Difference between revisions of "Distributed.net"
Jump to navigation
Jump to search
(→Notes) |
|||
| Line 83: | Line 83: | ||
</pre> | </pre> | ||
In addition, this run was done with hyperthreading enabled, so it is really not that valid. It is obvious that the cores were fighting for the same silicon a lot. | In addition, this run was done with hyperthreading enabled, so it is really not that valid. It is obvious that the cores were fighting for the same silicon a lot. | ||
| + | |||
4. I find it hard to believe that 12 Intel procs at 3.0 Ghz cannot beat the same number of AMD Opteron 4184 cores running at 2.8Ghz. Trying a newer client to see if this is the problem. | 4. I find it hard to believe that 12 Intel procs at 3.0 Ghz cannot beat the same number of AMD Opteron 4184 cores running at 2.8Ghz. Trying a newer client to see if this is the problem. | ||
Revision as of 11:42, 20 February 2012
Tested speeds
x86/ELF Client
| Processor Type | Freq | Cores | Sockets | Tot Cores | Key Rate (keys/sec) |
|---|---|---|---|---|---|
| Intel P4 | 3.2 | 1 | 1 | 1 | 6,043,121 |
| Intel P4 | 2.4 | 1 | 1 | 1 | 4,413,332 |
| Intex Xeon E5335 | 2.0 | 4 | 2 | 8 | 52.94M |
| AMD Opteron 244 | 1.8 | 1 | 2 | 2 | 15.39M |
AMD64 CUDA 3.1 Client
| Processor Type | Freq | Cores | Sockets | Tot Cores | Key Rate (keys/sec) | Client Ver | Notes |
|---|---|---|---|---|---|---|---|
| nVidia Tesla T10 | 1.3Ghz | 240 | 2 | 480 | 511.14M | 517 | |
| nVidia Tesla M2070 | 1.15Ghz | 448 | 8 | 3584 | 1504.08M | 517 | |
| nVidia Tesla M2070 | 1.15Ghz | 448 | 1 | 448 | 417.85M | 517 |
AMD64 Client
| Processor Type | Freq | Cores | Sockets | Tot Cores | Key Rate (keys/sec) | Client Ver | Notes |
|---|---|---|---|---|---|---|---|
| AMD Opteron 6220 | 3.0 | 8 | 2 | 16 | 118.06M | 509 | 4 |
| AMD Opteron 4122 | 2.2 | 4 | 1 | 4 | 34.34M | 518 | |
| AMD Opteron 4184 | 2.8 | 6 | 2 | 12 | 130.49M | 517 | |
| AMD Opteron 4184 | 2.8 | 6 | 1 | 6 | 65.80M | 517 | |
| AMD Opteron 6174 | 2.2 | 12 | 4 | 48 | 410.26M | 517 | 2 |
| AMD Opteron 6174 | 2.2 | 12 | 2 | 24 | 198.32M | 509 | 1 |
| AMD Opteron 8435 | 2.6 | 6 | 4 | 24 | 139.55M | 509 | |
| AMD Opteron 2427 | 2.2 | 6 | 2 | 12 | 75.24M | 509 | |
| AMD Opteron 2456 | 2.3 | 4 | 2 | 8 | 72.86M | 509 | |
| AMD Opteron 2350 | 2.0 | 4 | 2 | 8 | 52.98M | 509 | |
| AMD Opteron 2214 | 2.2 | 2 | 2 | 4 | 26.42M | 509 | |
| AMD Opteron 875 | 2.2 | 2 | 4 | 8 | 53.65M | 509 | |
| AMD Opteron 275 | 2.2 | 2 | 2 | 4 | 34.73M | 509 | |
| AMD Opteron 250 | 2.4 | 1 | 2 | 2 | 18.80M | 509 | |
| AMD Phenom II X4 945 | 3.0 | 4 | 1 | 4 | 41.03M | 518 | |
| Intel Xeon X5675 | 3.07 | 6 | 2 | 12 | |||
| Intel Xeon E5462 | 2.8 | 4 | 2 | 8 | 67.05M | 509 | |
| Intel Xeon E5620 | 2.4 | 4 | 2 | 8 | 50.10M | 517 | |
| Intel Xeon X5660 | 2.8 | 6 | 2 | 12 | 88.52M | 517 | |
| Intel Xeon E5-2680 | 2.7 | 2 | 16 | 32 | 97.44M | 518 | 3 |
| Intel Xeon E5-2680 | 2.7 | 2 | 8 | 16 | 114.40M | 518 |
Notes
1. As a note, dnetc was not able to detect the processor properly because it is a preproduction:
[Feb 19 16:35:07 UTC] Automatic processor type detection did not
recognize the processor (tag: "40010001")
2. Updated client to version 517. 3. Could not detect processor properly:
[Dec 08 10:30:40 UTC] Automatic processor type detection did not
recognize the processor (tag: "100062D6")
In addition, this run was done with hyperthreading enabled, so it is really not that valid. It is obvious that the cores were fighting for the same silicon a lot.
4. I find it hard to believe that 12 Intel procs at 3.0 Ghz cannot beat the same number of AMD Opteron 4184 cores running at 2.8Ghz. Trying a newer client to see if this is the problem.
