Microsoft leaks details on Xbox Next

[cybamerc]

But, surely, S3TC wasn't around at the time of the design of the GS chip in the PS2.

The Savage3D came out in 1998.

 

[Simon F]

But, surely, S3TC wasn't around at the time of the design of the GS chip in the PS2.

The Savage3D came out in 1998.

But how long do you think it takes to spec', synthesise, layout, manufacture, and debug a chip?

 

[cybamerc]

Simon F:

> But how long do you think it takes to spec', synthesise, layout,
> manufacture, and debug a chip?

I dunno. Nintendo licensed it in 1999 and had a product out two years later.

 

[Dio]

Surely, S3TC wasn't around at the time of the design of the GS chip in the PS2.

If memory serves me correctly, Savage3D had been shipping for close to a year when I heard the GS die size (I remember a discussion along the lines of 'How big???'). You are right that given the design cycle time it might not have been possible to add S3TC in time if the first they knew was at Savage3D ship time.

However, I would guess that S3 contacted Sony about licensing S3TC substantially before Savage3D shipped, though. They certainly tried to sell it to everyone else!

 

[London Geezer]

However, I would guess that S3 contacted Sony about licensing S3TC substantially before Savage3D shipped, though. They certainly tried to sell it to everyone else!

Can you blame them? It's probably the only good thing they've managed to invent EVER...

 

[maskrider]

Re: ..

Physics calculation is not dot product intensive.

Just happen to read an Intel Prescott article in anadtech.com, something interesting from Tim

In response to developer requests Intel has included the following instructions for 3D programs (e.g. games): haddps, hsubps, haddpd, hsubpd. Intel told us that developers are more than happy with these instructions, but just to make sure we asked our good friend Tim Sweeney – Founder and Lead Developer of Epic Games Inc (the creators of Unreal, Unreal Tournament, Unreal Tournament 2003 and 2004). Here’s what he had to say:

Most 3D programmers been requesting a dot product instruction (similar to the shader assembly language dp4 instruction) ever since the first SSE spec was sent around, and the HADDP is piece of a dot product operation: a pmul followed by two haddp's is a dot product.

This isn't exactly the instruction developers have been asking for, but it allows for performing a dot product in fewer instructions than was possible in the previous SSE versions. Intel's approach with HADDP and most of SSE in general is more rigorous than the shader assembly language instructions. For example, HADDP is precisely defined relative to the IEEE 754 floating-point spec, whereas dp4 leaves undefined the order of addition and the rounding points of the components additions, so different hardware implementing dp4 might return different results for the same operation, whereas that can't happen with HADDP.

As far as where these instructions are used, Tim had the following to say:

Dot products are a fundamental operation in any sort of 3D programming scenario, such as BSP traversal, view frustum tests, etc. So it's going to be a measurable performance component of any CPU algorithm doing scene traversal, collision detection, etc.

The HSUBP ops are just HADDP ops with the second argument's sign reversed (sign-reversal is a free operation on floating-point values). It's natural to support a subtract operation wherever one supports an add.

So the instructions are useful and will lead to performance improvements in games that do take advantage of them down the road. The instructions aren’t everything developers have wanted, but it’s good to see that Intel is paying attention to the game development community, which is something they have done a poor job of doing in the past.

From http://www.anandtech.com/cpu/showdoc.html?i=1956&p=9

 

[Dio]

DXT is half the size(against 8bit) OR way better quality(against 4bit), but not both at the same time.
8bit with a GOOD quantizator easily matches the quality of DXTC on average.

I would strongly disagree with 'easily matches' .

Both have good cases and bad cases, and I would argue that for the type of textures used in games DXTC is more usually superior (as well as not requiring a 1K-per-texture overhead).

However, I'm diplomatic enough to agree that I should revise my statement to 'half the size at typically similar image quality'.

 

[chapban.]

OT:

I think VQ is gooooood. I meana, SA1/2 textures still stand out against anything i see on the PS2. BUT Faf/nAo is right, whats hurting PS2 most is the mipmapping....URGH! MY EYES MY EEEEYES!!!!!! :?

THEN the 2nd thing is the low res output...640x448...no progressive scan...everything just look BLUR and PIXEL-VISIBLEABLE...

 

[Dio]

Can you blame them? It's probably the only good thing they've managed to invent EVER...

Savage3D was far better than most people give it credit for. It's a pity that there weren't that many fast SGRAM boards sold (they were substantially faster than anything else at the time).

 

[bbot]

Which is more expensive: a 10 gig hdd or 256MB of flash memory?

 

[Qroach]

I think the flash memory is much more expensive.

 

[chapban.]

I think the 10gig be more EX. No one is going sell 10gig by 2005, it be have to be specially made for Xbox2. Special = more moneeee!

AS i said, if Xbox2 really kick off the HDD at default, there is no stopping MS for a Live2 starter kit, includes a HDD, and that be more than 10gig!

 

[Guden Oden]

BUT Faf/nAo is right, whats hurting PS2 most is the mipmapping....URGH! MY EYES MY EEEEYES!!!!!! :?

You know what hurts MY eyes?

THIS:

PIXEL-VISIBLEABLE

What IS that? It's not even a word! You're American, right? So why the F**K can't you speak English instead of all that immature scriptkiddy JeffK "teh" cr*p I always see you fill your posts with?

 

[London Geezer]

Guden, u have a PM

 

[chapban.]

PIXEL-VISIBLE = seems to me the pixels in PS2 are more larger, visible than DC/Xbox. Possibly due to the lack of progressive scan and the low res output.

 

[London Geezer]

PIXEL-VISIBLE = seems to me the pixels in PS2 are more larger, visible than DC/Xbox. Possibly due to the lack of progressive scan and the low res output.

Chap, please, this was a serious thread until not so long ago... Leave the non-sense outside, or at least post something worth reading.

 

[Guden Oden]

I'm sliding off-topic here, but I can't say I see much of a difference regarding pixel size between PS2 and GC on my TV, which admittedly, is interlaced. 640-res is going to have the same pixel size no matter what the output device is.

Just drop the JeffK spelling, mmkay? It's fairly annoying and makes it difficult to take you seriously.

Thanks.

 

[chapban.]

OKAY BACK TO TOPIC!

 

[Megadrive1988]

I am wondering if Microsoft will be able and willing to use any memory beyond the various DDRII revisions. probably not, though ?

 

[London Geezer]

I'll ask again, or at least give the input again, i think this got lost in the OT bashing...



If the next Xgpu is a DX compliant derivative, then it should have no problem at all running what NV2A is running now. I was actually thinking what "extras" the Rx00 in Xbox2 could add on top of the original software... like AA, AF or additional resolution. Shouldn't be too hard to implement (see: FORCE IT) would it?

Playing PDO at 720p with LOADS of AA and AF would be sweet. Although i'm sure there will be games on Xbox2 that will make PDO look "below average" in 2 years time... I'm strictly speaking in techincal terms here! PDO will always look good because the art is good, much like other games that rely on art rather than features, which is not what PDO is, but you get my point...