Symmetry (someone.delete@this.somewhere.com) on August 17, 2014 6:32 am wrote:
> Ricardo B (ricardo.b.delete@this.xxxxx.xx) on August 16, 2014 8:51 am wrote:
> > To succeed, upcoming ARM server systems need to do those tasks you've just described with
> > better performance/price, performance/power or performance/ToC than x86 server systems.
> >
> > And what we've seen so far were ARM server systems failing at
> > that even for friendly workloads like front-end web servers.
> >
> > Yet, some "pro-ARM" supporters remain adamant in assuming that the upcoming ARM server SoC will succeed.
>
> According to the benchmarks I saw Calexda is actually somewhat better in terms of performance/power
> than the Xeon solution that existed when it was released. The problem was that the difference
> wasn't large enough, and that just didn't justify a large change for most organizations.
>
> http://www.anandtech.com/show/6757/calxedas-arm-server-tested/11
> http://www.anandtech.com/show/6757/calxedas-arm-server-tested/12

Another problem is that it was also more expensive.