
HOW WE
CONFIGURED THE TEST SYSTEM:
Due to the
significant variations in benchmark scores we have seen from one site to
the next, we feel it is necessary to explain exactly how we
configure our test systems before running any benchmarks.
When testing the Athlon XP 2600+, the first thing we did was
enter the EPoX 8K3A+'s system BIOS and set the board to
"Load Optimized Defaults". We then configured the Memory
manually to run at 166MHz, with the CAS Latency and other
memory timings set to 2-2-5-2, with 4-Way Bank Interleaving
and a 1T command rate.
The hard drive was then formatted, and Windows XP
Professional was installed. After the Windows installation
had completed, we installed the VIA 4-In-1 drivers and then
hit the Windows Update website and downloaded all of the
available updates, with the exception of the ones related to
Windows Messenger. Then we installed the rest of the
necessary drivers, and disabled then removed Windows
Messenger from the system. Auto-Updating and System Restore
were also disabled, and we setup a 768MB permanent paging
file. Lastly, we set Windows XP's Visual Effects to "best
performance", installed all of the benchmarking software,
defragged the hard drive and ran all of the tests at the
CPU's default and overclocked speeds.
The Athlon XP 2200+ benchmarks were run on a
Gigabyte
GA-7VRX KT333 Motherboard that was supplied to us by AMD
when the 2200+ launched back in June. Abit's IT7
MAX motherboard, based on the Intel 845E chipset, was used to
test the 2.2GHz and 2.4GHz Pentium 4s. Lastly, an Iwill
P4R533-N motherboard was used to test the 2.53GHz Pentium 4,
as this was the only i850E based motherboard we had in the
lab that has been certified to run the RDRAM at PC1066
speeds.
 |
The
HotHardware Test Systems |
The KT333 + 512MB of PC3200
- A Perfect Match |
|
CPUs:
AMD Athlon XP 2600+
(2133MHz)
AMD Athlon XP 2200+ (1800MHz)
Pentium 4 2.2GHz
Pentium 4 2.4GHz
Pentium 4 2.53GHz
Common
Hardware:
NVIDIA GeForce 4 Ti 4600
On-Board Sound
IBM DTLA307030 30GB ATA/100
7200 RPM HD
Creative 52X ATAPI CD-ROM
Standard 3.5" Floppy
320W Power Supply
Other
Hardware:
EPoX 8K3A+ VIA KT333 Based
Motherboard (2600+)
Gigabyte GA-7VRX KT333
Motherboard (2200+)
Abit IT7 i845e Motherboard
(2.2GHz & 2.4GHz)
Iwill P4R533-N Motherboard (2.53GHz.)
512MB of PC3200 DDR RAM @ CAS 2 (2200+, 2600+, P4 2.2 & P4
2.4)
512MB of Samsung PC800 RDRAM
(2.53GHz)
512MB of Samsung PC1066 RDRAM
(2.53GHz)
Software:
Windows XP Professional
Direct X 8.1
NVIDIA Detonator 4 reference drivers, version 28.32
VIA 4-in-1's v.4.42
VIA 4-in-1's v.4.38
Intel INFs v4.00
 |
Benchmarks &
Comparisons With SiSoft SANDRA |
Synthetic Number
Crunching |
|
SANDRA (the System
ANalyzer, Diagnostic and Reporting Assistant)
is an information and diagnostic utility put out by the good
folks at
SiSoftware. Not only is this program used for
benchmarking, but it can provide a host of other information
about your installed hardware and operating system. We
began our benchmarking with four of the built-in sub-system tests
that are part of the SANDRA 2002 suite (CPU, CPU Cache,
Multimedia and Memory Bandwidth). Default, and
overclocked scores are both represented below.
CPU
Test
Athlon XP 2600+
2133MHz (16x133)

|
CPU
Test OC
Athlon XP 2600+
2416MHz (16x151)
 |
CPU Cache Test
Athlon XP 2600+
 |
CPU Cache Test OC
Athlon XP 2600+
 |
Multimedia Test
Athlon XP 2600+

