by September 4, 2000 0 comments

AMD and Intel together have a near-duopoly in the x86 CPU
market, with little competition to be seen anywhere in the near future. However,
this market domination has also meant that both companies have had to
continually innovate and bring out newer and faster processors, to prevent the
rival company from taking over market share. Competition between the two has
been heated, with both claiming to have the fastest CPUs in the market.

year has certainly seen AMD giving Intel a run for its money. Though K6-2 and
K6-3 didn’t perform very well, the Athlon has continuously been beating the
PIII in the megahertz war. The Athlon has traditionally been strong in the FPU
(floating point unit) arena–making it perform well in applications that
require FPU-intensive operations, such as CAD. Due to this, it’s turning out
as a viable option in the high-end graphics market, as well as for gaming.

It must be kept in mind that you can’t switch between these
two processors without changing your motherboard. Even though the Athlon and
early PIIIs look similar, the underlying architecture is very different, and one
processor cannot be installed on a motherboard meant for the other. The greatest
difficulty arising from this situation, therefore, is deciding which one to go

October, we received the latest processors from Intel and AMD, both clocking at
600 MHz. The tests showed that Athlon had a clear edge over the PIII when it
came to graphics and floating-point intensive operations. They were both new
processors, and not readily available in the Indian market. This year, we’ve
compared the 700 MHz versions of both. Not only that, but the PIII that we’ve
used is the Coppermine processor, which is different from the conventional
Katmai (another name for PIII processors other than Coppermines) in several
ways. The primary difference is the 256 kB L1 cache running at full clock speed,
unlike the Katmai that had 512 kB running at half the clock speed. Also, the
earlier cache had a 64-bit data path to the processor, whereas the new cache has
a 256-bit path. Coppermines are manufactured using a 0.18-micron manufacturing
process whereas the Katmai used a 0.25-micron process. There are other
differences like ATC (Advance Transfer Cache) and ASB (Advanced System
Buffering) as well. The Athlon remains the same Slot-A processor with 128 kB L1,
and 512 kB L2 cache, running at 200 MHz FSB, unlike the PIII which uses a 100
MHz bus.

The good news this time is that both processors as well as
their motherboards are readily available in the market. For our benchmarking, we
used the Asus CUBX motherboard, based on the Intel 440BX chipset, and the Asus
K7M motherboard for the Athlon processor. Accompanying this was 128 MB of 100
MHz Simtronics SDRAM, a Creative GeForce AGP card with 32 MB RAM, and a Seagate
U8 5,400-rpm hard drive.

After our benchmarking, we found that neither of the two
processors holds a clear edge over the other. While the PIII performed
moderately better in business applications, the Athlon took a small lead in FPU-intensive
tests. However, this lead didn’t reflect in the gaming and graphics
benchmarks. For gaming, we ran Quake III Arena and the results were almost
identical for both, a frame more here or there. For graphics, we loaded the CPUs
using the Indy3D benchmark from Sense8. Besides having an official run, this
benchmark has the advantage of being highly customizable. So, you could, for
example, add certain 3D features to a test to see how it would affect the
rendering. The end result was the same here too. Barring the motherboards of
course, the remaining configuration was kept consistent for both. Whether
modifying this configuration would give one processor an edge over the other is
something we don’t know. However, for now it seems that the deciding factor
between the two would be the price and availability of the CPU and its

Overall, we have a tie on our hands as far as performance

No Comments so far

Jump into a conversation

No Comments Yet!

You can be the one to start a conversation.