MIME-Version: 1.0
References: 
 <CAOzEurSW8cNr6TPKsjrstnPfhf4QyQqB4tnPXGGe8N4e_v7Jig@mail.gmail.com>
 <CAN55FZ247JdiT8Sd1SRiyOJxk3Ei=pDCL4kpdP=HqLRjOhKf1Q@mail.gmail.com>
In-Reply-To: 
 <CAN55FZ247JdiT8Sd1SRiyOJxk3Ei=pDCL4kpdP=HqLRjOhKf1Q@mail.gmail.com>
From: Nazir Bilal Yavuz <byavuz81@gmail.com>
Date: Mon, 11 Aug 2025 11:52:25 +0300
Message-ID: 
 <CAN55FZ2AxiwSah7TiQoMB==r=JKT0bOtooCB7ov4xRrGkVmJ1A@mail.gmail.com>
Subject: Re: Speed up COPY FROM text/CSV parsing using SIMD
To: Shinya Kato <shinya11.kato@gmail.com>
Cc: pgsql-hackers@postgresql.org
Content-Type: multipart/mixed; boundary="0000000000009b5194063c130bbd"
Archived-At: 
 <https://www.postgresql.org/message-id/CAN55FZ2AxiwSah7TiQoMB%3D%3Dr%3DJKT0bOtooCB7ov4xRrGkVmJ1A%40mail.gmail.com>
Precedence: bulk

--0000000000009b5194063c130bbd
Content-Type: text/plain; charset="UTF-8"

Hi,

On Thu, 7 Aug 2025 at 14:15, Nazir Bilal Yavuz <byavuz81@gmail.com> wrote:
>
> On Thu, 7 Aug 2025 at 04:49, Shinya Kato <shinya11.kato@gmail.com> wrote:
> >
> > I have implemented SIMD optimization for the COPY FROM (FORMAT {csv,
> > text}) command and observed approximately a 5% performance
> > improvement. Please see the detailed test results below.
>
> Also, I did a benchmark on text format. I created a benchmark for line
> length in a table being from 1 byte to 1 megabyte.The peak improvement
> is line length being 4096 and the improvement is more than 20% [1], I
> saw no regression on your patch.

I did the same benchmark for the CSV format. The peak improvement is
line length being 4096 and the improvement is more than 25% [1]. I saw
a 5% regression on the 1 byte benchmark, there are no other
regressions.

> What do you think about adding SIMD to CopyReadAttributesText() and
> CopyReadAttributesCSV() functions? When I add your SIMD approach to
> CopyReadAttributesText() function, the improvement on the 4096 byte
> line length input [1] goes from 20% to 30%.

I wanted to try using SIMD in CopyReadAttributesCSV() as well. The
improvement on the 4096 byte line length input [1] goes from 25% to
35%, the regression on the 1 byte input is the same.

CopyReadAttributesCSV() changes are attached as feedback v2.

--
Regards,
Nazir Bilal Yavuz
Microsoft

--0000000000009b5194063c130bbd
Content-Type: text/plain; charset="US-ASCII"; name="v2-0001-Feedback.txt"
Content-Disposition: attachment; filename="v2-0001-Feedback.txt"
Content-Transfer-Encoding: base64
Content-ID: <f_me6vgbh10>
X-Attachment-Id: f_me6vgbh10
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--0000000000009b5194063c130bbd--