Keyboard Shortcuts in Stacks
You aren't limited to using the mouse or trackpad to select items within a stack. Once you've clicked a stack to open it, you can type a letter to jump to the first file starting with that letter, or press either the Tab or arrow keys to move among the stack's contents. Press Return to open the selected item.
Visit plucky tree
Series: Inside Your PowerPC
All about busses, chips, emulators, and the new "G3" PowerPC 750
Article 1 of 3 in series
by Geoff Duncan
In TidBITS-334, we looked at the PowerPC processor family and some of the terms and technologies associated with it. If you read the article, your probably know the difference between 68K and PowerPC chips, why clock speed and clock multipliers are important, the difference between Level 1 and Level 2 caches, and the differences among different PowerPC chipsShow full article
In TidBITS-334, we looked at the PowerPC processor family and some of the terms and technologies associated with it. If you read the article, your probably know the difference between 68K and PowerPC chips, why clock speed and clock multipliers are important, the difference between Level 1 and Level 2 caches, and the differences among different PowerPC chips. Part 2 builds on this information and examines additional software and hardware components of the Power Macintosh.
Emulators Forever -- If there's a single thing that made the Power Macintosh successful, it's the 68K emulator built into its system software. Conceptually, the 68K emulator sits between the PowerPC processor and executing code. If code is written for the PowerPC (such code is considered "native"), the emulator does nothing; if the code is written for 68K machines, the emulator translates it to PowerPC code (at a very low level) and passes it to the PowerPC processor. Without a 68K emulator, non-native programs wouldn't run at all on a Power Mac.
The 68K emulator enabled Apple to move the Macintosh to a new processor architecture while retaining strong compatibility with existing programs - undeniably a good thing. At the same time, 68K emulation is also the Achilles heel of the Power Mac because the performance of 68K emulation can't compare to that of native PowerPC code. When the Power Macs were introduced, Power Mac users often took a step backward in performance because the vast majority of Mac software was only available for 68K machines. Though some native applications appeared quickly, major tools like QuarkXPress, Microsoft Office, and FileMaker Pro took a while to become Power Mac-native.
Further, though Apple has ported many critical portions of the system software to take advantage of the PowerPC, much of the system still relies on the 68K emulator. Thus, even high-end Power Macintoshes are caught in a quagmire of 68K code, reducing their potential real-world performance even when running native applications.
If 68K code is so slow, then how long will 68K emulators be around? That's simple: Apple has to keep a 68K emulator in the system forever.
First, the Mac OS relies heavily on the 68K emulator, and though System 8 will contain substantially more Power Mac-native code than System 7.5.3, it's unlikely the entire operating system will ever be fully native. At a basic level, it's not worth the effort to port everything, particularly little-used, non-performance-related portions of the system.
Second, Apple has a vested interest in making sure 68K code and applications continue to run. Almost every Power Mac user owns software written for 68K machines that will never be ported to PowerPC. A good example is Ambrosia Software's arcade classic Maelstrom, which is largely written in 68K assembly language. Porting Maelstrom to the PowerPC would be an enormous undertaking; yet, more than two years after the introduction of the Power Macintosh, Maelstrom continues to run fine in emulation, and is actually a good test of 68K emulators.
Keeping 68K emulation in the system doesn't mean that improvements can't be made. Apple's original 68K emulator was static, translating 68K instructions to PowerPC code one at a time. Emulation performance can be improved with larger Level 1 or Level 2 processor caches (emulator performance is better on the PowerPC 603e chip than the original 603 due to a larger Level 1 cache); however, it's also possible to build a smarter emulator.
With the PCI Power Macs, Apple introduced a significantly faster dynamic recompilation (DR) emulator. The DR emulator watches the 68K code for loops and stashes the translated PowerPC code for later use, rather than translating the same 68K instructions over and over again. However, the DR emulator comes with a slightly higher price in terms of compatibility: programs that do not operate correctly on 68040 machines with their processor caches enabled may not run correctly. Also, Apple's DR emulator only works on PCI Power Macs; the ROMs of earlier Power Macs don't support it.
