The Bus Speed Guide — THG.RU

Introduction

The bus speed of modern Pentium and Pentium Pro systems is still a fairly neglected subject. If you want to know about the performance of a system, you always look first at the CPU speed. The higher the CPU speed, the faster the system — that’s how most computer users judge system performance. This is one reason why Cyrix had to invent the P-rating system, because without more information, why would a power user be interested in a 6×86 CPU running at only 133 or 150 MHz? This CPU/performance relationship is also responsible for the fact that Intel still is selling a fair number of Pentium 150’s, although they are barely faster than their cheaper brothers at 133 MHz. We read ‘150’ and hence we think it has to be faster!

I first wrote about this subject quite a while back, when I was creating my overclocking page. I told you about the limitation you put on your system by running it at a bus speed of less than 66 or even 60 MHz. However, one look at my overclocking survey and you will see how many users have ‘overclocked’ their P166 to 180 MHz or their Cyrix 166+ (133) to 3 x 50 = 150 MHz. Others have overclocked from 120 to 150 and still think they have done a really cool job.

Well, to bring this false belief to a final and irrevocable end, I have decided to create this page. The new Asus P/I-P55T2P4 rev. 3.0 motherboard has opened up a new dimension of bus speeds and performance, so somebody had to take the opportunity and do some decent research on it.

The work for this page has been the biggest project I have undertaken for this site to date. To be able to present you bullet proof facts I had to do three days of testing, measuring and tuning. Let me tell you, after the 30th run of Winstone96 you just feel tired, after the 20th crash of Winstone96 you are annoyed, frustrated and fed up. However, eventually I realized the fruits of all this testing and tuning, and now I can present you a page that’s the first of its kind in the entire computer publishing business. It will change the way we think about computer performance.

The Importance of the Bus Speed

First of all, I think I should explain to all those who are unsure what ‘Bus Speed’ means and what it does.

The ‘Bus Speed’ is the frequency a Pentium, Pentium Pro, or compatible is tacted with externally. Only in the first days of the Pentium did the CPU run with the same tact/speed externally as internally. These were the days of the P60 and P66. A Pentium 60 was clocked at 60 MHz from the motherboard and its core was running at the very same speed. Times were changing and faster CPUs were needed, so Intel came out with the P90 and P100. These two fellows were still running at 60/66 MHz externally, but the core was multiplied by a factor of 1.5, which made a 90/100 MHz CPU. A short while later, Intel released the P75 to replace the P60 and P66. The P75 also runs with a multiplied speed: 50 MHz x 1.5 = 75 MHz. You know what has happened from there. Now we have the current Pentium flagship, which runs at 66 MHz externally and 200 MHz (66×3) internally.

So what does this ‘runs externally’ mean?

To start with, the external clock is obviously supplied by the motherboard. At this speed/frequency/tact the CPU is communicating with all of the system components that it’s directly connected to. These are

• the Chipset
• the Ram
• the Second Level Cache

The chipset speaks to the PCI bus at exactly HALF the speed/frequency/tact, so that the CPU talks to all PCI devices through the chipset with HALF the external tact or bus speed.

Now, what do we know about the important performance components of a PC system?

• the CPU — well, that’s covered, it’s specified by it’s clock speed
• the Second Level Cache — aha, determined by type, size and … the bus speed
• the RAM — aha again, determined by the type, the size, the timing and … the bus speed
• the Video Card — ugh, determined by the type and the PCI speed … = bus speed/2
• the Harddisk(s) — allright, determined by the type, the bus type and of course the PCI Speed … hence the bus speed

Well, it seems there’s nothing in that box, which isn’t influenced by the bus speed, so learn: TAKE GOOD CARE OF IT !!!

Facts and Numbers

Now, to show you how important the bus speed really is, I took the opportunity given to me by the recent acquisition of an Asus P/I-P55T2P4 board to do just that. Fortunately, you can run this board at six different bus speeds: 50 MHz, 55 MHz, 60 MHz, 66 MHz, 75 MHz and 83 MHz.

