[Rpcemu] RPCEmu 0.8.11
george.greenfield at tiscali.co.uk
Fri Oct 25 02:56:14 PDT 2013
In message <1382637922.58886.YahooMailNeo at web172302.mail.ir2.yahoo.com
Tom Walker <tommowalker at yahoo.co.uk> wrote:
> On Thursday, 24 October 2013, 13:36, george greenfield
> <george.greenfield at tiscali.co.uk> wrote:
>>> There is the oft repeated story that !Desk_Bogo favours VRPC, but
>>> Dhrystones indicate rpcemu is faster.
>>Indeed: that suspicion is supported by Chris Hall's benchmarks here
>>where VRPC on a 2668MHz PC gives only 175% compared to a bog-standard
>>S/ARM RPC, i.e. 1524MHz = 100%; the corresponding figure for RPCEmu
>>here is approx. 1000MHz = 100%. Obviously that assumes a similar
>>processor type in the 2668MHz machine, which may not be the case.
> The frequently quoted 2668 MHz machine is most likely a Pentium 4. The
> Core i7 quoted for the RPCemu results will be vastly quicker
> clock-for-clock, therefore comparing performance across the different
> emulators based solely on clock speeds of these two very different
> machines is not very meaningful.
Fair enough, but the point is not whether exact measurements can be
made, but whether a trend is revealed. RPCEmu0.8.8/4.02 Recompiler on
my elderly Celeron 1500MHz XP laptop manages 112%, which is still
relatively better than the VRPC result on the 2668MHz machine (not
allowing for the >5% speed-up on later versions of RPCEmu), so the
suspicion that VRPC runs slower remains intact. But granted, until
someone with time on their hands /and/ both emulators on the same PC
runs ROMark on both we're never going to know for sure.
More information about the Rpcemu