A good alternative is Speed Emulator, part of Connectix's Speed Doubler. (See TidBITS-292.) Speed Emulator is also a DR emulator, and though it uses more memory than Apple's, it also significantly outperforms it and runs on any Power Mac. Speed Emulator's additional performance is particularly obvious in some areas; for instance, it significantly speeds up the Apple Event Manager, a feature particularly appreciated by AppleScript users.
Both Apple's and Connectix's emulators imitate a 68LC040, which is a problem if you need to use a 68K program with a program that specifically requires a floating point unit (FPU). In the 68K family, FPUs were originally a separate chip devoted to floating point math operations. With the 68040, Motorola built most FPU functions directly into the processor, then (in a cost-cutting move) removed them in the 68LC040. Programs requiring an FPU won't run under emulation on a Power Mac because they correctly determine that an FPU isn't available.
If you need to use programs requiring an FPU on a Power Macintosh, you have two choices: SoftwareFPU and PowerFPU, both from John Neil and Associates. These programs emulate a 68K FPU, allowing 68K programs that require an FPU to function. SoftwareFPU, a $10 shareware product, works fine, though it's not PowerPC native and must pump its math calls through the 68K emulator. PowerFPU is a $20 commercial product that provides PowerPC-native FPU emulation. Since native PowerPC floating point functions are speedy, PowerFPU's performance can be quite good.
The Magic Bus -- When evaluating the performance of a computer, most users refer to the machine's processor type and clock speed, primarily because these terms are common, occasionally comparable, and liberally used in marketing materials. However, another major factor in a computer's overall performance is its bus, the main data path between the processor and other components.
The easiest way to explain a bus is by analogy: think of your computer as a small, one-road town. Most of your computer's components live on the road, and the road must be used every time information has to travel between components. A traffic light controls travel, and a complex series of local laws governs who can go ahead, who has to wait, and how often people can get on or off the road. Two things control how fast traffic moves: how many lanes the road has ,and how often the traffic light changes. One thing controls how efficiently traffic moves: local traffic laws.
In this analogy, the bus width is the number of lanes in the road, the bus speed is how often the traffic light changes, and the hardware architecture and operating system are the traffic laws.
- Bus Width: The bus width is literally how many bits can move across the bus at the same time. Power Macs have a 64-bit bus, meaning 64 bits can travel across the bus simultaneously. Previous Macs had a 32-bit bus, and early Macs had a 16-bit bus. As you might expect, a 64-bit bus is about twice as fast as a 32-bit bus, since it can move twice as much material in the same amount of time. However, a 64-bit bus is also more expensive to manufacture.
- Bus speed: The clock oscillator controls bus speed, as well as processor speed. Basically, a clock oscillator is a tiny quartz crystal that vibrates a certain number of times per second. It's like a metronome for a computer, controlling everything from disk access and screen redraws to memory access and networking, and making sure everything happens in sync.
- Hardware architecture: How traffic flows over the bus is a function of the computer's hardware and operating system design. For example, in older computers, writing data from RAM to a disk meant every piece of information in RAM had to go across the bus to the processor, then back across the bus to the disk system, which would write the information and report back when it finished. These days, it's more common for computers to have a "private road" between RAM and disks. There are numerous other instances of hardware and software engineering in all Macintosh models that strive to improve bus efficiency.
Dishing Your Buses -- The analogy above is a vast over-simplification - in reality, a Macintosh has a number of different buses, most of which exist in sub-systems. SCSI, Ethernet, serial ports, RAM, expansion slots (NuBus and PCI), and input devices all have separate buses, each of which has its own width and (sometimes) its own oscillator.
