Why are handhelds so.......weak?

Are you sure they didn't say 80% more work out of 2 SH2's than a single SH2
at the same frequency..
The other quote doesn't really make sense..
 
Well, it can be true the way I heard it, considering the SH-2's aren't really SMP. They can't address memory at the same time under any circumstances.
 
Well a fighter also helps you as Saturn developer as you have two character on screen and you can divide the main work you have to do between the two SH-2s... it is a bit easier IMHO to synchronize the work between the SH-2s...

Other genres make a bit more difficult to divide the processing load between the twin SH-2s...
 
Single SH-2 with 4 KB of Cache and 4-8 KB of SPRAM plus a full speed SCU's DSP ( same speed as the SH-2 ) attached as co-processor ( I think it could have been possible ) would have been better than the set-up Saturn went with, but at the time rushing to beat PSOne's launch and to be competitive in terms of performance might have made this kind of idea very difficult to examine...
 
Wow, and I always believed 3dfx's line that in a perfect dual chip solution, the best you can hope for is about a 5%(something like that) decrease in efficiency, but the saturn got an increase.....
Which also led me to believe the VSA-100 was more powerful than the box said, but 3dfx lied since their chip link up wasn't that efficient. That, and I remember seeing voodoo 4 benchmarks that blew away a voodoo 3, and were more like 3/4th of a voodoo 5's speed and not half, but they may have been bogus benchmarks.
 
fox5, you're comparing processors (not easy to multithread) to graphics chips (insanely easy to multithread - come standard as such).

3dfx's SLI was almost 100% efficient, and resulted in 1 + 1 = 2, in nearly all situations (except in some, like when you're limited by space).

Anyway, I don't know about that AM2 quote, that's what I heard, and I didn't even hear it straight from that person. So oh well...
 
Tagrineth said:
http://steelkiel.tripod.com/vf2flash.jpg
http://steelkiel.tripod.com/vf2noflash.jpg
http://steelkiel.tripod.com/vf2sarah.jpg
http://steelkiel.tripod.com/vf2kage.jpg

First two are an attempt at reproducing your screens as closely as possible.

Third and fourth are close-ups of Sarah and Kage.

Admittedly, the TV's quality isn't very good (and it's over an RF signal), but you can see the shading is much more even.

I'll take pictures from my reverse-projection TV when I can (parents are watching something).

I wonder... with ATi's CATALYST drivers, if I put this All-in-Wonder Rage 128 Pro in my system, would there be any hassles? It'd be nice if I could use that to get some direct screen captures...

I know this is really old, but I just wanted to note something about the crapiness of the VF2 PC port....
It was only that crappy because I had the highlights option turned on. Not sure what it does(besides making the characters brighter) but it makes all the polygon seems very apparent. Well, my fault there.
BTW, I also noticed that some of the moves from shenmue 1 and 2 are in vf2 with virtually the exact same controls, that's kind of cool.
 
gba has the ability to play most if not all SNES type games with little or no porting work, not to mention their HUGE GB game library.
 
YeuEmMaiMai said:
gba has the ability to play most if not all SNES type games with little or no porting work, not to mention their HUGE GB game library.

I'm not sure that's true. The Game Boy Advance runs on the ARM architecture while the SNES runs on the 65816, two instruction sets that have nothing in common. This would force, at least, a recompilation of the source code, or more likely (since pre-PSX era console games were usually coded in assembly, and also since the 65816 didn't lend itself to high-level languages with its absolute lack of GPRs) a full recoding of the entire game. Emulation is also difficult given the the fact the the GBA isn't exactly an order of magnitude stronger than the SNES.

One thing I don't understand is this: The ARM series typically consumes power at about 0.6mW/MHz (or lower). The GBA has a power consumption rating of ~600mW, which means the ARM7 in the GBA (@ 16MHz) comsumes less than 2% of the GBA's power. In devices similiar to the GBA main CPUs are typically clocked 9-25 times higher without noticably lower battery life. If the GBA could have been made much more powerful without lowering its' battery life much, why wasn't it?
 
Back
Top