Wednesday December 14, 2005
Wednesday December 14, 2005
iGenCPU results on UltraSPARC T1 (T2000) and UltraSPARC IV+ (V490+)
This is an update on my previous entry, adding the UltraSPARC T1 CoolThread server T2000 and the V490+.The throughput increase between the UltraSPARC IV+ and the UltraSPARC IV is fairly proportional to the clock frequency.No surprise here. The simplicity of this microbenchmark does not allow taking benfits of many of the UltraSPARC IV+ innovations.
Regarding the T2000, please note that I am not afraid to publish the results. This is not a marketing blog...Indeed, this benchmark is not recommended for the UltraSPARC T1 due to the fact that 25% of the instructions are floating point operations (see two previous blog entries).
It does not prevent us to collect the data. Please remember that we want our customers to run the right platform for their workload. So, if your workload is generating more than 2% of floating point operations, the UltraSPARC T1 is probably not what you should choose....
Note : The column threads is the number of threads used to observe the highest throughput with a response time (RT) less than 100 ms.
| Processor | Frequency | # CPU | # Cores | Ram | OS | Threads | Fractals/s | RT (ms) |
| Intel XEON | 3 Ghz | 2 | 4 (HT) | 4 GB | S10 03/05 | 3 | 31.16 | 96.26 |
| Sun UltraSPARCIIIi | 1.2Ghz | 4 | 4 | 8 GB | S10 03/05 | 4 | 53.01 | 75.4 |
| AMD OPTERON | 2.4Ghz | 4 | 4 | 8 GB | S10 HW2 | 5 | 90.18 | 55.4 |
| Sun UltraSPARC IV | 1.2Ghz | 4 | 8 | 8 GB | S10 03/05 | 8 | 98.88 | 80.8 |
| Sun UltraSPARC IV+ |
1.5Ghx |
4 |
8 |
8 GB |
S10 HW1 |
8 |
123.48 |
78.36 |
| Sun UltraSPARC T1 |
1.2Ghz |
1 |
8 |
8 GB |
S10 HW2 |
9 |
18.62 |
93.08 |
Let me know your thoughts....
Next, I will publish a description of the iGenRAM 1.6 and related benchmark results. We will see that the T1 is pretty good at this....
Dec 14 2005, 02:24:01 PM PST Permalink
Comments:
Benoit,
A few nits..
Is it possible to have all systems running the same OS release. Does the HW1/HW2 refer to platform specific patches which are absolutely essential to get it running properly?
Same RAM for the Xeon i.e 8GB
Same CPU for the Xeon i.e 4-way
Is it possible that you are not comparing apples vs apples? :-) We are trying to make everything the same but the Niagara T1 stays at 1 CPU!
Do you think that benchmarking similar cost systems is essential or similar hardware configuration is essential? Your input is much appreciated.
I just read Walter Bays benchmark guidelines for Sun Bloggers post here.
Thanks in advance
A few nits..
Is it possible to have all systems running the same OS release. Does the HW1/HW2 refer to platform specific patches which are absolutely essential to get it running properly?
Same RAM for the Xeon i.e 8GB
Same CPU for the Xeon i.e 4-way
Is it possible that you are not comparing apples vs apples? :-) We are trying to make everything the same but the Niagara T1 stays at 1 CPU!
Do you think that benchmarking similar cost systems is essential or similar hardware configuration is essential? Your input is much appreciated.
I just read Walter Bays benchmark guidelines for Sun Bloggers post here.
Thanks in advance
Posted by Amit Kulkarni on December 14, 2005 at 06:12 PM PST #
Amit,
I guess RAM size does not matter on a microbenchmark calculating Mandelbrot fractals.
Posted by Florian Laws on December 15, 2005 at 12:59 AM PST #
Post a Comment:
Comments are closed for this entry.