The Low End Mac Mailbag

More Safari Tips, Beige G3 Stuck at 266 MHz, Rage Pro 128, 'Dark Side' CPU Comments, and More

Dan Knight - 2003.02.13 - Tip Jar

More Safari Tips

After reading the URL selection tips posted last week, Ed Hurtley suggests:

Try triple-clicking.

First click: puts pointer where you clicked.

Second click: selects the word you clicked in.

Third click: selects the whole line.

This works in both the URL line and the Google bar.

P.S. to me, in both Safari and IE, LEM's 'fave icon' is 'LEM' in blue letters, not the little computer icon (). (Odd, I just checked in IE on my PC, and it's the computer. But it's 'LEM' on two of my Macs.)

There's a reason for that. Safari never cleans out the favorite icon cache for a website if you have any bookmarks for it. Any, including making Low End Mac your homepage. I spent a week fighting that one myself following a step-by-step tutorial on RAILhead Design. It wasn't until I changed my home page in Safari that it finally changed the favicon.

Beige G3 Stuck at 266 MHz?

Having problems upgrading his beige G3, Greg Agri writes:

I have a G3 beige with a 350, computer came with a 266 MHz CPU. I put a 350 in I got from eBay. Anyways, its in, but the computer still thinks its a 266. Did I miss something? Maybe there is some support docs you guys know of. I am a bit new to this....

Back in November 1997, we posted Clocking the Power Mac G3, which explains how the J16 jumper block must be configured for different CPU speeds. In general, you should leave the bus speed at 66 MHz and adjust the multiplier. You should have no trouble at all running a 350 MHz G3 at 333 MHz and may well be successful at 366 MHz. Getting it to run reliably at 400 MHz may not work.

When I find the time, I want to try overclocking the G3/333 in the beige G3 we use here at Low End Mac.

More on Radeon 7000, a Beige G3, and OS X

In response to my reply in Radeon 7000, a Beige G3, and OS X, Francis Gibson notes:

Thanks so much for the reply. I have found an old Rage Pro 128 for around $50 from a dealer. Haven't had much luck on eBay with old PCI cards. The onboard video of this unit is very bad . . . loose connections, so I need a replacement. The Rage Pro will work with OS X.

CPU Competition

In reply to The Value of a Low End Mac, Steven Hunter writes:

In a recent email, Eric McCann listed several PC CPUs in an attempt to illustrate that there is some sort of "problem" with the wide variety of CPUs available over the years.

I disagree. The competition between AMD and Intel (and to a lesser extent Cyrix IDT, and others) has improved PC CPUs tremendously. If Motorola had any real competition for Apple's CPU business, the G5 would probably be out by now.

His list is also inaccurate:

Intel Chips

Pentium (P5): Socket 5 (60-200 MHz)

Pentium MMX : Socket 7 (200-266 MHz)
Pentium Pro: Socket 8 (150-200 MHz)
PPro Xeon: Socket 8 (150-200 MHz)
Pentium II: Slot 1 (233-300 MHz)
Pentium III: Slot 1 (300-450 MHz)
PII Xeon: Slot 1 (233-300 MHz)
Pentium III: Slot 2 (500-1033 MHz)
PIII Xeon: Slot 2 (500-850 MHz)
Celeron: Slot 2 (500-1000 MHz)
Pentium III: Socket 370 (750-1330 MHz)
Celeron: Socket 370 (750-1400 MHz)
Pentium 4: Socket 423 (1.2-2.4 GHz)
Pentium 4: Socket 478 (1.8-3.06 GHz)
PIV Celeron: Socket 478 (1.7-2.2 GHz)
PIV Xeon: Socket 603 (1.8-2.8 GHz)

Cyrix/National Semiconductor Chips

5x86 (486 class): Socket 5

6x86(P5 class): Socket 5/7
MII (PII class): Socket 7/"Super" Socket 7 (100 MHz bus)
MediaGX (PII class): "Super" Socket 7

AMD Chips

K6 (P5 Class): Socket 5/7 (two versions)

K6-2/3 (PII Class): Socket 7/"Super" Socket 7
Athlon: Slot A (300-1100 MHz)
Athlon: Socket A (900-1400 GHz)
Duron: Socket A
Athlon XP: Socket A

As you can see, the only real confusion comes from Intel's product lines. AMD and Cyrix had quite straight forward products. This is due in part to Intel trying to drive them out of business.

I don't think McCann was trying to be exhaustive.

According to your list, Intel has used nine different slots/sockets for the Pentium over the years, Cyrix has used three (although in some cases, the same CPU was designed for two different sockets), and AMD has used four (with each version of the K6 available in two different connections). In all, 12 different CPU connectors have been used.

