iXBT Labs - Computer Hardware in Detail

Platform

Video

Multimedia

Mobile

Other

SiS645DX and SiS648 Mainboards Roundup

October 25, 2002



 

In spite of such a good success of SiS on the arena of high-performance chipsets gained during the last several years the company keeps on reaping the fruits of the "lost 90s". Its image has recently fallen down to the ALi's level and now it's ought to restore it with great difficulties. Although SiS648 is proved to be the most efficient chipset among those designed for Pentium 4 with DDR support, and SiS645DX was one of the best solutions in its category, the developers consider these chipsets just a cheap alternative to the Intel's brand. 

And the consequences are the following: all manufacturers provide standard functions for their solutions based on the SiS chipsets: instead of normal audio controllers or expensive codecs they mount something cheaper, a network interface is obligatory, and other features are provided only if they are supported on the chipset's level. Although there are no rules without exceptions, the general situation looks like the described above, which is proved by a narrower width of such boards which is typical of office products. Besides, such boards usually have a no-frills design, i.e. there is no a model row where the boards could differ in controllers and connectors. 

So, SiS products still bear a label of average inexpensive solutions, and the manufacturers still consider them cheap products which, without added functions, will remain only budget chipsets. We hope that our regular readers are able to give a just mark to SiS's products and choose a high-performance, feature-laden and inexpensive board based on the SiS645DX or SiS648. 

Test conditions

Testbed: 
  • Processor: Intel Pentium 4 2.53 GHz (19x133 MHz), Socket 478 
  • SiS645DX based Mainboards: 
  • SiS648 based Mainboards: 
  • Memory: 2x256 MB PC3200(DDR400) DDR SDRAM DIMM Winbond, CL 2.5 (also used as DDR333 with CL 2) 
  • Video card: Palit Daytona GeForce4 Ti 4600 
  • Hard drive: IBM IC35L040AVER07-0, 7200 rpm
Software: 
    OS and drivers: 
    • Windows XP Professional 
    • DirectX 8.1 
    • SiS AGP Driver 1.09f 
    • NVIDIA Detonator XP 29.42 (VSync=Off) 
    Test applications: 
    • Wstream 
    • VirtualDub 1.4.10 + DivX codec 5.02 Pro 
    • WinAce 2.11 
    • BAPCo & MadOnion SYSmark 2002 
    • Gray Matter Studios & Nerve Software Return To Castle Wolfenstein v1.1 
    • Croteam/GodGames Serious Sam: The Second Encounter v1.07 

Given to the early release of the SiS645DX chipset and an unstable operation of boards based on it with the PC3200 memory, most board makers didn't provide support for that memory speed, that is why we used DDR400 modules as DDR333 with lowered timings. For the SiS648 based models we used the memory in its rated mode with the increased timings. Which approach bears more fruit will be shown in the test results and now take a look at the characteristics of the today's contestants:


