Sun's W2100z Dual Opteron Workstation
by Kristopher Kubicki on October 27, 2004 12:05 AM EST- Posted in
- Systems
The Test
Testing the w2100z is difficult, since there are very few dual Opteron 250 rigs on the market right now. We had the opportunity to set up a dual Opteron 250 workstation using store-bought components for our tests. This will likely change as more Tier 1 workstation companies start working closer with AMD. The majority of our tests are conducted in the same manner as our Linux CPU and GPU tests in the past. Since this is a workstation, we want to focus the majority of our benchmarks on applications that we would use in a workstation: compiling, rendering and encoding tests.We took various benchmarks that we are familiar with and ran them on various configurations of the w2001z. As we mentioned earlier, the w2100z does not come pre-installed with any operating system. As a result, we took several different operating systems (Solaris 10, JDS 2.0, RedHat 9, SuSE 9.1) with similar gcc/kernels and benchmarked our programs on them. For reference, we took our off-the-shelf Opteron 250 rig and ran the same benchmarks on it using SuSE 9.1 as well. Although we upgraded the GCC libraries and glibc for the JDS 2.0 configuration, we did not change the kernel - JDS was designed around the 32-bit SMP Linux 2.4 kernel, and upgrading/hacking it would put us well outside the scope of the analysis. As we will see in the next few pages, performance on the proven 2.4 kernel is actually surprisingly good!
Finding properly threaded applications on Linux that are capable of taking advantage of multiple CPUs is a difficult task. While lots of server applications are designed for multiple threads, as the need for multiple users exist, applications that fully utilize multiple processors in a single user environment are difficult to come by. This makes it very challenging for us to benchmark applications effectively using multiple processors. Almost all of our benchmarks only utilize a single CPU unless stated otherwise.
All of our tests take place in the 64-bit environment, with the exception of the JDS rig. Although there is nothing to prevent us from running on 32-bit operating systems, (most) of our applications are mature enough to take a performance hit in a 64-bit environment.
| Performance Workstation Configurations | ||
| Processor(s): | (2) AMD Opteron 250 (130nm, 1MB L2, Socket 94) | |
| RAM: | 4 x 1024MB Buffered ECC PC-3200 CL3 (400MHz) | |
| Motherboards: | Sun K85AE | Tyan K8W S2885ANRF |
| Hard Drives | SCSI u320 Seagate Cheetah 10,000RPM | SCSI u320 Quantum Atlas 10,000RPM |
| Memory Timings: | Default | |
| Video Card(s): | GeForce QuadroFX 3000 | |
| Operating System(s): | SuSE 9.1 Professional RedHat 9 JDS 2.0 Solaris 10 |
SuSE 9.1 Professional |
| Kernel: | Linux 2.6.8 Linux 2.4 (JDS 2.0) SunOS 5.10 s10_63 (Solaris 10) |
|
| Compiler: | linux:~ # gcc -v Reading specs from /usr/local/lib/gcc/i686-pc-linux-gnu/3.4.2/specs Configured with: ./configure Thread model: posix gcc version 3.4.2 |
|
| Drivers: | NVIDIA Linux 1.0-6111 | |
The majority of the differences between these two systems result from different motherboards and slightly different hard drives. Most of our benchmarks do not particularly stress hard drive IO, and focus more on CPU and memory IO. Generally, we average three passes of our results unless stated otherwise. Variance is noted, if above 2%.
Facebook
Twitter
RSS
47 Comments
View All Comments
KristopherKubicki - Wednesday, October 27, 2004 - link
PrinceGaz: Multiple probes.Thanks for all the positive feedback!
Kristopher
slashbinslashbash - Wednesday, October 27, 2004 - link
#12, the Power Mac G5 is Apple's high-end workstation/server. The iMac is Apple's consumer-oriented product.I'd say that the PPC970 (G5) should be roughly comparable to an Opteron/Athlon64 at the same speeds -- at least much more comparable than a P4 at the same clock speeds. It's got a 16-stage integer pipeline, and the FPU units kick serious ass, especially with code that utilizes AltiVec (similar to SSE for x86). It's got a high-speed bus comparable to HyperTransport. AMD's integrated memory controller does give an advantage, but still I think the G5 could hang in there. Really though, any benchmark would come down to optimized code. Companies like Adobe and Apple that have the time, money, and inclination to optimize their applications for Apple hardware will see serious performance advantages over PC hardware. gcc-compiled apps, not so much, although more PPC-specific optimizations should come sooner or later.
