I think there is a typo on the first line, no? It should be:
>t1 – 1.00 sec (12M rows/sec ; __ 800MB__ /sec)

>>Wouldn’t using tables with equal row size give a better to compare result?

>This is why I specify MB/sec too
>
>t1 – 1.00 sec (12M rows/sec ; 80MB/sec)
>t1c99 – 1.71 sec (7M rows/sec ; 676MB/sec)
>t100 – 1.77 sec (7M rows/sec ; 653MB/sec)
>t99v1 – 12.36 sec (1M rows/sec ; 93MB/sec)

I see.

You said:
> So there is surely the penalty for dynamic rows, however it is not very significant
> if number of columns is small. For large number of columns dynamic rows become very expensive and you have to watch out.

I fully agree with your conclusion that there is a penalty for “dynamic” rows.
Looking at your numbers we see the result of 653MB/sec versus 93MB/sec.
Which means that a static row is 7 times faster than a dynamic row.

But I’m not sure if the number of columns is of any relevance or if its the row length.
I would assume that a 2 column but longer row should get a huge penalty also.

Could you be so kind and test what result are getting for a table like “t1v99″?
CREATE TABLE `t1v99` (
`t1` tinyint(3) UNSIGNED NOT NULL,
`v99` varchar(99) NOT NULL DEFAULT ’12345678901234567890123456789012345678901234567890123456789012345678901234567890123456789′
) ENGINE=MyISAM DEFAULT CHARSET=latin1

Could we compare this result with “tac99″ please?

Cheers
Gunnar

This is why I specify MB/sec too

The sizes are only different when I want them to be. t1c99, t100 and t99v1 all have same row length in MyISAM and yet they have different scan speed. Other tables are of different row length and mainly provided for comparison so we can both see how number and type of columns affect scan speed.

Thanks for the benchmark.
I wonder if the test is comparing apples to oranges by accident.

The 4 test tables that you use have very different sizes.
How much influence has the different size of the test tables on the results?
Wouldn’t using tables with equal row size give a better to compare result?

Take care

Sorry about captcha. Looks like some other glitch. What kind of error message did you get if any ?

Thanks corrected.