 
Board ABIT SA7 FIC VI11 MSI 645E Max2-LRU ABIT SR7-8X ASUS P4S8X ECS L4S8A2 Gigabyte 8SG667 MSI 648 Max Shuttle AS45GTR
Links ABIT SA7 FIC VI11 MSI 645E Max2-LRU ABIT SR7-8X ASUS P4S8X ECS L4S8A2 Gigabyte 8SG667 MSI 648 Max Shuttle AS45GTR
Chipset SiS645DX (SiS645DX + SiS961 B0) SiS648 (SiS648 + SiS963)
Processor support Socket 478, Intel Pentium 4 (400/533 MHz), Intel Celeron
Memory slots 3 DDR
Expansion slots AGP/ 5 PCI AGP/ 6 PCI AGP/ 5 PCI/ CNR AGP/ 5 PCI AGP/ 6 PCI AGP/ 5 PCI/ CNR AGP/ 5 PCI AGP/ 6 PCI AGP Pro/ 5 PCI
I/O ports 1 FDD, 2 COM, 1 LPT, 2 PS/2 1 FDD, 2 COM, 1 LPT, 2 PS/2 1 FDD, 2 COM, 1 LPT, 2 PS/2 1 FDD, 2 COM, 1 LPT, 2 PS/2 1 FDD, 2 COM, 1 LPT, 2 PS/2, 2 IEEE 1394 1 FDD, 2 COM, 1 LPT, 2 PS/2, 2 IEEE 1394 1 FDD, 2 COM, 1 LPT, 2 PS/2 1 FDD, 2 COM, 1 LPT, 2 PS/2 1 FDD, 2 COM, 1 LPT, 2 PS/2, 3 IEEE 1394
USB 2 USB 1.1 + 2 connectors for 2 USB 1.1 2 USB 1.1 + 2 connectors for 2 USB 1.1 2 USB 1.1 + 1 connector for 2 USB 1.1 + 2 connectors for 2 USB 2.0 2 USB 2.0 + 2 connectors for 2 USB 2.0 4 USB 2.0 + 1 connector for 2 USB 2.0 4 USB 2.0 + 1 connector for 2 USB 2.0 2 USB 2.0 + 2 connectors for 2 USB 2.0 4 USB 2.0 + 1 connector for 2 USB 2.0 2 USB 2.0 + 2 connectors for 2 USB 2.0
Integrated IDE controller ATA133
External IDE controller - - AChip ATP865-A - Promise PDC 20376 - - - HighPoint HPT372 + Marvell 88i8030-TBC
Sound AC'97 codec, C-Media CMI9738 AC'97 codec, Avance Logic ALC650 AC'97 codec, Avance Logic ALC650 AC'97 codec, Avance Logic ALC201A AC'97 codec, C-Media CMI9739A AC'97 codec, C-Media CMI9738 AC'97 codec, Avance Logic ALC650 AC'97 codec, Avance Logic ALC650 AC'97 codec, Avance Logic ALC650
Integrated network controller Realtek RTL8100B 10BaseT/ 100BaseTX Realtek RTL8100L Realtek RTL8100B 10BaseT/ 100BaseTX 10BaseT/ 100BaseTX - Broadcom BCM5702CKFB -
I/O controller ITE IT8705F ITE IT8705F Winbond W83697HF Winbond W83697HF Winbond W83697SF ITE IT8705F ITE IT8700F Winbond W83697HF ITE IT8705F
BIOS 2Mbit Award BIOS v6.00PG 2Mbit Phoenix AwardBIOS v6.00PG 2Mbit AMI BIOS v3.31a 2Mbit AMI BIOS v3.31a 4Mbit Award BIOS v6.00 2Mbit Phoenix AwardBIOS v6.00PG 2Mbit Award BIOS v6.00 2Mbit AMI BIOS v3.31a 4Mbit Phoenix AwardBIOS v6.00PG
Form-factor, dimensions ATX, 30.5x24.5 cm ATX, 30.5x24.5 cm ATX, 30.5x24.5 cm ATX, 30.5x24.5 cm ATX, 30.5x22 cm ATX, 30.5x19 cm ATX, 30.5x20 cm ATX, 30.5x22 cm ATX, 30.5x24.5 cm
Set of accessories ordinary ordinary good ordinary good ordinary ordinary good good
Adjustment and overclocking settings average average average rich average average average average average

 

Test results

I'm setting off with the synthetic memory test Wstream to explore the difference between the chipsets. 

 

Well, even with the increased timings the flow throughput of the DDR400 is greater by around 200 MB/s than that of the DDR333. As the test is synthetic we are not going to compare the boards. 

 

In the MPEG4 encoding which also uses a flow memory load, only the ASUS's board (on SiS648) gained a marginal breakaway, and the scores of the other participants make a regular staircase (the difference between two its ends is 3%), and all the SiS645DX based boards go last. 

 

The WinAce archiving is one more test which shows how the boards work with memory, and here the boards on the different chipsets are mixed up because the memory speed is more important factor here. The leaders are the boards from ASUS and Gigabyte, and the outsider is the FIC VI11, the gap between the other samples reaches 5%. 

