News Intel will retire rarely-used 16x MSAA support on Xe3 GPUs — AI upscalers like XeSS, FSR, and DLSS provide better, more efficient results

[Admin]

Intel is retiring 16x MSAA support in its Xe3 GPUs, citing limited vendor adoption and the rise of more efficient upscaling methods like XeSS, FSR, and DLSS. The change begins with Mesa 25.3 and aligns with Xe3’s launch in upcoming Panther Lake SoCs.

Intel will retire rarely-used 16x MSAA support on Xe3 GPUs — AI upscalers like XeSS, FSR, and DLSS provide better, more efficient results : Read more

 

[coolitic]

I don't know why we care about Redditor opinions for news articles, but, the lack of MSAA in most games has little to do w/ the performance-cost (it's not that much more expensive than the more complicated temporal techniques), but rather mainly to do w/ its incompatibility w/ deferred shading techniques, which have dominated the games industry since like 2010.

The "incompatible w/ transparency" isn't strictly true either (but there are certainly complications), and it's actually quite amusing/ironic considering that deferred-shading itself pairs quite poorly w/ transparency, in a way that's (imo) significantly worse in terms of performance and/or appearance, versus a simple absence of MSAA ("absent" meaning strictly in the transparent portions).

As an aside: the reason deferred-shading had become dominant at that time is because it allowed for negligible per-light cost for dynamic lighting (at least, for shadow-less lights), versus traditional forward-shading which scales quite poorly w/ dynamic light-count (4-ish being the conventional peak).

However, "clustered-forward" shading catches up to deferred-shading w/ regards to dynamic-light cost, and doesn't suffer from many of the same limitations/difficulties that deferred does (ie. transparency, MSAA). The two main reasons (that I see) for clustered-forward not being widely-used is 1. complacency/tech-lag and 2. many screen-space techniques make use of the G-buffer that deferred provides.

 

[applito1994]

I don't know why we care about Redditor opinions for news articles, but, the lack of MSAA in most games has little to do w/ the performance-cost (it's not that much more expensive than the more complicated temporal techniques), but rather mainly to do w/ its incompatibility w/ deferred shading techniques, which have dominated the games industry since like 2010.

The "incompatible w/ transparency" isn't strictly true either (but there are certainly complications), and it's actually quite amusing/ironic considering that deferred-shading itself pairs quite poorly w/ transparency, in a way that's (imo) significantly worse in terms of performance and/or appearance, versus a simple absence of MSAA ("absent" meaning strictly in the transparent portions).

As an aside: the reason deferred-shading had become dominant at that time is because it allowed for negligible per-light cost for dynamic lighting (at least, for shadow-less lights), versus traditional forward-shading which scales quite poorly w/ dynamic light-count (4-ish being the conventional peak).

However, "clustered-forward" shading catches up to deferred-shading w/ regards to dynamic-light cost, and doesn't suffer from many of the same limitations/difficulties that deferred does (ie. transparency, MSAA). The two main reasons (that I see) for clustered-forward not being widely-used is 1. complacency/tech-lag and 2. many screen-space techniques make use of the G-buffer that deferred provides.

This is a solid comment that explains things better than the article