Contrast that with Apple. The PCI Power Macs (7300-7600, 8500-8600, 9500-9600) all used the same CPU slot. The beige G3 and blue & white G3 used the same ZIF socket, although at different bus speeds. The Power Mac G4 models with AGP likewise seem to all use the same CPU socket, although at three different bus speeds.

Much simpler, which is exactly the point McCann was making.

Despite your assertion of "tremendous competition," the vast majority of computers sold today have Intel processors. To control the market, Intel has been creating proprietary CPU sockets since the Pentium Pro came out. You can't plug an Athlon into a motherboard designed for a Pentium 4.

The so-called competition has also turned the CPU business into a marketing game, seeing who can bluff the consumer more - AMD with their equivalency ratings or Intel with their less efficient CPUs that run at higher clock speeds. With these kinds of marketing games, the average consumer loses.

Motorola does have competition, a small company known as IBM. The IBM that was showing 1.1 GHz PowerPC chips five years ago. The IBM that wanted to create overall faster CPUs for the Mac, but Apple chose Motorola's slower G4 because it had AltiVec. The same Motorola that once had MHz parity with Intel and falls further behind almost every year.

If Motorola took Apple seriously - 3 million computers a year is not a bad market - they would have a faster G4 or even a G5 today. But Apple is just a small market for Motorola, so they put less resources into the PPC family than needed to keep up with the rest of the industry.

Apple's decision to adopt the G4 has haunted them ever since. It was months between announcing a 500 MHz G4 and shipping it. When Motorola couldn't provide faster CPUs, Apple went the dual processor route - and then abandoned it almost as soon as faster G4s became available. Again Motorola ran stuck. Again Apple went dual processor.

Apple's only hope is to break their dependence on Motorola, which may never ship a G5. The PowerPC 970 provides a way of doing that without abandoning the PowerPC architecture.

In a follow-up email, Hunter comments:

Despite your assertion of "tremendous competition," the vast majority of computers sold today have Intel processors. To control the market, Intel has been creating proprietary CPU sockets since the Pentium Pro

Yes they have. But you don't think Intel went with the current P4 design (one that sacrifices efficiency for a higher clock speed) because they don't feel threatened by AMD, do you? AMD is a huge threat to Intel's monopoly, as they are the only company producing a competitive chip.

came out. You can't plug an Athlon into a motherboard designed for a Pentium 4.

And of course you can't plug a stock G4 into a stock G3 motherboard either...

But in a larger sense, who cares? Competition does not mean that they have to be 100% interchangeable. I can't drop the engine from a Ford pickup into my Geo Prizm, but it doesn't mean that Ford isn't competing with other car manufactures.

I don't really know how much experience you had with PCs when the first Athlon CPUs came out, but let me tell you, it was a real turning point for PC users. The Athlon CPU was the first viable competition to the Intel monopoly to ever come along. Ever. The PII and PIII blew the pants off of the K6-2 and the Cyrix MII/MediaGX. The Athlon stood it's ground, clock for clock, against the PIII. In fact in floating point math it was a 1/3rd faster than a PIII.

It became almost impossible to find one. <http://www.penny-arcade.com/view.php3?date 99-08-11&res=l>

And not because of a shortage of CPUs, but because motherboard manufacturer's were afraid that if they made Slot A motherboards that Intel would bring their wrath down upon them and not sell them Intel brand chipsets. But a brave few stuck it out, and now you have a choice between literally hundreds of different motherboards and several different CPUs.

The so-called competition has also turned the CPU business into a marketing game, seeing who can bluff the consumer more - AMD with their equivalency ratings or Intel with their less efficient CPUs that run at higher clock speeds. With these kinds of marketing games, the average consumer loses.

Not true. The average consumer has more choices now than ever before. However the stupid consumer is far more easily baffled and will pick the system with the biggest numbers for the lowest cost. Invariably this is a Dell PC running XP Home.

However, these people were, in a sense, never in the market for a Mac or AMD based PC. They are the people who want a computer, but don't know why. Ask all the people you know that have PCs why they run the version of Windows and not some other OS or other version. You will hear:

  1. It's what came on the computer.
  2. Everything else is "too complicated" or "too expensive".

AMD's chips have been pushing Intel to make even faster chips. And P4s are in fact faster than PIIIs (But it takes a third again as many CPU cycles to do it.)

The interesting thing will be when AMD's Hammer CPUs and Intel's Itanium II CPUs hit the market.