 

The SYSmark 2002 benchmark doesn't let us say that the SiS648 is definitely a winner (its real business applications doesn't need much memory), and the MSI 645 Max2-LRU looks excellent, though the performance crown still belongs to the board coupled with DDR400 - ASUS P4S8X. 

 

In the 3D games the contest looks more intriguing than ever. Remember that in low resolutions in the low detailing mode the test turns into a competition of the processor-memory bundles (with some deviation depending on the game engine), that is why this test looks more theoretical, like the Wstream. We can see, however, that in the heavy conditions we took measurements in the difference between the chipsets/memory types is still noticeable and the performance is not limited by a video accelerator (it will be limited in higher resolutions and with more complicated graphics). So, even the fastest 3D accelerator can be coupled today with a SiS648 based board and used with the DDR400 for games. The models from ASUS or Gigabyte will be quite a good choice (their scores are a little higher compared to the rest). 

Conclusion

The SiS648 based board with the DDR400 will be a justified purchase if you believe in SiS and want a high-performance base for a Pentium 4 system. Also remember that the SiS648+DDR333 bundle is no better compared to the SiS645DX. 

If you want my advice on what to choose among to today's contestants, take a look at the ASUS P4S8X and Gigabyte 8SG667; besides, the ASUS, MSI 648 Max and Shuttle AS45GTR have a nice set of components and rich functions; and the ABIT SR7-8X and MSI 648 Max provide decent adjustability and plenty of room for overclocking . 

The general recommendations of choosing an appropriate system will be to look through brief characteristics of the boards and read descriptions of the models you like. After that, ignoring insignificant performance difference you should make your final choice taking into account the prices. 
 
  


Dmitry Mayorov (destrax@ixbt.com
Sergei Pikalov (peek@ixbt.com

Write a comment below. No registration needed!


Article navigation:



blog comments powered by Disqus

  Most Popular Reviews More    RSS  

AMD Phenom II X4 955, Phenom II X4 960T, Phenom II X6 1075T, and Intel Pentium G2120, Core i3-3220, Core i5-3330 Processors

Comparing old, cheap solutions from AMD with new, budget offerings from Intel.
February 1, 2013 · Processor Roundups

Inno3D GeForce GTX 670 iChill, Inno3D GeForce GTX 660 Ti Graphics Cards

A couple of mid-range adapters with original cooling systems.
January 30, 2013 · Video cards: NVIDIA GPUs

Creative Sound Blaster X-Fi Surround 5.1

An external X-Fi solution in tests.
September 9, 2008 · Sound Cards

AMD FX-8350 Processor

The first worthwhile Piledriver CPU.
September 11, 2012 · Processors: AMD

Consumed Power, Energy Consumption: Ivy Bridge vs. Sandy Bridge

Trying out the new method.
September 18, 2012 · Processors: Intel
  Latest Reviews More    RSS  

i3DSpeed, September 2013

Retested all graphics cards with the new drivers.
Oct 18, 2013 · 3Digests

i3DSpeed, August 2013

Added new benchmarks: BioShock Infinite and Metro: Last Light.
Sep 06, 2013 · 3Digests

i3DSpeed, July 2013

Added the test results of NVIDIA GeForce GTX 760 and AMD Radeon HD 7730.
Aug 05, 2013 · 3Digests

Gainward GeForce GTX 650 Ti BOOST 2GB Golden Sample Graphics Card

An excellent hybrid of GeForce GTX 650 Ti and GeForce GTX 660.
Jun 24, 2013 · Video cards: NVIDIA GPUs

i3DSpeed, May 2013

Added the test results of NVIDIA GeForce GTX 770/780.
Jun 03, 2013 · 3Digests
  Latest News More    RSS  

Platform  ·  Video  ·  Multimedia  ·  Mobile  ·  Other  ||  About us & Privacy policy  ·  Twitter  ·  Facebook


Copyright © Byrds Research & Publishing, Ltd., 1997–2011. All rights reserved.