By: Michael S (already5chosen.delete@this.yahoo.com), November 10, 2008 8:58 am
Room: Moderated Discussions
Henrik S (henrik.stahl@oracle.nospam.com.nothanks) on 11/10/08 wrote:
---------------------------
>Michael S (already5chosen@yahoo.com) on 11/10/08 wrote:
>---------------------------
>>Henrik S (henrik.stahl@nospam.oracle.nothanks.com) on 11/10/08 wrote:
>>---------------------------
>>>The best L5420 score with JRockit (from NEC) is *much* higher: 1010 ssj_ops/watt
>>>http://www.spec.org/power_ssj2008/results/res2008q3/power_ssj2008-20080714-00064.html
>>>
>>>This is on Windows (will yield a slightly higher score), with different BIOS settings
>>>and with 2x4 GB DIMMs instead of 8x2.
>>>
>>>Interesting enough, the NEC submission has a higher peak score (288502 vs 267804)
>>>despite the fact that it is uses a smaller Java heap and the slower (?) 4 GB DIMMs.
>>>This is counterintiutive to me and makes me suspect that the AMD Xeon submission
>>>is on a machine with a less optimized BIOS.
>>>
>>
>>Did you find a specification for NEC ECO CENTER in English?
>
>No.
>
>>I couldn't.
>>If, like Supermicro 6015C-MTB, it is based on San Clemente chipset then Blackford/Seaburg
>>based experience with regard to which memory is faster could not apply. It's possible
>>that a couple of dual-rank DIMMs is the fastest memory option for that chipset.
>
>Supermicro result here for reference:
>http://www.spec.org/power_ssj2008/results/res2008q4/power_ssj2008-20080917-00079.html
>>
>
>-- Henrik
There is rather small 3.3% difference in favor of NEC.
Could it be explained by better tuning (different compiler options, in particular, -XXlargePages, bigger initial heap) and smarter configuration (disabling unused NIC)?
In ECO CENTER product announcement letter NEC claimed especially efficient power supply unit. May be that's what helped them to match Supermicro power scores despite of inherent disadvantage of 780W PSU competing against 520W PSU.
---------------------------
>Michael S (already5chosen@yahoo.com) on 11/10/08 wrote:
>---------------------------
>>Henrik S (henrik.stahl@nospam.oracle.nothanks.com) on 11/10/08 wrote:
>>---------------------------
>>>The best L5420 score with JRockit (from NEC) is *much* higher: 1010 ssj_ops/watt
>>>http://www.spec.org/power_ssj2008/results/res2008q3/power_ssj2008-20080714-00064.html
>>>
>>>This is on Windows (will yield a slightly higher score), with different BIOS settings
>>>and with 2x4 GB DIMMs instead of 8x2.
>>>
>>>Interesting enough, the NEC submission has a higher peak score (288502 vs 267804)
>>>despite the fact that it is uses a smaller Java heap and the slower (?) 4 GB DIMMs.
>>>This is counterintiutive to me and makes me suspect that the AMD Xeon submission
>>>is on a machine with a less optimized BIOS.
>>>
>>
>>Did you find a specification for NEC ECO CENTER in English?
>
>No.
>
>>I couldn't.
>>If, like Supermicro 6015C-MTB, it is based on San Clemente chipset then Blackford/Seaburg
>>based experience with regard to which memory is faster could not apply. It's possible
>>that a couple of dual-rank DIMMs is the fastest memory option for that chipset.
>
>Supermicro result here for reference:
>http://www.spec.org/power_ssj2008/results/res2008q4/power_ssj2008-20080917-00079.html
>>
>
>-- Henrik
There is rather small 3.3% difference in favor of NEC.
Could it be explained by better tuning (different compiler options, in particular, -XXlargePages, bigger initial heap) and smarter configuration (disabling unused NIC)?
In ECO CENTER product announcement letter NEC claimed especially efficient power supply unit. May be that's what helped them to match Supermicro power scores despite of inherent disadvantage of 780W PSU competing against 520W PSU.
