* DDJ Home

* Today's Headlines
* Past Headlines
* Microprocessor Articles
* Intel Secrets
* Intel Errata
* Undocumented Corner
* Processor Manuals
* Motherboard Manuals
* Links

Microprocessor Resources

Microprocessor
Headline News

Top Stories for October 20, 1999 (details below)

ZD Net News AMD eyes 1,000MHz chips with new fab
The Register Files
The Register Intel's i840 is a chipset that works
The Register Camino egg on face to be wiped away by i840?
The Register Camino egg on face to be wiped away by i840?

 

Microprocessor Headline News

Collected By Robert R. Collins

Week of October 17, 1999

Older News

October 20, 1999

AMD eyes 1,000MHz chips with new fab

By John G. Spooner

October 19, 1999
ZD Net News

Advanced Micro Devices is expected to escalate its battle with Intel Wednesday with the dedication of a fabrication plant capable of producing 1,000MHz processors.

The plant, called Fab 30, is key to AMD's (NYSE:AMD) plan to keep ahead of Intel Corp. (Nasdaq:INTC), because its state-of-the-art facilities will allow for the manufacture of Athlon processors at higher clock speeds than are currently possible.

The Register Files

Intel's i840 is a chipset that works

By Mike Magee

October 19, 1999
The Register

When we met Raghu Murthi, the workstation marchitecture manager at Intel Dupont last Friday, he would not publicly comment about the i840 chipset and associated motherboard which the company launches on the 25th of October next.

But he did enthuse about how his team had designed a powerful and very wonderful chipset and fell into a cunning British hack trap when we asked: "Shouldn't you guys have designed the i820 then?" To his credit, he blushed. Murthi had obviously not had time to read the now famous Intel document on how to deal with British hacks.

Camino egg on face to be wiped away by i840?

By Peter Sherriff

October 19, 1999
The Register

If you were a major chip manufacturer with a problem chipset, you might like to have something on the back burner to take its place.

And as luck would have it, Chizilla is in just such a position.

Camino may have deposited a lot of egg on Intel's corporate face, but the Great Satan of Chips appears to have the luck of the devil in that the i840 workstation chipset is just around the corner and actually works.

Intel pulls into fast lane as workstation plans unfold

By Mike Magee

October 19, 1999
The Register

UK readers will be aware of Desperate Dan, a cartoon character in the Dundee publication The Dandy. Pictured as a cowboy with an insatiable appetite, Dan used to shave with a blowtorch and eat enormous cow pies*, complete with horn and tails.

What has Desperate Dan to do with Intel? Well, aside from the fact that its CEO Craig Barrett likes huntin', fishin' and, presumably, eatin', Intel wants the biggest cow pie in world -- and that's a huge chunk of the lucrative market.

October 19, 1999

Taiwan's chipset suppliers brace for Intel's PC133 retort

By Jack Robertson

October 18, 1999
Electronic Buyers' News

It's marketing tremors from Intel Corp., not the aftermath of Taiwan's recent earthquake, that is weighing on the minds of the island's three independent core-logic- chipset vendors.

Although the trio is beating Intel to the punch by rolling out faster chipsets equipped with a PC133 SDRAM interface, Via Technologies Inc., Silicon Integrated Systems Inc. (SiS), and Acer Laboratories Inc. (ALI) are bracing for the MPU giant to come roaring back next year with a competitive PC133-enabled device of its own.

Chip set suppliers look for edge in Intel-dominated world

By Jack Robertson

October 18, 1999
Electronic Buyers' News

It's marketing tremors from Intel Corp., not the aftermath of Taiwan's recent earthquake, that weighs on the minds of the island's three independent core-logic chip set suppliers -- Via Technologies Inc., Silicon Integrated Systems Inc. (SiS), and Acer Laboratories Inc. (ALI).

Although each of the trio has beaten Intel to the punch by rolling out chip sets equipped with a PC133 SDRAM interface, they are bracing for Intel to come roaring back next year with a competitive PC133-enabled device of its own. After months of characterizing the PC133 standard as a non-starter, Intel threw its support behind the memory interface in August. But even after embracing the technology due to the lack of Direct Rambus DRAM availability, Intel wants to retard PC133 adoption until it can get its own chip-set into the market next year, according to many analysts. 

