Re: Whence the Opterons? - Mailing list pgsql-performance

The DP+DC isn't available yet, from Sun. Only QP+DC is, for which the
bid opens at 38k, that is a bit pricey -:)


-----Original Message-----
From: William Yu [mailto:wyu@talisys.com]
Sent: Monday, May 09, 2005 1:24 PM
To: pgsql-performance@postgresql.org
Subject: Re: [PERFORM] Whence the Opterons?

Unfortunately, Anandtech only used Postgres just a single time in his
benchmarks. And what it did show back then was a huge performance
advantage for the Opteron architecture over Xeon in this case. Where the

fastest Opterons were just 15% faster in MySQL/MSSQL/DB2 than the
fastest Xeons, it was 100%+ faster in Postgres. He probably got rid of
Postgres from his benchmark suite since it favors Opteron too much. As a

general hardware review site, makes senses that he needs to get more
neutral apps in order to get free systems to review and (ahem) ad
dollars.

That being said, I wouldn't get a quad Opteron system anyways now that
the dual core Opterons are available. A DP+DC system would be faster and

cheaper than a pure quad system. Unless of course, I needed a QP+DC for
8-way SMP.






Anjan Dave wrote:
> Wasn't the context switching issue occurring in specific cases only?
>
> I haven't seen any benchmarks for a 50% performance difference.
Neither
> have I seen any benchmarks of pure disk IO performance of specific
> models of Dell vs HP or Sun Opterons.
>
> Thanks,
> Anjan
>
>>EMC you can file an RPQ via your sales contacts to get it approved,
>>though not sure how lengthy/painful that process might be, or if it's
>>gonna be worth it.
>>
>>Read the article devoted to the v40z on anandtech.com.
>>
>>I am also trying to get a quad-Opteron versus the latest quad-XEON
>
> from
>
>>Dell (6850), but it's hard to justify a difference between a 15K dell
>>versus a 30k v40z for a 5-8% performance gain (read the XEON Vs.
>
> Opteron
>
>>Database comparo on anandtech.com)...
>>
>>Thanks,
>>Anjan
>>
>
>
> 15k vs 30k is indeed a big difference. But also realize that Postgres
> has a specific benefit to Opterons versus Xeons. The context switching
> storm happens less on an Opteron for some reason.
>
> I would venture a much greater benefit than 5-8%, more like 10-50%.

---------------------------(end of broadcast)---------------------------
TIP 3: if posting/reading through Usenet, please send an appropriate
      subscribe-nomail command to majordomo@postgresql.org so that your
      message can get through to the mailing list cleanly