facebook rss twitter

Review: DFI NB76-EC

by Ryszard Sommefeldt on 5 July 2002, 00:00

Tags: DFI (TPE:2397)

Quick Link: HEXUS.net/qamb

Add to My Vault: x

System Setup




Before we move onto a few pages of benchmarks, lets take a look at the test platform.

DFI NB76-EC, Intel i845G Chipset, Socket 478 Intel Pentium 4 DDR Motherboard
• MSI 645E Max2 LRU, SiS645DX Chipset, Socket 478 Intel Pentium 4 DDR Motherboard
• Intel Pentium 4 1.8A 'Northwood' @ 133MHz FSB (2.4GHz, 2.4B effective)
• 2 x 256Mb Samsung PC2700 DDR Memory Modules (CAS2, 2-5-2)
• MSI G4Ti4400-VTD (MS-8871) GeForce4 Ti4400 128MB
• Adaptec 39160 PCI SCSI Dual Channel U160 controller
• 2 x 73Gb Seagate Cheetah U160 10,000rpm SCSI disks
• Plextor 12/10/32S SCSI CDRW
• Pioneer 6x Slot-load SCSI DVD
• Creative Soundblaster Audigy Player Retail

• Windows XP Professional Build 2600.xpclient.010817-1148
• DetonatorXP 28.80 NVIDIA drivers
• Aquamark v2.3
• Quake3 v1.30
• 3DMark 2001 Professional Second Edition
• Serious Sam: The Second Encounter Demo
• Comanche 4 Demo Benchmark
• PiFast version 4.1, by Xavier Gourdon
• LAME v3.91 MMX and RazorLame 1.1.5 front end

Broadly similar to the P4 system used in the AMD XP2200+ review recently, the 1.8A does a sterling job here in the i845G. A quick note on the Intel Extreme graphics controller. Any attempt to run the CPU at 2.4GHz or the memory above DDR266 caused the onboard graphics to cease working within a short time of booting the system into Windows XP.

The 2.4GHz results were obtained using the MSI GeForce4 Ti4400 which enabled me to run the system at any speeds the memory and CPU could handle, in this case, 2.4B and DDR266.

Windows XP was installed fresh on both systems (GHOST is a lovely tool) for the purposes of this review, the NVIDIA 28.80 Detonators installed and then the test software installed and run and the results collected.

Being an Intel board, the option to install Intel Application Accelerator was there but not taken.

As always, benchmarks were run 3 times and the middle of the 3 runs taken as the result. If any of the recorded results deviated by more than 5% up or down, the 3 results were discarded and run again until 3 consistent results were obtained. In the case of the OcUK SETI Benchmark, due to the time it takes and the need to leave the machine alone while running it for a fair result, it was only run once on each platform.

For the LAME MP3 encoding tests, CDex, a popular CD audio extraction tool, was used to extract Fat Boy Slim's album, Half Way Between The Gutter And The Stars into .WAV format. 11 resulting WAV tracks totalling some 689MB were encoded using the following LAME parameters from RazorLame 1.1.5: -b 320 -m j -h. That resulted in 320kb/sec CBR (constant bit rate) encodings into MP3 format and RazorLame was used to provide visual confirmation of the time taken to do the full encode.

Lets take a look at a few of our favorite 3D based benchmarks first of all.