Performance tests

Testbeds:

• Motherboards:
  • Gigabyte MA78GM-S2H rev. 1.0 BIOS F4 on AMD 780G.
  • ASUS M3N78-EMH HDMI BIOS 0404 on NVIDIA GeForce 8200
  • Gigabyte MA69G-S3 BIOS D2 on AMD 690G
  • ASUS P5E-V HDMI BIOS 0505 on Intel G35
• Processors: AMD Athlon 64 X2 4000+ and Intel Pentium E2160
• Memory: 2 GB DDR2-800 Corsair XMS2-6400 (DDR2-800, 5-5-5-15-2T)
• Discrete video: ATI Radeon X1900 XTX, 512 MB GDDR3
• HDD: Seagate Barracuda 7200.10 (SATA-II, 7200 rpm)
• Power supply unit: Chieftec CFT-560-A12C

Software:

• OS and drivers:
  • Windows XP Professional SP2, Windows Vista Ultimate SP1 (32-bit versions)
  • DirectX 9.0c, DirectX 10.1
  • ATI Catalyst 8.4
  • NVIDIA ForceWare 174.74
  • Intel Chipset Drivers 15.9.0.1472 dated 25.04.2008
• Benchmarks:
  • 7-Zip 4.10b
  • XviD 1.0.2 (29.08.2004)
  • Doom 3 (v1.0.1282)
  • FarCry (v1.1.3.1337)
  • Unreal Tournament 2004 (v3339)
  • Futuremark 3DMark06 (1024x768)
  • Serious Sam II 2.07 (Shield Generator28 demo, HDR ON, benchmark)
  • F.E.A.R. (v1.04, Maximum system, Middle graphics)
  • S.T.A.L.K.E.R. (v1.0005, ixbt3 demo)
  • Company of Heroes (v1.2)
  • World in Conflict (v1.0, Low Quality)
  • CyberLink PowerDVD 7.3.2911 (we used only decoders for VC1 and H.264), we played video in Media Player Classic (Home Cinema)

As always, we'll start with computational tests in order to find out how the integrated graphics core affects memory performance.

We are pleased to note that engineers managed to optimize operations between the graphics core and memory (in comparison with AMD 690G). RMMA results illustrate these changes in the mode with enabled integrated graphics core. The new chipset from AMD also outperforms GeForce 8200, which comes forward only in the maximum write rate test (however, its average read rate (much more important from the practical point of view) is noticeably lower even versus AMD 690G). This ought to be expected, because AMD must be better informed how to optimize chipsets to interact with its own processors than third-party companies.

Motherboard performance differences between tests with enabled/disabled integrated graphics core has been minimal for a long time already. We can notice the difference only in synthetic memory bandwidth tests and in several sensitive applications (to be more exact, in 7-Zip). When the integrated graphics core is disabled, performance depends on a chipset in the minimum degree (for the AMD platform in general). The strongest influence on performance is exerted by a CPU, memory modules, and BIOS settings (including default values loaded into registers of a memory controller) of a tested motherboard. According to RMMA test results, a leader is the Gigabyte motherboard on AMD 780G. However, test results in 7-Zip come as a surprise. A minimal difference in tests with an enabled/disabled graphics core contradicts synthetic tests and results in other programs. Perhaps, it's because of the BIOS version, which does not disable the integrated graphics core completely, even when a graphics card is installed and Surround View is disabled (but why are there no strange results in other tests?) What's even more amazing, ASUS M3N78-EMH HDMI based on GeForce 8200 demonstrates similar results. However, when we tested Palit N78S on the same chipset, 7-Zip did not drop off the general line. Besides, objective RMMA results of both motherboards are very close to each other. We've come up with the following intuitive-practical conclusion - it's time to update our motherboard test procedure; at least our archivers and encoders. It will be done in the nearest future.

Write a comment below. No registration needed!