Motorola does have competition, a small company known as IBM. The IBM that was showing 1.1 GHz PowerPC chips five years ago. The IBM that wanted to create overall faster CPUs for the Mac, but Apple chose Motorola's slower G4 because it had AltiVec. The same Motorola that once had MHz parity with Intel and falls further behind almost every year.

If Motorola took Apple seriously - 3 million computers a year is not a bad market - they would have a faster G4 or even a G5 today. But Apple is just a small market for Motorola, so they put less resources into the PPC family than needed to keep up with the rest of the industry.

Apple's decision to adopt the G4 has haunted them ever since. It was months between announcing a 500 MHz G4 and shipping it. When Motorola couldn't provide faster CPUs, Apple went the dual processor route - and then abandoned it almost as soon as faster G4s became available. Again Motorola ran stuck. Again Apple went dual processor.

Apple's only hope is to break their dependence on Motorola, which may never ship a G5. The PowerPC 970 provides a way of doing that without abandoning the PowerPC architecture.

I agree, but Apple doesn't seem to want to use IBM's chips for whatever reason. So Motorola has a virtual strangle hold on Apple. I think that Apple may have signed some sort of deal with Motorola a while back, and can't switch. Or maybe Jobs holds a bunch of Motorola stock! :)

BTW, in case you were curious, I'm a tech support for about 200 PCs and Macs, split about 60/40. I've ran OSes from Win3.x, to System 7, from OS/2 to OS X, and lots in between. In my opinion, OS X is the best thing to ever come out of Apple (after FireWire). Our machines run a mix of Win2K, WinXP (which I despise), OS 8.6, 9.x, and X.

I cut my computing teeth on an Apple II+, used Commodore for a few years, and was an all-out DOS geek until Apple gave me a Mac Plus. (Okay, I earned it in a holiday sales promotion....) I left the PC world around 1991/2 and have never looked back.

I've done the computer support thing and can only comment how easy it usually is to get a Mac up and running again. It's no wonder the world needs so few Mac geeks in computer support positions - things rarely go wrong, and when they do, there usually easy to fix.

Why Apple chose Motorola's way over IBM's way is beyond me, although I think the sheer horsepower of AltiVec (the supercomputer angle) is probably what got the marketing side excited. Incremental progress with IBM vs. a seeming quantum leap from Motorola.

Although no longer a PC user, I do follow developments over there. Intel has shot itself in the foot so many times - the Pentium math bug, the terrible first generation Celeron, the second generation Celeron that nearly matched the Pentium III, actually reducing CPU efficiency in the Pentium 4 - that it's amazing they can still walk, let alone appear to lead the CPU industry in innovation.

IBM and AMD see the better way, more efficient processors. Motorola and Intel would rather create a more marketable one. And, as in the days of Beta vs. VHS, the consumer usually bases the decision on the wrong factor. Not Beta quality or processing efficiency, but tape capacity and the highest GHz number.

I sincerely hope Apple adopts the PowerPC 970, even if it's only for the Xserve at first....

Nitpick on Eric McCann's Comments

Also in response to The Value of a Low End Mac, Ed Hurtley explains:

Eric lists out all the various processors and slots/sockets that have existed in the x86 world. He does make a few minor oversights:

One, he compares it to the 'one slot' that accepts 603, 604, G3, and G4 upgrades. Yes, but in addition to that 'one slot', we have seen the various Performa methods of madness, the ZIF socket, the iMac board, and the G4 boards. Still fewer than PCs, yes, but it's not just 'the one processor slot'.

His listing of Processors is also misleading. Here are the PC processor slots/sockets he listed as used since 1998: (Total of 8 'desktop' slots/sockets, 5 from Intel, 3 basically from AMD.)

Intel desktop slots/sockets:

  • Socket 5: The original Pentium 60 and Pentium 66. This was completely, totally obsolete by 1996. Intel has an unfortunately penchant for designing new processor cores to use proprietary slots/sockets in their first iteration.
  • Socket 7: This was the socket for the Pentium 75 through Pentium MMX 233, along with the entire K6 line and Cyrix line. By 1998, this was the extreme low end of new computers.
  • Slot-1: This slot lived for quite a while, and was home to the Pentium II, Celeron, and Pentium III. It was new in late '97, and was the primary processor slot through late 2000. The Pentium II, Celeron, and Pentium III were all based on the Pentium Pro core. Companies make adapters that let you use newer Socket 370 processors in these slots. They are very easy to find, and work with even the latest (last?) 1.4 GHz processors (I know someone who has upgraded a formerly 233 MHz computer to 1.4 GHz with one of these.)
  • Socket-370: This was the socketed form of the Celeron and Pentium III. It was introduced for the Celeron processor in 1999, and is just now dying out.
  • Socket 423: Yet a third 'introductory' socket. This was the original Pentium 4 socket, Intel abandoned it before the Pentium 4 became a widely used processor. (Of course, early adopters of the P4 now suffer from their current lack of upgradeability.) The Pentium 4 is an all-new architecture.
  • Socket 478: The current Pentium 4 socket.