Arstechnica has a very lengthy article on the PPC970 architecture. Take a look, if you have the time, and I don't think you'll walk away with a bad opinion of the G5.
BTW, I'm glad to see a Sun review here :) Interesting stuff. I would've liked to see another shot or two of the case itself, with all the components in it; plus a shot from the rear of the system. I'd like to see just how much space is in between that mezzanine card and the motherboard, and I'm kind of confused as to how the extender card fits in. It looks like it simply lies parallel with the motherboard (attaching along the bottom of the motherboard), but most of the time that there's a separate card, it is perpindicular.
thesix - Wednesday, October 27, 2004 - link
#12:"mlittl3 is attempting to compare a consumer product with a high end workstation/server."
You talk like a typical marketing person.
From pure HW design/layout point of view,
G5 _feels_ so much better than w2100z.
What makes me feel better with w2100z is the (new) Solairs OS and Opteron CPU.
Plus, I want to show support for both Sun and AMD :)
thesix - Wednesday, October 27, 2004 - link
I assume AMD CPUs/systems are faster. I haven't got a chance to do any serious testing between my dual-246 w2100z and my friend's dual-2.0GHz G5.I guess I simply don't care that much, anything less
than 10% of performance difference is not that significant to me. However, it's always good to know "my computer is faster than yours". :)
Reflex - Wednesday, October 27, 2004 - link
The problem with the recent Mac article appears to be that it has attracted the zealots to this site in force. mlittl3 is attempting to compare a consumer product with a high end workstation/server. They target different market segments and are designed for different tasks. Furthermore, the support offered for each, and the type of support is considerably different. Its Apples to Oranges.#11: We'll never really know how fast a G5 is in comparison, Apple does not allow independant hardware reviews, leaving it up to websites to purchase and test their own hardware. I know of no site as comprehensive as Anandtech on the Mac side of things as a result, so any claims of G5 performance are pretty much guesstimates and marketing unfortunatly. I wish this would change, it would be nice to know how Apple hardware really does perform as that would assist network admins in determining where it fits/if it fits into their current setup. Just being told by a company that "Its fast and stable" is not enough to make a corporation invest in a platform.
PrinceGaz - Wednesday, October 27, 2004 - link
Great review, the thermal graph is especially good in showing how the cooling performs. Did you attach many probes and take all the measurements in one go, or have to do it multiple times?----
The dual G5 uses some sort of liquid-cooling system inside the case I believe. That probably helps it reduce the noise.
Is a dual 2.5GHz G5 machine as fast as a dual Opteron 250 (2.4GHz) box anyway? Results posted by Apple themselves don't count of course :)
fic - Wednesday, October 27, 2004 - link
#8 you could replace the FX3000 with a 6800 and save another $1000.I would like to see a comparison between the w2100z and 2nd tier companies like Boxx, XI, Aspen, etc. I have been trying to talk my company into moving to a 2x 250 system from a 2x Xeon system to ship with our product. BTW, we do digital medical imaging and need the processing power to do image processing on our 4096x5625 images.
thesix - Wednesday, October 27, 2004 - link
I am surprised to see w2100z is rated as _quiet_ here. I _personally own_ w2100z for almost 3 months now, at home. Apple G5 is _much_ quieter than w2100z. In fact, noise is my biggest complain about w2100z, and I spent lots of time trying to "fix" it. I am confused.http://www.pbase.com/taochen/w2100z
mlittl3 - Wednesday, October 27, 2004 - link
Well, I guess all you apple-haters' comments that x86 computers are cheaper than apple computers can eat your words.Whitebox system $5284
Sun w23100z $8695
Apple Dual G5 2.5 GHz $5570
(configed with 4 GB RAM, larger 160 GB harddrive, Geforce 6800 which spanks the FX3000's ass and combo drive)
Oh that's right, you guys think the dual G5 system was meant for your grandmother to check email and your baby brother to play spongebob squarepants video games.
Get a clue. Good x86 systems with dual processors from reputable companies like Sun (or even whitebox systems) and dual apple systems cost about the same. Let's end at least the price part of apple vs x86 right here and now.
Araemo - Wednesday, October 27, 2004 - link
Wow.. you guys outdid yourselves, especially with the thermal graphs. I also want to say how nice it is to have a couple new reviews every week again. Keep up the good work guys. :)