Bus speed is an important factor when considering upgrades. Clock chipping, a popular, inexpensive method for upgrading Quadras and first-generation Power Macs, involves replacing the computer's clock oscillator with a faster one. Although it invalidates Apple's warranty and not all Macs can be clock chipped successfully (success rates are around 90 percent), replacing the clock chip speeds up the computer's processor and bus, often making for a good all-around performance improvement. For detailed information on clock-chipping, check out Marc Schrier's clock chipping FAQ.
Many PCI Power Macs and clones have both their clock oscillators and processor chips on a removable CPU daughter card, providing a built-in upgrade path to faster clocks and processors. This design permits you to replace the processor and the clock oscillator at the same time. However, in many cases there's still a limit to how fast the main bus can go. In Apple's current models, the upper limit is 50 MHz; Power Computing's PowerTowers go to 60 MHz. This doesn't mean that daughter card upgrades won't be worthwhile for these machines, but rather that they won't improve the performance of every aspect of the system beyond a certain point.
Similarly, upgrade cards for from vendors like Apple and DayStar for earlier Mac models (from the IIci through the Quadra series) should be evaluated not only on the basis of the promised clock speed of the PowerPC chip, but also in terms of the performance constraints imposed by other hardware. In many cases, these cards must traverse a comparatively slow, narrow bus to get data from disks, ports, other devices, and/or RAM, yielding real-world performance levels considerably lower than Power Macs with equivalent processor speeds. Though these upgrades might be adequate for being able to run PowerPC code, they're rarely equivalent to the performance of a used Power Mac and often cost just as much.
The Myth Of Clock Speed -- In the end, what does all this mean for buying a Macintosh these days?
Be wary of hype surrounding the raw clock speed of a particular machine. Although processor speed is (of course) related to performance - and many computer vendors trumpet little more than the clock speeds of their machines - many other factors (processor type, cache, emulation, bus speed, system software, and more) contribute to a machine's real-world performance.
As an example, Power Macs achieve their high processor speeds by using clock multipliers built into their PowerPC processors, allowing the chips to run faster than the machine's clock oscillator. There's no question this improves performance, but there are limits to how much bang-for-the-buck this technique will produce. There's a real performance difference between a 120 MHz machine using a 6x clock multiplier on 20 MHz bus and a 120 MHz machine using a 2x clock multiplier on a 60 MHz bus. Though both machines would function, the first machine will take much more time to access disks, networks, memory, and peripheral cards than the second machine. Even though they'd be roughly equivalent in terms of raw processor performance, the first machine is going to spend more of its processor cycles waiting for its hardware.
Also, pay attention to what processor a particular machine uses. In terms of raw processor power, a 120 MHz PowerPC 604 is significantly (50 to 75 percent) faster than a 120 MHz PowerPC 601 or 603e, just by the nature of the chip designs. However, in real world terms, a machine with a PowerPC 604 might only mildly outperform a 120 MHz 603e with a fast bus, fast video, fast disks, and a good emulator.
If you can't judge computers by their clock speed, what can you use? Increasingly, the only meaningful measures of real-world performance are produced by benchmark applications like Speedometer, MacBench, and Norton Utilities System Info.
I don't feel the results of these programs can be accepted as gospel. Though tests on my Macs produced results in the right ballpark for each machine, none of these applications produced consistent results in repeated testing. Still, programs like these at least attempt to analyze more than a processor's performance, and if results are sufficiently averaged across a wide range of configurations, they might give a reasonable idea of a machine's real-world performance.
For More Information -- These two articles have covered a lot of territory, and I hope they dispelled some confusion about what different bits of hardware do and how you can relate their specifications to real world performance. If you'd like more information, I'd recommend the following technical sources.
