Macintosh fans have a well-deserved reputation for insane loyalty. But it's expensive to upgrade, especially if your new computer can't run your current stash of software. That's where Rosetta comes in. The emulation module - which uses binary translation technology from Transitive of Los Gatos, California - will enable Macs built with chips from Intel (Apple's new supplier) to run software made for IBM PowerPC chips in older systems. Sounds simple, but no one's ever been able to cross platforms without crippling performance. Such inter-operability would be a big deal if it worked; tech makers profit from locking customers into one system. We asked Transitive's CEO Bob Wiederhold why he thinks Rosetta is different.
WIRED: Previous emulation revolutions fell flat. Why should software makers trust you?
WIEDERHOLD: We realize that we have to overcome skepticism, and we feel like we've made tremendous progress. The people who have seen this technology, who have dived into the details, are hugely impressed. We understand that software vendors are going to do extensive testing. The proof is once you see your application running - a sophisticated application that you never thought could run under translation.
But what does that mean for someone trying to run, say, old Photoshop on an Intel-based Mac?
I can't comment on any work that we've done with Apple, but in general, computational performance is going to be around 80 percent of what you can achieve with a natively compiled application. Graphics and interactive performance will be transparent, meaning you won't be able to tell the difference between a natively compiled application and a translated one. You say Transitive does "binary translation."
How's that different from plain old emulation?
Emulation typically involves converting one instruction at a time. We do something very different. We look at blocks of instructions and convert them to an intermediate representation that allows us to understand the higher-level semantics of the code. This is what allows us to achieve breakthrough performance.
If your code works, does that mean users will no longer be locked into their current systems?
The key problem that we're solving is hardware-software dependency, which causes all kinds of problems in the electronics industry. As our technology becomes ubiquitous, you'll see a lot more applications running on a lot more platforms. That's something customers have wanted for 30 years.
But hardware-software dependency is a basic business model: You have a PC, so you buy Office, so your next machine is another PC. Companies like Dell and Microsoft rely on that cycle.
Rosetta will lower the barriers to getting more software on more platforms. Some companies may be threatened by that, but technology marches forward. The companies that get behind new technologies and find ways to take advantage of them are usually the ones that are successful in the long term. Companies that try to keep new technologies from taking hold in the market are typically the ones that are hurt most.
That sounds like a politely worded threat.
It's not meant as a threat. We're solving an important problem in the industry, and when you do that the various players have to figure out how to adjust their businesses to take advantage of this kind of technology. They do that all the time.
Email hotseat@wiredmag.com.
Transitive CEO Bob Wiederhold
credit PJ Richardson
START
Can Apple Make the Switch?
