, in any case. When something fails, I usually like to know about it.

[HAL9000]

no advantages at all that I'm aware of to big-endian except for making hex dumps look nice.

Some data are processed more efficiently with big-endian storage.

The x86 little-endian architecture is a relic of Intel's calculator chip designs of the 1970s. Even Intel's engineers admit that little-endian was an bad decision for x86. I wish that Intel would add big-endian access to their next generation of chips.