Client Speed Comparisons
To satisify the curiosity of users that want to see the performance difference between different processors and architectures, we provide these data tabulation pages listing client processing rates that have been reported to us by other users.
The recommended way for obtaining reliable computation rates from your client's benchark is to run the "long" client benchmark 3 times, and taking the best rate of the three. Also, when possible, it is best to run the clients in as similar an environment as possible. Not all UNIX/multiuser machines have that luxury, so there may be some fluctuation between measured and actual speed.
To view some of our statistics, click on any of the following links:
| All | DES | CSC | OGR | OGRNG | |||
|---|---|---|---|---|---|---|---|
| x86 processors: Intel/AMD/Cyrix... | View | View | View | View | View | View | View |
| Motorola 68k processors | View | View | View | View | View | View | View |
| Power PC Processors | View | View | View | View | View | View | View |
| RISC Processors: Sparc/HP/Alpha/MIPS | View | View | View | View | View | View | View |
| GPU Processors: CUDA/Stream | View | View | — | — | — | — | — |
Other ways to view the data...
- Frequently Asked Questions about the speeds database
- Execute a custom query and specify the platform, architecture, or contest type that you want to view data for.
- Multi-Processor machines: regardless of CPU architecture.
- View a custom graph that charts the client speed versus processor MHz speed for the platforms that you choose.
How to contribute to the speeds database...
- Submit new keyrates to our statistics database
- Submit a bug request to correct an issue with the Client Speeds Database pages. (Note that the recommended way of reporting bad entries is to click the new "Report as error" links instead of using Bugzilla now.)
- Currently pending issues and bugs relating to the Client Speeds Database pages.