| From: | Pierre Frédéric Caillaud <lists(at)peufeu(dot)com> |
|---|---|
| To: | "Scott Marlowe" <scott(dot)marlowe(at)gmail(dot)com>, "Scott Carey" <scott(at)richrelevance(dot)com> |
| Cc: | "Josh Kupershmidt" <schmiddy(at)gmail(dot)com>, "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org> |
| Subject: | Re: PG-related ACM Article: "The Pathologies of Big Data" |
| Date: | 2009-08-08 09:26:34 |
| Message-ID: | op.uyb66kq6cke6l8@soyouz |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
>> I don't see how on any recent hardware, random access to RAM is slower
>> than
>> sequential from disk. RAM access, random or not, is measured in
>> GB/sec...
>
> I don't think anybody's arguing that.
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795&p=5
These guys mention about 50 ns memory latency ; this would translate into
20 million memory "seeks" per second, which is in the same ballpark as the
numbers given by the article...
If you count 10GB/s bandwidth, 50 ns is the time to fetch 500 bytes.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2009-08-08 16:02:06 | Re: [PERFORM] BUG #4919: CREATE USER command slows down system performance |
| Previous Message | Fizu | 2009-08-08 06:02:47 | ORDER BY ... LIMIT and JOIN |