What does Microsoft need to show at their reveal to interest you?

[WereKitten]

Chris Hu said:

Well it only a rumor so there is really no reliable source but I don't think it will be as hard as you think the orginal XBox and 360 where completely different and the pulled of a 20% compatability with older titles (orginal X-Box Intel based CPU and NVIDIA GPU, 360 IBM based CPU and AMD/ATI GPU).  This time around MS is sticking with AMD/ATI for there GPU I don't think it will be that impossible to achieve a high rate of backwards compatatibility.  PLus based on rumors the Nextbox is a lot more powerful then the 360 3 core vs 8 core and 512MB RAM vs 8GB.

Well, of course it is much more powerful than the 360. Gflops or MHz comparisons mean very little between different architectures, but let's say its something like a factor 3x-4x in practical CPU tasks. For emulating a different architecture, that's generally not enough.

To put things in perspective: by the time 360 and PS3 came out, they were powerful machines by PC standards. They were multi-core/threaded machines with innovative features in CPU or GPU setup. Xbox 360 with its 3 multithreaded cores at 3.2 GHz has probably close to 10 times the CPU power of the Pentium III in the original XBox. And yet, the XBox emulation in the 360 relied on pre-transcoded code you had to download, and the PS3 never got a full software real-time PS2 emulator.

When Durango and PS4 will come out, they will be good bang-for-bucks machines, with more efficient architectures and custom-tailored software and services, but quite under-powered in brute strength terms when compares to mid-range PCs of their time.

Now, if emulating a 360 in software on current PC architecture (because that's what Durango and PS4 basically will be) were easy, or just reasonably doable, there would already be an ecosystem of 360 emulators, at least for high end PCs.

So, if we're talking of pure software emulation here, that seems extremely unlikely to happen. Hardware BC, or at least hardware-assisted emulation is much more likely to be technically feasible: It becomes a problem of costs, engineering, manufacturing yield and policy over the content. And we know very little of MS' business plan with their next machine so that can only be a big question mark.