For details on PowerPC processors, look at Motorola's and IBM's information, as well as the PowerPC FAQ:
If you're interested in how processors are officially benchmarked (and what a SPECint95 means!), check with the source:
Finally, if you're curious about how the PowerPC chip works in the middle of a Macintosh, I recommend this introduction from Apple's Developer University:
Article 2 of 3 in series
by Geoff Duncan
When Apple introduced the Power Macintosh back in 1994, it pulled off an engineering feat that's rarely been equalled in the computing industry: Apple successfully migrated an operating system and the vast majority of existing applications from the 68000 family of processors to RISC-based PowerPC processorsShow full article
When Apple introduced the Power Macintosh back in 1994, it pulled off an engineering feat that's rarely been equalled in the computing industry: Apple successfully migrated an operating system and the vast majority of existing applications from the 68000 family of processors to RISC-based PowerPC processors. For those of you who are unfamiliar with the jargon, 68000-based Macs are often called "68K" Macs, and RISC stands for Reduced Instruction Set Computing.
More than two years after their introduction, however, understanding the relative merits of PowerPC processors can be confusing, and Apple has further muddied the situation through its use of cryptic model names. What's the difference between a PowerPC 601 and a 603? How much does clock speed matter? What's a Level 2 cache? And what does any of this say about the difference between a Performa 5400 and a Power Mac 7600?
Answers to questions like these are hard to find, and are all but absent from materials Apple and other Mac systems manufacturers make available. Further, news sources (TidBITS included) rarely explain these terms, since we have much to do just to keep up with the latest releases. So, with that in mind, what follows is an overview of PowerPC processors and some of the terms and technology associated with them. Next issue, I'll cover real-world aspects of PowerPCs, including emulators, system software, and performance tuning.
Worth the RISC? All PowerPC processors are software compatible, so as long has you have a PowerPC chip in your Macintosh, you can run any PowerPC-native Macintosh software. PowerPC-based Macintoshes can also run older software written for 68K Macs, but in emulation mode, which tends to be a little slower than what you'd expect from machines touted as blazingly fast. 68K Macs, however, cannot run software written solely for the PowerPC.
This doesn't mean 68K Macs suddenly become useless; most of these machines will be useful for years to come. I certainly plan to continue using mine. In a way, this is a problem for Apple and other software developers, since this long life span means plenty of people will use 68K Macs for years into the future, and these people will want to be able to upgrade their software in order to take advantage of new features.
But, the writing is on the wall. As time goes on, current system and application software will increasingly only work with the PowerPC. It's unlikely that System 8 will be available for 68K Macs, although certain technologies will probably be broken out and made available for older machines. Similarly, software will be optimized for better performance on more recent PowerPC processors, so more recent processors have potential benefits.
Of Clocks & Cache -- I'll just take a moment to define some terms commonly used to describe PowerPC-based Macintoshes:
Clock speed: Clock speed measures how fast a processor processes instructions, and clock speeds are measured in megahertz (MHz); 1 MHz is one million operations per second. Current clock speeds on PowerPC-based Macs range from 50 to 180 MHz, and you can expect 200+ MHz models soon. Before you get excited about a Mac carrying out millions of operations per second, note that - unfortunately - this doesn't mean millions of menu commands per second! An operation is a tiny thing - moving data into a memory location, moving data out of a memory location, or performing a logical transformation. Choosing a menu item requires untold thousands of operations. Similarly, one assembly-language instruction can conceivably consume hundreds of operations - particularly if it's emulated.
Level 1 Cache: A Level 1 cache is a bit of high speed memory built into PowerPC processor. The processor can cache frequently-needed data here and access it rapidly, saving it the trouble of requesting data from RAM or disk. Level 1 caches vary among PowerPC designs, but loosely speaking, PowerPCs have between 16K and 32K of Level 1 cache. Because the cache is built into the processor, you can't upgrade it separately from the processor.
Level 2 Cache: A Level 2 cache works much like a Level 1 cache, but it is separate from the processor and you can upgrade it. Some Macs ship with no Level 2 cache, though most currently ship with a 256K Level 2 cache, and you can often upgrade to 512K or 1 MB. Results vary, but increasing Level 2 cache can improve performance somewhere between 5 and 30 percent, with best results for processor-intensive functions common to science, engineering, or high-end graphics applications. For many users, increasing the Level 2 cache is an inexpensive way to improve the performance of their Macs.
