Linux.com Article DB: Benchmarking the Tyan Trinity 400

Haven't We Done This Already?

A few weeks ago, Jeff Garzik reviewed Tyan's non-Intel chipset based motherboard for the Celeron and Pentium II/III processor line. The Tyan Trinity 400 takes any Socket 370 or Slot 1 processor, but uses the VIA Apollo Pro chipset. The Apollo Pro has wonderful features such as UDMA/66 support and the ability to use PC133 memory, even if the CPU only runs at 100 MHz.

Since I've been on a benchmarking spree lately (see my three-part Camino article), and Jeff didn't put any benchmark scores in his review (tsk, tsk), I've decided to benchmark this board to see how it stands up to Intel 440BX & i820 based motherboards.

What's In The Box, Brian?

To make direct comparisons to my other benchmark platforms easier, I'm using the same base configuration as in the Camino review:

Intel Pentium II 400 MHz (4X100Mhz)
64MB PC100 SDRAM
Seagate 8.4GB UDMA/66 Hard Drive (with 80-pin cable)
ATI Rage Pro AGP 2X (8 MB)
Acer 50X ATAPI CD-ROM
Intel Pro/100 Ethernet

Note: Detail-oriented readers will note the lack of my Seagate 20.4 GB 7200 RPM hard drive. This has been removed from my test suite for the noble cause of MP3 storage.

What's The Story, Morning Glory?

I'm glad you asked that musical question. My tests use my usual arsenal: Linux-Mandrake 7.0, installed as "development" with all 1395MB of packages; BYTEmark v2; and UNIX Bench 4.0.1. The system BIOS was set to use a 100 MHz DRAM clock. It's important to note that Mandrake was installed with no hard drive optimizations. This is important, and will be explained later. Here are the results:

BYTEmark* Linux/UNIX v2 (10/95)

Memory Index (Linux): 1.679
Integer Index (Linux): 1.48
Floating Point Index (Linux): 2.618

UNIX Bench 4.0.1 (index scores)

Arithmetic Test (type = double): 33.3
Dhrystone 2: 73.4
Excel Throughput: 159.3
File Copy (1024 buffsize, 2000 maxblocks): 188
File Copy (256 buffsize, 500 maxblocks): 217.7
File Copy (4096 buffsize, 8000 maxblocks): 5
Pipe Throughput: 186.1
Process Creation: 255.3
Shell Scripts (8 concurrent): 126.7
System Call Overhead: 148.9
Final Score: 96.7

Compared to the SuperMicro PIIISCA and the AMI S781 I reviewed earlier, this board has very similar scores. Considering this is a very inexpensive motherboard, it's a good choice for the value-oriented consumer. I'm sure the ability to use PC133 memory will improve the performance (sorry, I didn't have any at the time this test was run).

But You Didn't Use Hard Drive Optimizations!

Jeff did use hard drive optimizations in his review, which I haven't used yet. That whole "system may be unstable" warning on Mandrake's install screen made me hesitant. It would be nice to see if the optimizations did make the system faster.

So stay tuned for yet another mind-numbing Linux benchmark. My next article will compare Mandrake with hard drive optimizations versus Mandrake with no optimizations. Same Tux time, same Tux channel!

If Brian Richardson had won the lottery, each and every Linux user in America would have received a free stuffed penguin. Since you have no stuffed Tux, that means he's still a middle-class citizen of Metro Atlanta. Money can't buy love, and Linux is free, so who cares?

Originally Published: Monday, 12 June 2000	Author: Brian Richardson
Published to: enhance_articles_hardware/Hardware Reviews	Page: 1/1 - [Printable]
Benchmarking the Tyan Trinity 400 A few weeks ago, Jeff Garzik reviewed Tyan's non-Intel chipset based motherboard for the Celeron and Pentium II/III processor line. This article benchmarks that motherboard.

	Page 1 of 1
Haven't We Done This Already? A few weeks ago, Jeff Garzik reviewed Tyan's non-Intel chipset based motherboard for the Celeron and Pentium II/III processor line. The Tyan Trinity 400 takes any Socket 370 or Slot 1 processor, but uses the VIA Apollo Pro chipset. The Apollo Pro has wonderful features such as UDMA/66 support and the ability to use PC133 memory, even if the CPU only runs at 100 MHz. Since I've been on a benchmarking spree lately (see my three-part Camino article), and Jeff didn't put any benchmark scores in his review (tsk, tsk), I've decided to benchmark this board to see how it stands up to Intel 440BX & i820 based motherboards. What's In The Box, Brian? To make direct comparisons to my other benchmark platforms easier, I'm using the same base configuration as in the Camino review: Intel Pentium II 400 MHz (4X100Mhz) 64MB PC100 SDRAM Seagate 8.4GB UDMA/66 Hard Drive (with 80-pin cable) ATI Rage Pro AGP 2X (8 MB) Acer 50X ATAPI CD-ROM Intel Pro/100 Ethernet Note: Detail-oriented readers will note the lack of my Seagate 20.4 GB 7200 RPM hard drive. This has been removed from my test suite for the noble cause of MP3 storage. What's The Story, Morning Glory? I'm glad you asked that musical question. My tests use my usual arsenal: Linux-Mandrake 7.0, installed as "development" with all 1395MB of packages; BYTEmark v2; and UNIX Bench 4.0.1. The system BIOS was set to use a 100 MHz DRAM clock. It's important to note that Mandrake was installed with no hard drive optimizations. This is important, and will be explained later. Here are the results: BYTEmark Linux/UNIX v2 (10/95)* Memory Index (Linux): 1.679 Integer Index (Linux): 1.48 Floating Point Index (Linux): 2.618 UNIX Bench 4.0.1 (index scores) Arithmetic Test (type = double): 33.3 Dhrystone 2: 73.4 Excel Throughput: 159.3 File Copy (1024 buffsize, 2000 maxblocks): 188 File Copy (256 buffsize, 500 maxblocks): 217.7 File Copy (4096 buffsize, 8000 maxblocks): 5 Pipe Throughput: 186.1 Process Creation: 255.3 Shell Scripts (8 concurrent): 126.7 System Call Overhead: 148.9 Final Score: 96.7 Compared to the SuperMicro PIIISCA and the AMI S781 I reviewed earlier, this board has very similar scores. Considering this is a very inexpensive motherboard, it's a good choice for the value-oriented consumer. I'm sure the ability to use PC133 memory will improve the performance (sorry, I didn't have any at the time this test was run). But You Didn't Use Hard Drive Optimizations! Jeff did use hard drive optimizations in his review, which I haven't used yet. That whole "system may be unstable" warning on Mandrake's install screen made me hesitant. It would be nice to see if the optimizations did make the system faster. So stay tuned for yet another mind-numbing Linux benchmark. My next article will compare Mandrake with hard drive optimizations versus Mandrake with no optimizations. Same Tux time, same Tux channel! If Brian Richardson had won the lottery, each and every Linux user in America would have received a free stuffed penguin. Since you have no stuffed Tux, that means he's still a middle-class citizen of Metro Atlanta. Money can't buy love, and Linux is free, so who cares?
	Page 1 of 1