Since there are 4 different multiplier settings of a P54C (1.5, 2, 2.5, 3), I had 24 different CPU Speed/Bus Speed combinations to test. Unfortunately, my P166 would not run any faster than 208 MHz. At 208 MHz, I could not even run Winstone 96 without crashing (though I tried about 25 times with 25 different settings). Hence, the combinations were reduced to 22 under DOS and 21 under Windows 95.

I decided to use two well known benchmarks.

• For DOS, which nowadays only seems to be used by gamers, I chose the Quake 1.06 Timedemo Benchmark. I actually fell in love with this benchmark because it revealed a very nice surprise to me, which will be extremely interesting to you as well. It’s also very easy to execute and does not take long to run.
• For Windows 95 I chose the well-respected Winstone 96, since I just could not get my hands on the new Winstone 32. The practical thing about Winstone 96 is that it’s also used for the P-rating system. Hence I was able to make my own P-rating for the different configurations. The big downside of this benchmark is the time it takes to complete. Together with the defragmentation of the HDD before each test run, it took between 20 and 40 minutes to run, which adds up to about 12 hours of non-stop testing.

Let’s start with the Quake benchmark.

Quake 1.06 Timedemo Benchmark Results, Grouped by Multiplier

I am presenting you this kind of grouping to show you that — amazingly — within the same multiplier setting, the Quake frame rate is directly proportional to the CPU speed and hence a linear function. The gradient is different from multiplier to multiplier and of course depends on your system. This opens up a completely new way of interpreting the results of my survey—but more about this soon on my Quake Survey Page. It gives you the chance to actually calculate the expected frame rate for each CPU/bus speed, once you have one result with the corresponding multiplier.

Quake 1.06 Timedemo Benchmark Results for different CPU and Buspeeds

Now there you are! Hope that everybody can see the importance of the bus speed and also the new possibilities it’s opening for us. To make it even more clear, I have another graph for you. It shows the frame rates of my system with the same CPU speed, but with different bus speeds:

Can you see, what I mean now?

A Pentium 166, running at 2 x 83 MHz is faster than a Pentium 200 at 3 x 66 MHz, and a Pentium at 125 MHz outclasses a Pentium at 180 MHz!!! Doesn’t that give you a completely new perspective about overclocking at all cost? Is anybody still seriously interested at running his 6×86 at 3 x 50 MHz or his Pentium at 180 MHz??

But wait for the Winstone results.

Winstone 96 Results, Grouped by Multiplier

I ran Winstone under 1024x768x65536x75 (sorry, the previous ’16’ ment ’16bit’) on my system. I know, this is not the default configuration used for the P-rating, which would yield higher scores. However, the only thing that counts is the ratio.

I hope that you will excuse me for not running Winstone through all the configurations with the 50 and 55 MHz bus speed. Nobody is running his system at these speeds anyway (I hope) and it would have taken me another four hours.

Unfortunately Winstone would crash at 208 MHz and above, regardless what I would try.

Winstone 96 Results, Grouped by Bus Speed

Now what do you say? The system with a Pentium 166 at 2 x 83 MHz outclasses even a P200, as does the P187.5 at 2.5 x 75 MHz !!! Why bother overclocking your P166 to 200 MHz anymore? Reach 208 MHz or leave it to 166!!! This is great news to all of you criticizers of overclocking!!

I can now say that a normal Pentium 166, running at 2 x 83 MHz, should really be called a P200+ !!! A P150 running at 2 x 75 MHz is actually a P166+, a P120, running at 1.5 x 83 MHz is a P150+, and a P100, running at 1.5 x 75 MHz is a P120+!!

Conclusion

Well, so much for the importance of the bus speed in the real world!! Hope everyone has gotten the message.

These results should change a lot of preconceptions regarding system performance.