Intel: Speed Does Matter
Faster CPUs may not be vital, but they'll keep coming, and customers will keep buying.

By Matt Hamblen

October 18, 1999
Computerworld

Intel Chief Executive Officer Craig Barrett defended his company's release of ever-faster Pentium III processors last week, despite complaints about the limited number of applications for corporate users.

Barrett concedes that 95 percent of the time, corporate users won't need the latest, fastest processors, as new desktop models exceed 600 MHz. But the other five percent of the time, he says, users might want that faster processor to create a graphics-intensive, Web-based application.

"Unless you buy that capability, you'll never be able to take advantage of that 5 percent," Barrett told 8000
attendees at the Gartner Group Symposium/ITxpo '99 here last week.

Micron contemplates DDR chipset licensing program

By Jack Robertson

October 15, 1999
Electronic Buyers' News

Hoping to stimulate sales of higher-speed memory, Micron Technology Inc. said it may put its double-data-rate, SDRAM-enabled Samurai chipset out for license to accelerate the market's adoption of PC266 DDR devices.

Dean Klein, vice president of the integrated-products group at Boise, Idaho-based Micron, said the company “isn't in the chipset business, but [wants] to enable the early penetration of DDR memory into the market.” Klein said widespread availability of DDR chipsets is the gating function to ramping up PC266 sales.

Intel fixes cache glitch on its 8-way motherboards

By Ephraim Schwartz

October 15, 1999
InfoWorld Electric

Three weeks after a problem with Intel 8-way motherboards and on-chip Level 2 cache first came to light, company officials said today that the glitch which caused systems to hang when using the 1MB on-chip Level 2 cache configuration is fixed.

However, the problem persists in 512KB Level 2 cache configurations, according to Otto Pijpker, an Intel spokesman.

"We don't have a solution yet for the 512KB configuration," Pipker said.

Intel moves could prompt graphics showdown

By Brooke Crothers

October 15, 1999
C/Net

A new wave of budget products from Intel may force a showdown in the booming market for graphics chips which are bringing greater realism to personal computers.

Graphics processors, which control the images users see on their screens, have improved at a frantic pace over the last three years and many are now as sophisticated--some packing in more than 20 million transistors--as the flashy Pentium III or PowerPC G4 central processors just hitting the market.

USB to go the distance

By Ephraim Schwartz

October 15, 1999
InfoWorld Electric

Universal Serial Bus (USB) technology will continue to scale next year not only in terms of its performance, but also its topology thanks to enhancements currently under development.

Icron Systems will ship during the first quarter of next year ExtremeUSB, a USB technology that will extend to 130 meters the maximum range from PC to end-device. Currently, USB can support up to 5 meters between devices and with hubs linked, it has a maximum range of 30 meters. Icron's technology also uses standard Category 5 cable, making it compatible with the vast installed base which this type of cable has.

The Register Files

Intel 1100MHz 'Athlon killer' to launch in December

By Peter Sherriff

October 18, 1999
The Register

Intel knows in its heart that Coppermine just can't cut it in the race against Athlon, so Chipzilla has a little surprise up its sleeve - the next generation of IA32 processor, codenamed Willamette, could be here a staggering nine months early.

US sources say the chip will have a paper launch at the end of December, with product in the shops two months later, although if AMD keeps up the pressure it could be even sooner. Presumably this two month gap is to enable OEMs to shift bucketloads of Coppermine systems before they're rendered unsaleable by the new super chip.

1GHz copper Athlon production starts next week

By John Lettice

October 15, 1999
The Register

The first copper Athlon production should roll out of AMD's Dresden Fab 30 next week, after the plant's official opening on Wednesday. That means that despite problems with cost, and associated negotiations with Motorola to help fund Dresden development, AMD is probably on schedule for 1GHz copper Athlon parts for Q1 next year.

AMD has an unpleasant history of not quite managing to get its production schedules together, but so far Dresden doesn't look at all bad. It's the first plant outside of the US capable of using a copper process, and as it comes into commission this year it seems to have been hitting its targets.

1GHz Athlon to arrive 10 Jan 2000

By Mike Magee

October 18, 1999
The Register

Sources close to AMD's plans have said that the company is likely to intro a 1GHz Athlon early next year.

That is likely to throw Intel into near-panic and makes the likelihood of an 800MHz Coppermine, followed by higher clock speeds shortly afterwards, even more likely.

