facebook rss twitter

Review: SavRow's Katana 3D-9 Portable Workstation

by Ryszard Sommefeldt on 12 April 2005, 00:00

Tags: NVIDIA (NASDAQ:NVDA), Savrow

Quick Link: HEXUS.net/qabca

Add to My Vault: x

SPECviewperf, 3D Studio Max v6, Cinebench 2003

SPECviewperf v8.01

SPEC's OpenGL benchmark starts us off. It's the most interesting test, since it should highlight the difference between the Go and desktop versions of the Quadro FX 1400.

SPECviewperf v8.01

The results highlight a couple of things. Firstly, when the SavRow system is ahead, it's down to the extra framebuffer memory on the card. The speed that memory is clocked at has little effect on the results, it's just important for the test that it's there. When the SavRow system is behind, it's entirely down to the HEXUS system having more available system memory bandwidth and less access latency. I'll show you that later. So SPECviewperf is, at times, system limited, rather than graphics subsystem limited.

3D Studio Max v6

Using a custom script, our HEXUS Superstress benchmark for Max v6 tests, amongst other things, having the Mentalray render core render large amounts of particles, do multiple object copy and transformation, perform alpha blending and do heavy texturing.

3D Studio Max v6

The HEXUS comparison system is faster than the SavRow in our 3D Studio Max v6 test, mainly down to memory performance and slightly because of a faster disk subsystem, when loading the large texture data sets.

Cinebench 2003

Cinebench 2003

Highlighting the fact the SavRow system appears a little lacking in basic platform performance, Cinebench is faster on the HEXUS system.