Other slots/sockets:

  • 'Super' Socket 7: Technically the same pinout as Socket 7, this was used solely by Cyrix and AMD, and it refers to the capacity of these boards to support a faster processor bus speed. (Just like the Macs have moved from 50 to 66 to 100 to 133, and now to 166 MHz bus.) By 1998, most 'socket 7' motherboards were really 'super socket 7'. It lived on (or limped along, as the case may be) until AMD came out with the Athlon in 1999, when it died very quickly.
  • Slot-A: AMD's answer to Slot-1. This was the initial interface for the Athlon processor.
  • Socket-A: AMD's more recent socket, which takes the later Athlon processors, as well as Duron and Athlon XP.
  • Socket 8: Another short lived one. This was for the Pentium Pro (which was a completely new core, not based on the Pentium at all.) It was never used in mainstream 'desktop' computers, as the Pentium Pro was very expensive, and was primarily designed for servers and workstations. A few enthusiasts used them as high end PCs, though. Some companies make an adapter that lets you use Socket 370 processors in these computers. Due to bus speed issues, they are limited to the Celeron at 766 MHz, though.
  • Slot-2: This was never a 'desktop' interface. It was the interface for Intel's Pentium II Xeon and Pentium III Xeon processors. As these processors went for $1000 to $6000 for a single processor, this can not be considered to have ever been a desktop design.
  • Socket-604: The interface for the new 'Xeon' Processor. This is based on the Pentium 4, despite the confusion of its name.

P.S. Sorry, this message ended up a lot longer than I originally thought it would. (Normally when that happens, I just decide that it was too long winded and cancel it. This time I'll let it go. :-)

I've got nothing against a long email when it's got useful content. Much of this duplicates the email from Steve Hunter, but you organize things by slot/socket rather than CPU. Regardless, there have been a lot of different CPU connections used just since the Pentium began shipping.

McCann never said that the Mac has a single upgrade system, only that three generations of Power Macs used the same one, so upgrades are readily available.

To the best of my knowledge, no Performa has ever had upgradable CPUs, so I don't know what madness you refer to. Likewise, the iMac was never designed as an upgradable computer, so it's irrelevant what kind of connection they put between the CPU/cache card and motherboard.

As for ZIF, Apple has used two - one for G3s and the "Yikes!" G4, another for Power Macs with AGP graphics. Much simpler, and with the right adapter ZIFs for the G3s can be used on PCI Power Macs.

Display Resolution

In response to Pixels and Points, Screens and Paper, Ed Hurtley writes:

Being the Excel aficionado that I am, I tend to make spreadsheets about all sorts of esoteric things. I happen to have recently made one with the sizes, resolutions, dpi, etc. of Apple's monitors. Here you go! (If you prefer, I can provide it in OpenOffice format.)

Thank heavens for the MacLink Plus translators. I double-clicked the file, MacLink converted the file, and I was able to view it in AppleWorks. (Why does the whole world assume everyone can read Microsoft Word and Excel files?) Too bad AppleWorks doesn't have an export to HTML function in the spreadsheet.

Join us on Facebook, follow us on Twitter or Google+, or subscribe to our RSS news feed

Dan Knight has been publishing Low End Mac since April 1997. Mailbag columns come from email responses to his Mac Musings, Mac Daniel, Online Tech Journal, and other columns on the site.

Today's Links

Recent Content

About LEM Support Usage Privacy Contact

Custom Search

Follow Low End Mac on Twitter
Join Low End Mac on Facebook

Favorite Sites

MacSurfer
Cult of Mac
Shrine of Apple
MacInTouch
MyAppleMenu
InfoMac
The Mac Observer
Accelerate Your Mac
RetroMacCast
The Vintage Mac Museum
Deal Brothers
DealMac
Mac2Sell
Mac Driver Museum
JAG's House
System 6 Heaven
System 7 Today
the pickle's Low-End Mac FAQ

Affiliates

Amazon.com
The iTunes Store
PC Connection Express
Macgo Blu-ray Player
Parallels Desktop for Mac
eBay

Low End Mac's Amazon.com store

Advertise

Open Link