[
Lists Home |
Date Index |
Thread Index
]
On Wednesday 25 May 2005 06:02, Ken North wrote:
>
> In other words, can we quantify how much benefit we receive from pure
> hardware acceleration when processing content with one of these appliances
> -- before adding the effect of the security layer?
>
I don't think parsing/validation offload has come up directly in a independent
benchmark yet, although there are gains to be had. In the Sarvega case, there
is an internal parsing/validation benchmark available from the web site that
might be some use, I am sure the other vendors have similar information
available.
One area to be careful of is the context in which the figures are generated.
For example, the performance of parser/validation off-load from Java can be
more dependant on Java<->appliance glue issues than the appliance performance
itself. Needless to say, figures without specifying any context can be very
misleading.
Regards,
Kev.
|