Via hit by chipset glitch

By Mike Magee

October 14, 1999
The Register

Reliable sources at customers of chipset manufacturer Via are reporting that the initial stepping of product is displaying bugs which will have a knock-on effect on motherboard, and therefore PC, shipments.

Via production has also been adversely affected by the Taiwanese earthquake, the sources add. Despite bullish statements by the Taiwanese government about the integrity of fabs and the quick restoration of supplies to firms in the Hsinchu science park, our information is that delicate re-alignment of equipment is taking longer than expected.

Whatever happened to... Intel's processor serial number?

By Peter Sherriff

October 15, 1999
The Register

Remember all the fuss about Intel's processor serial number back at the beginning of the year when The Processor Formerly Known As Katmai was launched?

Gone a bit quiet, hasn't it?

A quick trawl around the web failed to reveal anything going on at all. Even Zero Knowledge Systems, the Canadian outfit that tried to cash in on the initial paranoia, seems to have gone quiet.

Chipzilla risks humiliation with breakneck i820 launch

By Peter Sherriff

October 15, 1999
The Register

For a company which still likes to remind its employees of the FDIV flaw that caused early Pentia to lose their grip on mathematical reality, Chipzilla doesn't seem to learn (see Satan's little helpers get hammers out to i820 mobos).

The knee-jerk reaction to early pressure at the low end from what Intel lovingly refers to as "imitators", resulted in the tragically-underpowered, cache-free Celeron being rushed to market. Had Chipzilla waited for the far superior Mendocino part, the precious Intel brand would have survived untarnished.

Satan's little helpers get hammers out to i820 mobos

By Mike Magee

October 15, 1999
The Register

Information from insiders close to Intel's plans is painting a picture of feverish activity as staff get rid of a third slot to fix an embarrassing problem with Rambus memory on i820 motherboards.

It reminds The Register of Santa Claus surrounded by hordes of gnomes and elves frantically attempting to make toys quickly so that good children will be able to get tangerines, apples and i820 motherboards when they open their stockings on Christmas Day.

Celeron III, anyone?

By Peter Sherriff

October 14, 1999
The Register

It pays to read the small print. When we looked at some leaked Intel roadmap pictures last week, we concentrated on the usual stuff about clock and bus speeds. But on a second glance, another interesting factoid can be seen.

The entry level Celeron systems arriving in the next six months or so bear the legend "Celeron III" and, in tiny type at the bottom, there's a note referring to Timna -- Chipzilla's first system on a chip (SoC) device. But it's not clear to which processors this refers.

Xeon to beat desktop Pentium IIIs in MHz stakes

By Mike Magee

October 15, 1999
The Register

Intel will implement higher MHz Coppermine speeds in its future workstation and server chips before it migrates that technology to desktops.

That emerged from a briefing this morning with Raghu Murthi, director of marketing workstation products at Intel's plant in Dupont.

Murthi also disclosed that Intel will implement up to five flavours of its IA-64 product line, and said that hundreds of systems will be in the hands of its partners by the end of this year, and thouands in early 2000.

Intel'sh Cashcades to cash in on cache inshide

By Mike Magee

October 16, 1999
The Register

As we reported last Sunday, Intel will introduce its Xeon Coppermine processors with 256K of cache on die on the 25th of October next. These were formerly codenamed Cascades, or Cashcades as we prefer it and will neatly sit in Slot 2 of 840 mobos, as revealed earlier.

We will see 512K and 1Mb versions of these chips with much higher clock frequencies probably in Q1 of next year, as Intel tweaks its .18 micron (Coppermine/Goldmine) technology. The Celeron III will also benefit from this ramp and have 128K of cache within the processor.

Major Intel roadmaps ahead: please keep left

By Mike Magee

October 17, 1999
The Register

It's a while since we sat down and took a cool hard look at Intel's processor plans. Plans, a little like budgets, are subject to the famous Grove's Law. We've just invented this law, but it has validity as it has served Andy & Co pretty well ever since the 286 debacle 15 years back.

Grove's Law states: "You will draw up roadmaps which follow the terrain unless you need to suddenly dig tunnels, build bridges, or scrap the whole project if it all goes terribly wrong."

Advertisement
Copyright © 2008 Dr. Dobb's Journal