Socket 939 Roundup: Battle at the Top
by Wesley Fink on July 30, 2004 8:00 AM EST- Posted in
- Motherboards
General Performance and Encoding
We never expected the nVidia nForce3 Ultra to be a better performer in Winstone benchmarks than the VIA K8T800 PRO. However, both the nF3 Ultra boards are outperforming the VIA boards by a significant percentage. Since the nVidia 6800 Ultra video card was used for all benchmarking in the roundup, we plan to verify these results with an ATI X800 XT as soon as that board is available to the Motherboard Lab for testing.
The performance of the Socket 939 boards in the new AuotGK 2-pass Media Encoding benchmark was also a mild surprise. We expected Intel to retain their lead in media encoding due to our logical choice of DivX 5.1.1 as the encoding codec. This codec fully exploits some of the features of the Intel chipset, like SSE3, and generally favors the Intel processors. The DVD2AVI engine at the heart of AutoGK, on the other hand, generally favors AMD processors. It appears that you can really show whatever you wish in encoding these days depending on the Encoding program and codec used. Anand did extensive research on media encoding and chose the new AutoGK/DivX 5.1.1. as our new standard due to the fair, balanced and respected results obtained with this combination.
It certainly looks like the combination of Socket 939, FX53, and AutoGK/DivX 5.1.1 balances the field for media encoding. The difference between top systems from AMD and Intel is now less than 3% with Divx, which can be considered equivalent. A quick check using a 2-pass test with the Xvid codec shows the Intel 3.6 at 35.2 and the AMD FX53 at 35.8, again essentially a dead heat. The upcoming move to 64-bit encoding should significantly boost Athlon 64 performance, giving A64 a formidable lead in encoding benchmarks.
Facebook
Twitter
RSS
83 Comments
View All Comments
Richdog - Wednesday, July 21, 2004 - link
bigtoe33 - Wednesday, July 21, 2004 - link
If you are looking for the K8NN939 beta M03 again its availabale at bleedinedge.com on bigtoe's bios bin.http://www.bleedinedge.com/forum/showpost.php?p=32...
grdh20 - Wednesday, July 21, 2004 - link
http://forum.abit-usa.com/attachment.php?s=&po...abit av8 1.3 bios link
grdh20 - Wednesday, July 21, 2004 - link
The msi K8N Neo2 will hit the states in 2 weeks according to MSI.How will the A8V rev. 2.0 be sold in terms of order by model #. Same or different? Rev. # is only printed on the PCB, not the box.
Brickster - Wednesday, July 21, 2004 - link
I love my new Gigabyte 939 board! It kicks a*s and is perrrty in the dark! :)Wesley Fink - Wednesday, July 21, 2004 - link
#36 - The new Gigabyte BIOS DOES have multipliers in BIOS as I detailed in the review specifications. See the link in #35.RyanVM - Wednesday, July 21, 2004 - link
With performance getting so similar between the various boards/chipsets, it seems to me that AT is overlooking some of the bigger differentiating factors: integrated features. It would be nice to see how the USB, Firewire, IDE/SATA, Ethernet, Sound, etc. implementations differ in CPU utilization and overall performance.For example, the Gigabyte (inexplicably) uses a 3rd party GigE controller instead of the NF3-250's controller. What impact on throughput/CPU utilization does this have? How is VIA's SATA implementation compared to nVidia's?
It seems to me that those are become more relevant of questions than "Which board has a higher Winstone rating?"
Parc - Wednesday, July 21, 2004 - link
I had heard the Gigabyte board would not let you adjust the multipler in the bios but I saw a picture where it could be adjusted from the software in windows? Is this true? Any disadvantages to this? Also just how bad is the voltage limit holding the Gigabyte board back? Is there anyway Gigabyte could allow you any more voltage adjusts through the bios in the future or is this board hopeless for volts?bigtoe33 - Wednesday, July 21, 2004 - link
For all who want the AV8 bios...first join the forums over at www.bleedinedge.com, then look in bigtoe's bios bin. you will find the bios under abit amd boards.http://www.bleedinedge.com/forum/showpost.php?p=32...
Wesley Fink - Wednesday, July 21, 2004 - link
#31 - A friend will post a host for the BIOS soon, as I can't handle all the individual requests. Until then email me and I will be happy to send it to you.#32 & #33 - The Socket 939 boards WILL handle ECC and non-ECC memory, but I suspect the question is larger than this. Socket 940/Opteron requires REGISTERED memory and the ECC can be on or off. We have tested the 939 with Registered memory and it is definitely NOT compatable with Registered memory. If you want to keep using Registered memory (Which you may have bought for 940) you will have to stick with an Opteron.