There’s a review of Via’s C3 800MHz processor here.
Included in the review are some SysMark2000 benchmarks.
I thought it would be interesting to compare it to something more along the lines of what you’re likely to have. So I took some old benchmarks of a TBird running at 9X150, or 1350Mhz, and compared them.
Here’s how the C3 did against the TBird (measured as a percentage of the TBird’s score):
Sysmark 2000: 42%
Now Lost Circuits did the job right and provided all the numbers that went into that 42%. As you’ll see,
that 42$ covers a lot of ground. That’s certainly not their fault, but it shows what a single number can hide:
| Application | Via C3 Performance as % of TBird Perf. |
| Bryce 4 | 29% |
| CorelDraw(TM) 9 | 31% |
| Elastic Reality ® 3.1 | 26% |
| Excel 2000 | 52% |
| NaturallySpeaking ® Pref 4.0 | 42% |
| Netscape ® Communicator | 60% |
| Paradox ® 9.0 | 69% |
| Photoshop ® 5.5 | 39% |
| PowerPoint ® 2000 | 53% |
| Premiere ® 5.1 | 30% |
| Word 2000 | 95% |
| Windows Media Encoder 4.0 | 28% |
Obviously, this isn’t a CPU for graphics or MP3 ripping. But what about Grandma? If she:
this might do rather nicely. 🙂
I took a look around, and you can get an ECS mobo with (granted, poor) integrated video and the 733Mhz version of this chip for a hundred bucks.
Even if you wouldn’t touch this with a ten-foot-pole even for Grandma, this does illustrate how a single benchmark number can hide more than it reveals.
This is something you should keep in mind and ask for when you’re making more meaningful comparisons, like between XPs and PIVs.
Be the first to comment