AnandTech Exclusive: Sun's sub-$1000 SunFire X2100 Server
by Kristopher Kubicki on September 12, 2005 1:29 AM EST- Posted in
- Systems
Rendering Benchmarks
Mental Ray 3.3.3
Below, we are running the 32-bit binaries provided by Alias Wavefront. You may be interested to see how some single CPU setups perform on the same test render here. Once again, we are running the same Maya benchmark file found in our other reviews. We ran Mental Ray via Maya using the command:
# maya_render_with_mr -file Benchmark_Mental.mb
Notice that there is a minute difference in performance between the X2100 and the V40z, even though both are essentially running a single core 2.2GHz processor. Some of this is due to the Aquarius's advantage of unbuffered DDR, but also, the different core logics play a part in each system's performance.
Shake 3.5c
We are planning on moving this benchmark to Shake 4.0, but since we had all of our existing benchmarks in 3.5c, we opted for 3.5c one last time. We took the opportunity to run a benchmark script by Lindsay Adams, which you can download here. The benchmark script renders 10 frames under various effects using one or multiple CPUs. We sum the render times and display them below. The times recorded are the averages of three runs. The command run for this benchmark is:
# shake -exec hardware_test_v01.shk -vv
Our SunFire X2100 actually took a bit of a performance hit on this test, something that we hadn't expected; although, we had a similar issue when running the four socket V40z in 2CPU render mode.
Mental Ray 3.3.3
Below, we are running the 32-bit binaries provided by Alias Wavefront. You may be interested to see how some single CPU setups perform on the same test render here. Once again, we are running the same Maya benchmark file found in our other reviews. We ran Mental Ray via Maya using the command:
# maya_render_with_mr -file Benchmark_Mental.mb
Notice that there is a minute difference in performance between the X2100 and the V40z, even though both are essentially running a single core 2.2GHz processor. Some of this is due to the Aquarius's advantage of unbuffered DDR, but also, the different core logics play a part in each system's performance.
Shake 3.5c
We are planning on moving this benchmark to Shake 4.0, but since we had all of our existing benchmarks in 3.5c, we opted for 3.5c one last time. We took the opportunity to run a benchmark script by Lindsay Adams, which you can download here. The benchmark script renders 10 frames under various effects using one or multiple CPUs. We sum the render times and display them below. The times recorded are the averages of three runs. The command run for this benchmark is:
# shake -exec hardware_test_v01.shk -vv
Our SunFire X2100 actually took a bit of a performance hit on this test, something that we hadn't expected; although, we had a similar issue when running the four socket V40z in 2CPU render mode.
24 Comments
View All Comments
dilidolo - Monday, September 12, 2005 - link
If you want to conpare hardware, then run same OS.If you want to benchmark OS/App, then use same hardware.
With Anand's testing method, you can't really tell if it's OS or hardware that makes the difference.
Furen - Monday, September 12, 2005 - link
Considering that the 175 is supposed to be cheaper than the 152, Sun's margins on the Dual-core system are huge. Then again, these systems cannibalize their 2-way single-core systems, so of course they have to be priced accordingly.MCSim - Monday, September 12, 2005 - link
https://www.sun.com/nc/05q3/">Sun NC05Q3 eventMCSim - Monday, September 12, 2005 - link
https://www.sun.com/nc/05q3/">Fixed linkgibhunter - Monday, September 12, 2005 - link
Everyone retains their badge number in our company too. I think it makes it easier from an accounting standpoint.When I left in 99 and came back in 03 I still had the same badge number even though new employees' numbers have gone up by about a hundred.
splat1 - Monday, September 12, 2005 - link
The board is a rebranded tyan K8E.surprized the guys from anandtech didnt see that.
ftp://ftp.tyan.com/img_mobo/i_s2865.tif">ftp://ftp.tyan.com/img_mobo/i_s2865.tif
Looks like they pulled the pci slots and some other stuff but Im pretty sure the board layout was tyan's design.
Furen - Monday, September 12, 2005 - link
I would call it a similar design, it's not easy to "pull" stuff out of a motherboard.splat1 - Monday, September 12, 2005 - link
Can anyone find a better pic of the sun motherboard. From what I can tell the board layout is identical. Even the smdc card looks to be the same one that tyan uses. I guess I could try to flash the k8e with the sun bios and see what happens.MCSim - Monday, September 12, 2005 - link
Sun is doing really well with the Opterons.Questar - Monday, September 12, 2005 - link
Yeah, look at those profits they're making!