log in | register | forums
Show:
Go:
Forums
Username:

Password:

User accounts
Register new account
Forgot password
Forum stats
List of members
Search the forums

Advanced search
Recent discussions
- WROCC Newsletter Volume 41:11 reviewed (News:)
- WROCC March 2024 meeting o... Hughes and Peter Richmond (News:1)
- Rougol March 2024 meeting on monday with Bernard Boase (News:)
- Drag'n'Drop 13i2 edition reviewed (News:)
- South-West Show 2024 talks (News:4)
- February 2024 News Summary (News:1)
- Next developer fireside chat (News:)
- DDE31d released (News:)
- South-West Show 2024 Report (News:)
- South-West Show 2024 in pictures (News:)
Related articles
- Native versus emulation for running RISC OS in 2017 (Part 1)
- ABug provide more interesting retro talks to pass the time this Christmas
- Pass the time this Christmas with a selection of RISC OS and BBC Micro talks
- Native versus emulation in 2016 (Part 3)
- First Impressions of RComp's TiMachine
- Newsround
- RISC OS - the week in comments
- Wakefield 2006 show report
- Wakey Wakey, it's show time again!
- RISC OS on TV take 2
Latest postings RSS Feeds
RSS 2.0 | 1.0 | 0.9
Atom 0.3
Misc RDF | CDF
 
View on Mastodon
@www.iconbar.com@rss-parrot.net
Site Search
 
Article archives
The Icon Bar: News and features: Native versus emulation in 2016 (Part 2)
 

Native versus emulation in 2016 (Part 2)

Posted by Mark Stephens on 09:47, 16/12/2016 | , ,
 
In part 1 we outline our plans to pit a Titanium (running RISC OS 5) against VirtualRPC (running RISC OS 6) on the latest MacBookPro laptop. Here are the scores, Titanium first each time.
 
Titanium versus VirtualRPC
 
Processor - Looped instructions (cache)
3762909 2115%, 524989 295%
Memory - Multiple register transfer
27373 16896%, 9184 5669%
Rectangle Copy - Graphics acceleration test
3253 1344%, 3067 1267%
Icon Plotting - 16 colour sprite with mask
42279 2113%, 26035 1301%
Draw Path - Stroke narrow line
11622 745%, 5337 342%
Draw Fill - Plot filled shape
17204 1179%, 6021 412%
HD Read - Block load 8MB file
115924 3887%, 608316 20399%
HD Write - Block save 8MB file
84216 2769%, 530070 17430%
FS Read - Byte stream file in
2493 1204%, 2634 1272%
FS Write - Byte stream file out
2553 1329%, 1170 609%
 
The 2.9gig Intel processor is not quick enough as an emulator to outperform a 1.5gig ARM chip on the Titanium but they said the new Macs have very fast SSD drives and that gives the Emulator an edge on block filesystem operations. You can compare these figures with those on Chris Hall's website (and try the tests yourself).
 
Does RISC OS performance on the MacBookPro vary if you run fullscreen or from battery?
 
A second question I posed last time was whether there would be any changes. Let us run it 3 times (in a window, in fullscreen mode and in fullscreen mode unplugged to see)...
 
Processor - Looped instructions (cache)
525238 295%, 522668 293%, 522822 293%
Memory - Multiple register transfer
9090 5611%, 9370 5783%, 9160 5654%
Rectangle Copy - Graphics acceleration test
2947 1217%, 2913 1203%, 2899 1197%
Icon Plotting - 16 colour sprite with mask
20316 1015%, 20308 1015%, 24903 1245%
Draw Path - Stroke narrow line
5396 345%, 5531 354%, 5457 349%
Draw Fill - Plot filled shape
5994 410%, 6098 417%, 6045 414%
HD Read - Block load 8MB file
592095 19855%, 559651 18767%, 567762 19039%
HD Write - Block save 8MB file
481882 15846%, 527207 17336%, 522039 17166%
FS Read - Byte stream file in
2593 1252%, 2657 1283%, 2653 1281%
FS Write - Byte stream file out
1141 594%, 1168 608%, 1192 620%
 
So no notable differences between modes.
 
Conclusions
This article was partly intended as a bit of Christmas entertainment and the advantages will vary with your exact usage or requirements (the Mac laptop does use a lot more power for example, whereas the VirtualRPC solutions does provide a 'free' Apple Mac as part of the package). And you could always run Jeffrey Lee's excellent VNC server with the free VNC viewer built into the Mac as another combination.
 
Is this a fair test? Mazzeo's Law says that the answer to any big question is 'It all depends'. It will not give a true answer for every single use case but I would argue that it is a 'valid' test in that it makes a reasonable and repeatable comparison. The Acorn Emulator offers several different chip emulations (700,7500, StrongArm) and may well be better suited to colour modes with less than 16M colours. Stay tuned for Part 3 which will see if these result in a faster Acorn experience (or feel free to test yourself)...
 
My main conclusion is that both my systems offer a very viable RISC OS solution which allows me to use my favourite operating system at home (my Titanium) or on the move (my laptop).
 
And in the future...
I generally replace my work Mac every 2-3 years and the rumour mills like to speculate that Apple may move to ARM chips in the future. At the moment, Apple uses Intel chips, and you can currently use a program called BootCamp to intall Windows directly onto Apple Laptops. Commerial software from VMware, and Parallels, and the free VirtualBox allow you to create virtual machines using the Intel hardware. This is much faster as it does not have to emulate another chip. Running RISC OS on a high end ARM laptop with a solution like this would be very appealing....
 
  Native versus emulation in 2016 (Part 2)
  Bucksboy (18:09 18/12/2016)
  markee174 (17:32 19/12/2016)
    Bucksboy (18:22 19/12/2016)
 
George Greenfield Message #123969, posted by Bucksboy at 18:09, 18/12/2016
Member
Posts: 89
Thanks for that, Mark: any chance of some 'real world' comparisons?
  ^[ Log in to reply ]
 
Mark Stephens Message #123970, posted by markee174 at 17:32, 19/12/2016, in reply to message #123969
Does all the
work around here

Posts: 142
Do you have any suggestions?
  ^[ Log in to reply ]
 
George Greenfield Message #123971, posted by Bucksboy at 18:22, 19/12/2016, in reply to message #123970
Member
Posts: 89
As you probably gathered from my comments to part one, my interest is in image manipulation, and I have found time to rotate a large (52 MB ) sprite or tiff to 45deg using Photodesk with a large* cache set - involving resizing as well as rotation - to be a good test of system speed. But Andrew Rawnsley mentioned Datapower file processing, and no doubt other commenters to pt. one will have their own favourites. The ideal test should call equally on processor, memory and disc access speed IMO.
*Around 120,000k for the file size mentioned.
  ^[ Log in to reply ]
 

The Icon Bar: News and features: Native versus emulation in 2016 (Part 2)