Intel Invites Gamers to join a Global Graphics Odyssey
- Thread starter WYP
- Start date
Dawelio
Active member
Rather old article and not sure if it's legit or not, but looking at that from 2017 and how much they spend on R&D (which you mentioning in the article that gave me the interest to look up)... It's literally insane.
https://www.electronicsweekly.com/blogs/mannerisms/markets/intel-spends-everyone-rd-2017-02/
https://www.electronicsweekly.com/blogs/mannerisms/markets/intel-spends-everyone-rd-2017-02/
Yeah NVidia has the fastest growing R&D budget & MediaTek has the highest as a percentage of sales but in terms of raw numbers Intel is in a completely different league to everyone else.
On its own I wouldn't say that really means anything for their new ventures because we've seen so many of them collapse under them in the last 10 or so years, obviously ultra-mobiles/smartphones being their biggest & most expensive failure recently, burning a massive hole in their R&D budget for around a decade with nothing to show for it when they eventually scrapped the whole scheme.
But, here they've got a wide range of relevant experience & talent hand picked from other silicon valley companies, as well as the fact that GPUs are becoming more & more like CPUs in certain aspects with modern heavily programmable pipelines, the wide range of execution units & ever growing memory hierarchys. Larrabee may have been very badly suited for fixed-pipeline Direct3D style rendering paths but for programmable pipelines & low level APIs with mixed compute it was actually ahead of its time in many aspects.
On its own I wouldn't say that really means anything for their new ventures because we've seen so many of them collapse under them in the last 10 or so years, obviously ultra-mobiles/smartphones being their biggest & most expensive failure recently, burning a massive hole in their R&D budget for around a decade with nothing to show for it when they eventually scrapped the whole scheme.
But, here they've got a wide range of relevant experience & talent hand picked from other silicon valley companies, as well as the fact that GPUs are becoming more & more like CPUs in certain aspects with modern heavily programmable pipelines, the wide range of execution units & ever growing memory hierarchys. Larrabee may have been very badly suited for fixed-pipeline Direct3D style rendering paths but for programmable pipelines & low level APIs with mixed compute it was actually ahead of its time in many aspects.