A problem with Level 2 caches is figuring out how much you have - the About This Macintosh dialog doesn't report such information, and it's tough to figure out unless you know what your Mac model shipped with or you feel like opening your Mac and reading cryptic numbers on the cache module. Newer Technologies has a free tool that reports on a Power Mac's Level 1 and Level 2 caches (up to 1 MB). Its results have been accurate on machines I've tested.
Clock Multipliers (or Bus Divider Ratio): A clock multiplier allows a processor to run faster than a computer's bus oscillator, and it's one way recent machines have achieved such astoundingly high clock speeds. As an example, the PowerTower 180 sports a PowerPC 604 running at 180 MHz. Power Computing did this by using the 3x clock multiplier built into the PowerPC 604 in combination with a 60 MHz bus speed on the PowerTower motherboard. Similarly, Apple's Power Mac 9500/150 runs at 150 MHz, three times the unit's 50 MHz bus speed. Different PowerPC chips have different clock multipliers available; for instance, the Performa 6300 uses the PowerPC 603e's 2.5x multiplier to get to 100 MHz using a 40 MHz bus, speed. The upcoming PowerPC 603e-200 and 604e also have 4x, 5x, and 6x multipliers.
Current PowerPCs -- Here's a brief outline of the PowerPC processor family as it relates to the Macintosh.
PowerPC 601: The 601 has the honor of having been the first PowerPC processor available, and it's at the heart of many systems from Apple, IBM, Power Computing, Radius, and other vendors. Mac systems based on the 601 range from 60 to 120 MHz. Development of the 601 has basically ceased in favor of newer processors; however, 601-based systems are certainly still viable today.
PowerPC 603: The 603 is intended to be a low-power version of the 601, aimed at laptops and other devices where power consumption and heat are significant design factors. The PowerPC 603 typically uses between one-quarter and one-third the power of a PowerPC 601 running at the same clock speed. The 603 is also supposed to be equivalent in performance to a 601 at the same clock speed. However, that didn't prove to be the case in Apple's early 603-based 5200 and 6200 series LCs and Performas, or prototype PowerPC-based PowerBooks, mostly due to the 603's small Level 1 cache. A 75 MHz 603 delivered roughly the same real-world performance as a 60 MHz 601.
PowerPC 603e: The PowerPC 603e (also known as the 603+) is basically a 603 with a larger cache and higher clock speed, and is equivalent in performance to a PowerPC 601 at the same clock speed. Most 603-based Mac systems shipping today (including desktop units and PowerBooks) use the 603e chip. Machines based on the 603e should be around for some time, and their speed and performance should continue to improve. Right now, shipping 603e systems peak at 120 MHz.
PowerPC 604: At the moment, the PowerPC 604 chip comes at the high end of the line, with configurations currently shipping at speeds of 120 to 180 MHz. The PowerPC 604 is intended for high-end workstations and servers, and a PowerPC 604 is, roughly speaking, about 50 to 75 percent faster than a 601 running at the same speed, making it the chip of choice for users with processor-intensive tasks. It also consumes two to three times the power of a 601, so don't expect to see a 604 in a laptop or hand-held device.
PowerPC 602: The 602 is a lower-end chip intended for set-top boxes and similar devices. I don't know of any Macintosh-related projects using the 602, but 3D0 plans to use it in a 64-bit game console codenamed M2.
Future PowerPCs -- The PowerPC shows no signs of slowing down in terms of developments of faster processors. Future processors should include the PowerPC 603e-200, which is essentially a 200 MHz version of the PowerPC 603e, sporting that processor's low power requirements and higher clock multipliers. If you'd rather think about improvements to the 604, think about the PowerPC 604e, an enhanced version of the 604, offering higher speeds (166, 180, and 200 MHz, to start with), larger clock multipliers, and increased processor cache size. Quantities of the 604e are shipping right now, and you can expect to see high-speed 604e-based machines from Apple, Power Computing, and other vendors later in 1996.
