Big deal, my 1996 IBM Apteva had MMX
that actually IS impressive since MMX came out in 97
i’m not sure if that’s true, but i’ll see if anyone can get a better setup for this joke.
I just want to know why Meteor Lake didn’t get XMX.
#I like your funny words, magic man!
“Intel’s ZetaFlux CPU debuts with Z-AX3 Core, V-TECH Sync, and NQI Integration, ensuring blazing speeds and futuristic computing.”
It’s those cores that accelerate XeSS yeah ?
XMX extensions are designed for matrix multiplication operations in FP64, FP32, FP16, and bfloat16 formats, which are crucial in many AI algorithms, including neural networks
From the article.
bfloat16
I have no idea why this is important or really what it even is but Apple had a pretty video about it in their WWDC catalog so I guess it’s a trend now.
Standard float16 uses 1bit sign + 5bit exponent + 10bit fraction.
bfloat16 uses 1bit sign + 8bit exponent + 7bit fraction.
bfloat16 basically gives the same exponent precision as a standard float32. But most neural networks don’t require a huge fraction range. So bfloat16 gives you the possibility of executing 2x 8bit NP FLOPs vs using a float32 to do the same 1x8bit NP FLOP.
Having the ALU support this format allows for the scheduler to pack 4xbfloat16 that can be executed in parallel in a standard 64bit ALU. So basically you double or quadruple the 8bit NP FLOPs that you would get from using traditional float16/32 representations.
Makes sense. Thanks for that!
Xe-LP iGPU have nothing in common with Xe dGPU currently, they have no RT capability and no XeSS capability, they’re mostly glorified Intel HD iGPU.
This is not true. Here’s Intel’s official slide on it
This is not true. Here’s Intel’s official slide on it
This is not true. Here’s Intel’s official slide on it.
Your slide is Xe, not Xe-LP
Intel even admitted themselves Xe-LP doesn’t even have ANY Xe-cores
The picture literally has the Xe core written on it.
Good thing this is Xe-LPG then. LPG is Alchemist based and goes into Meteor Lake. LPG+ goes into Arrow Lake.
the hell does that mean
New GPU architecture, now also integrating Intel’s ‘tensor cores’ for better XeSS among other things
It means it’ll be able to use the best version of XeSS.
… what?
Intel competing for top spot in this benchmark
They seem to have forgotten the x1950 XTX by XFX
XFX should have gone with MERX this gen to further expand their lead
idk call me when they release some XXX or some XNXX version of this
Xenon and liquified petroleum gas. Interesting ingredients…
Xenon’s critical point is nearly at room temperature, so I guess it’s not impossible for Xe-LPG to exist in a computer.
