Virus TI2 DSP power in relation to todays Intel chips

[izonin]

It's 2x200MHz DSPs, but equals a Core2 in processing power. Parallel processing is much faster than the versatile CPUs.

[EvilDragon]

So a modern i7 still kills them real hard (the 6-core-plus-hyperthreading i3770K especially). Core2 is quite old now.

[kmonkey]

So a modern i7 still kills them real hard (the 6-core-plus-hyperthreading i3770K especially). Core2 is quite old now.

i3770K isn't six core and if something have more brute force it doesn't act as suddenly better. After all TI is your practical real life example. It doesn't have latest i7 inside yet is worldwide instrument accepted by many amateurs and pros.

In fact even old viruses are still used extensively.

Quality of an algorithm my friend. Not mhz of speed...

[urosh]

Quality of an algorithm my friend. Not mhz of speed...

Yes, but this thread is about speed. Old Lexicon reverbs had pathetic processing muscles under the hood and yet they are still unsurpassed, but that is different topic. And if TI uses those Freescale processors EvilDragon mentioned than no, it isn't equal or even close to Core2 for most of MI applications.
As I said, apple to apple comparasion would be same algo on both systems. Now, since there is no native Virus here is second best thing: there was/is Powercore version of Virus and there are native ports of some Powercore plugins, so it shouldn't be too difficult to get some realistic numbers that would estimate how much would one Virus voice tax modern x86.

[ericj23]

Quality of an algorithm my friend. Not mhz of speed...

Yes, but this thread is about speed. Old Lexicon reverbs had pathetic processing muscles under the hood and yet they are still unsurpassed, but that is different topic. And if TI uses those Freescale processors EvilDragon mentioned than no, it isn't equal or even close to Core2 for most of MI applications.
As I said, apple to apple comparasion would be same algo on both systems. Now, since there is no native Virus here is second best thing: there was/is Powercore version of Virus and there are native ports of some Powercore plugins, so it shouldn't be too difficult to get some realistic numbers that would estimate how much would one Virus voice tax modern x86.

Doesn't really work - because of the different ways processors work you need to reprogram your algorithms for x86 or freescale - but some algos run very well on one and not on the other because of the advantages and disadvantages . Basically you would need to run lots of tests on various plug-ins in order to arrive at an accurate estimate of how the CPUs scale.

I don't think there are that many plug-ins to do this on

Having said that I am pretty convinced that the Virus TI is not even vaguely powerful by today's standards - I have to ask why it matters

I am willing to bet that we will not be getting a native virus anytime soon

[Cyforce]

The difference/comparision will not work. Also if the DSP runs on a much lower speed and RAM - code's for native plugins working different as codes for DSP platform.

It's like comparing apples and mangos

[Gamma-UT]

One thing to bear in mind is that in custom hardware, the algorithm's access to the processor is predictable. It's not getting constantly swapped in and out to make way for other processes all the time - that swapping process wastes a lot of CPU cycles.

Even the environment inside a PoCo I would imagine is much more predictable than in a native DAW as I would guess that it uses the same approach as ProTools HD where the algorithms get X many slices of DSP time at regular intervals. Contrast this with a DAW that says "I need x samples right now, hand them over", where x is a highly variable number.

That is likely to complicate porting. However, I would guess that the reason you haven't seen a native Virus is that it's hard to pirate the software a custom box or a PoCo as you need an identical box to run the software. Unless Access's hardware sales begin to dry up, there isn't really a compelling reason for them to port to native. Ditto Roland, Kurzweil, Yamaha et al.

[UltraJv]

One thing to bear in mind is that in custom hardware, the algorithm's access to the processor is predictable. It's not getting constantly swapped in and out to make way for other processes all the time - that swapping process wastes a lot of CPU cycles.

Even the environment inside a PoCo I would imagine is much more predictable than in a native DAW as I would guess that it uses the same approach as ProTools HD where the algorithms get X many slices of DSP time at regular intervals. Contrast this with a DAW that says "I need x samples right now, hand them over", where x is a highly variable number.

That is likely to complicate porting. However, I would guess that the reason you haven't seen a native Virus is that it's hard to pirate the software a custom box or a PoCo as you need an identical box to run the software. Unless Access's hardware sales begin to dry up, there isn't really a compelling reason for them to port to native. Ditto Roland, Kurzweil, Yamaha et al.

+1

Access have no intention of porting to X86. They, like the other big players dont see any money in native plugs, they make thier money on hardware. As for processing power, can Virus run Cubase or Logic as well as other plugs? Can it run Diva? Of course not. As mentioned earlier, the only thing thats special about Virus is its algorithms. They would easily run on any modern x86 hardware if ported.

[EvilDragon]

So a modern i7 still kills them real hard (the 6-core-plus-hyperthreading i3770K especially). Core2 is quite old now.

i3770K isn't six core

Sorry, I meand i3830K.