If you think the 604 is fast, the forthcoming PowerPC 620 is the first 64-bit PowerPC implementation, and it's an even higher-performance processor designed for very high-end systems. The PowerPC 620 uses the same basic design process as the 604e. Although the 620 has been delayed more than a year by problems with technology and reported staffing problems, I expect to see 620-based machines available from Apple and other vendors by early 1997, and some manufacturers have versions of the 620 in-hand now, reportedly running at 200 MHz. The 620 is geared toward multi-processor implementations and transaction processing, and could support up to a whopping 128 MB of Level 2 cache.
IBM and Motorola are currently the sole providers of PowerPC chips, but a little company in San Jose could change that. IBM has granted Exponential Technology a licence to develop PowerPC-compatible processors. Headed by CEO Rick Shriner, a former Apple vice president, and other industry veterans, Exponential plans to use BiCMOS technology to form its processors' core logic, while using more conventional CMOS for on-chip memory - sort of the reverse of the way Pentium chips are manufactured. Although Exponential hasn't made specific speed claims, it anticipates achieving twice the performance of today's microprocessors, which would put their processors in the 300 to 400 MHz range. Exponential still has to prove the feasibility of its technology, but the company has significant financial backing from Apple and other investors, and it claims that its chips will be ready in early 1997.
Stay Tuned -- Next issue, I'll talk about emulators, system software, real-world performance, and how to use this information when buying a Power Mac. Please note that we'll be taking a brief vacation for the Fourth of July and there will be no issue next week.
Article 3 of 3 in series
by Geoff Duncan
Beginning in TidBITS-334, we published a series of articles explaining the technical guts of a PowerPC-based Mac. We examined differences between PowerPC 601, 603, and 604 processors; Level 1 and Level 2 processor caches, the importance of the system bus, the 68K emulator, and other items. Since then, the PowerPC world has changedShow full article
Beginning in TidBITS-334, we published a series of articles explaining the technical guts of a PowerPC-based Mac. We examined differences between PowerPC 601, 603, and 604 processors; Level 1 and Level 2 processor caches, the importance of the system bus, the 68K emulator, and other items.
Since then, the PowerPC world has changed. What is the PowerPC 750, and how is it different than the 603e and 604e chips in other Macs? Why is Apple touting the 750 so heavily? What's a backside cache? This article will answer those questions - and maybe a few more.
PowerPC 750 -- The newest member of the PowerPC processor family is the PowerPC 750, codenamed G3 by Motorola and Arthur by IBM. Like the 601, 603, and 604 series of processors, the PowerPC 750 is a 32-bit RISC processor that's software compatible with the rest of the PowerPC line - meaning that the PowerPC 750 should have virtually flawless compatibility with all PowerPC-based Macintosh software.
The PowerPC 750 provides incremental improvements over previous PowerPC chips. It has larger data and instruction caches (32K each) and is optimized for integer operations, which makes it more spritely at common computing tasks. (The potential downside is that the PowerPC 750 isn't as fast at floating point math as the earlier 604e.) The PowerPC 750 can run from two to eight times faster than a computer's clock, so (in theory) PowerPC processor upgrade cards running as fast as 528 MHz could be designed for Apple's just-introduced G3 Macs.
In addition, branch prediction in the PowerPC 750 has been improved, providing another across-the-board performance increase. In general terms, branch prediction is a low-level technique that processors use when code can do different things depending on the value of particular data. When the processor reaches a point where it must wait to learn a value in order to continue executing code, it makes a reasonable guess at what the value is likely to be and continues processing, instead of waiting simply around for an answer from RAM or an even slower subsystem. When the appropriate data is returned from memory or a subsystem, the processor looks at it and makes a decision. If the processor "predicted" the right path, it's already well on its way to finishing the task (or even done); otherwise, if the prediction was inaccurate, the processor starts again from the decision point, which is what it would have had to do if it hadn't guessed in the first place. Earlier PowerPC processors also do branch prediction; the PowerPC 750 improves on their model by making cached instructions immediately available once a path is resolved, rather than loading the cached instructions separately.