|
Multimedia Test OC
Athlon XP 2600+
 |
Memory Test
Athlon XP 2600+
 |
Memory Test OC
Athlon XP 2600+
 |
At its default
clock speed of 2.133GHz, the Athlon XP 2600+ owned all of
SANDRA's reference systems in ALU performance, but its FPU
scores were second to the 2.66GHz Pentium 4. However,
when we overclocked our CPU to 2.416GHz, nothing could touch
it. In the CPU Cache tests, the 2600+ didn't fare as
well, getting beat by the 2.66GHz Pentium 4 at every block
size except for 8Kb and 16Kb. At its overclocked
speed, the 2600+ did a little better though, besting the P4
at 8Kb, 16Kb and 32Kb. The CPU Multimedia tests show
the Athlon XP 2600+ dominating all of the reference systems
at its default clock speed, and with the processor
overclocked it pulled even further ahead of the competition.
The memory bandwidth tests tell a much different story.
Due to the fact that the Athlon XP 2600+ is still "stuck"
with a 133MHz (266MHz effective) FSB, its memory bandwidth
scores did not increase much with the processor's higher
clock speed. When we overclocked our CPU with a
151MHz FSB, the scores did increase and are actually
pretty good, but they look abysmal next to PC1066 RDRAM.
Rumor has it, AMD will be switching to a 166MHz (333MHz
effective) FSB with future Athlon XPs...Let's hope this
turns out to be true.
 |
Benchmarks
&
Comparisons With The ZD Winstones |
Simulated
Application Performance |
|
NOTE: From this
point forward, we'll be comparing the performance of the
Athlon XP 2600+ to an Athlon XP 2200+, and three Pentium 4s
ranging in clock speed from 2.2GHz to 2.53GHz. We
should mention that the Athlon XP 2200+ was NOT
tested on the same EPoX motherboard as the 2600+, it was
tested on a Gigabyte KT333 based motherboard that AMD sent
out when the 2200+ launched. We also used a different
stick of memory with the 2600+, that was able to run at far
more aggressive timings than the TwinMos modules we used in
our 2200+ review. These changes, along with the
updated VIA 4-In-1 driver package, seem to have generated
some benchmark scores that don't quite scale exactly with
the 333MHz clock speed increase. We want you to keep
that in mind when viewing our benchmark results. Had
we retested the 2200+ on the newer test bed, its scores
would no doubt have been higher. Now, onto the rest of
the benchmarks...
Our first "Real
World" tests we taken with ZD Labs' Business
Winstone 2001 benchmark. The quote below, taken directly quote ZD's
eTestingLabs website, explains exactly what this test
is comprised of:
"Business Winstone is a system-level, application-based
benchmark that measures a PC's overall performance when
running today's top-selling Windows-based 32-bit
applications on Windows 98 SE, Windows NT 4.0 (SP6 or
later), Windows 2000, Windows Me, or Windows XP. Business
Winstone doesn't mimic what these packages do; it runs real
applications through a series of scripted activities and
uses the time a PC takes to complete those activities to
produce its performance scores."
The application used in the Business Winstone tests include:
-
Five Microsoft
Office 2000 applications (Access, Excel, FrontPage,
PowerPoint, and Word)
-
Microsoft
Project 98
-
Lotus Notes R5
-
NicoMak WinZip
-
Norton
Antivirus
-
Netscape
Communicator

The Athlon XP 2600+ simply
crushed every other system we tested in the Business
Winstone 2001 benchmark! This score seemed unusually
high, but was within 2 points of the reference score provided to us by AMD, so we decided to publish it
anyway. The 2.53GHz Pentium 4, even when using
PC1066 RDRAM, couldn't even come close the 2600+.
Score one for AMD... Next
up, we have ZD's
Content Creation Winstone 2002. This benchmark runs a
similar series of scripted activities to Business Winstone
2001, but the tests are comprised of more "bandwidth hungry",
multimedia type applications. The applications used in the Content Creation
Winstone 2002 tests include:
-
Adobe
Photoshop 6.0.1
-
Adobe Premiere
6.0
-
Macromedia
Director 8.5
-
Macromedia
Dreamweaver UltraDev 4
-
Microsoft
Windows Media Encoder 7.01.00.3055
-
Netscape
Navigator 6/6.01
-
Sonic Foundry
Sound Forge 5.0c (build 184)

We got an excellent score of 42.3 in the
Content Creation Winstone 2002 benchmark with the Athlon
XP 2600+. This wasn't fast enough to beat the
2.53GHz Pentium 4 when using PC1066 RDRAM though, but it
did outperform every other system we tested.
MPEG
Encoding & PC Mark 2002 |