Finally - and perhaps most interestingly - the PowerPC 750 is the first PowerPC chip designed for the Mac OS. This means the chip recognizes and efficiently handles types of byte sequences produced by (reportedly) Apple's and Metrowerks' compilers. All other things being equal (which they aren't; see above) a PowerPC 750 is better at running typical Macintosh software than a PowerPC 604e at the same speed.
Watts the Deal? Most Macs available today use either the 603e or 604e processor. The 604e was meant to be a high-performance chip for workstations, while the 603e was designed as a low-power, higher-speed version of the original PowerPC 601. That's why you've never seen a 604 processor in a PowerBook: they're too hot and they consume too much power.
However, unlike the PowerPC 604 series, the PowerPC 750 offers high performance and low power consumption, using just five watts of power at 250 MHz. In comparison, the PowerPC 604e consumes nearly 20 watts at 200 MHz. Portable machines using the 750 can (in theory) rival the performance of desktop workstations. The PowerPC 750 features four power-saving modes which kick in automatically when functional units of the processor are idle, reducing power consumption and heat dissipation without impacting performance. The PowerPC 750 also includes a thermal assist unit which enables manufacturers to interrupt or slow down processing in response to temperature increases.
Exposing Your Backside -- So what's with the "backside" caches always mentioned in relation to PowerPC 750 systems? Well, a backside cache is just a faster version of a Level 2 cache.
Level 2 caches are comparatively small units of high speed memory (256K to 1 MB) where PowerPC processors stash frequently used instructions and bits of data. Remember that in comparison to almost everything else on a PowerPC-based Macintosh, the processor itself is quite fast. That means it spends much of its time waiting for other systems, like RAM, video, networks, and disk drives. A fast Level 2 cache makes it so the processor can rapidly access frequently needed items and thus spend less time twiddling its thumbs, waiting for slower systems to respond. Increasing the amount of Level 2 cache in a PowerPC-based Mac is one of the cheapest and most effective ways to enhance performance.
The problem with Level 2 caches on earlier PowerPC- based Macs is that the processor accesses the Level 2 cache by crossing the system bus, which acts like a traffic cop for almost every subsystem on the computer. On PCI-based machines, the system bus runs at a comparatively slow 33 to 50 MHz; Apple's new G3 desktop Macs use a 66 MHz system bus.
The PowerPC 750, however, offers a built-in controller for Level 2 cache on a separate, private bus, so the processor need not touch the system bus to use the Level 2 cache. This separate bus can run anywhere from one-third the speed to the full speed of the PowerPC 750, so it's almost always faster than the main system bus. The bus handling the backside cache for Apple's G3 Power Macs runs at half the speed of the PowerPC processor, although some third-party PowerPC 750 processor upgrade cards run at the same clock speed as the PowerPC chip.
Why not make the main system bus faster, rather than having a separate bus for Level 2 cache? In theory, that would be great: we'd all love for our system buses to run at 250 or 300 MHz. In reality, it's much harder and more expensive to engineer a system bus (and its requisite controllers for RAM, disks, networking, and other systems) to run that fast than it is to make a high-speed bus that does just one thing. However, increasing the speed of the system bus always improves performance, and you can expect to see 83 MHz system buses from Apple in future models.
More Info -- Motorola has good technical information about the PowerPC at the first URL below (although, unfortunately, mostly in PDF format). In addition Apple has posted a succinct overview of the PowerPC architecture (including the G4 processors expected in 1999), and IBM also has made some information available.
<http://www.chips.ibm.com/products/ppc/documents /datasheets/ 750/604_750.html>