Intel's ability to upgrade from 14 nm to 12 cores next year depends on their manufacturing roadmap and market demands.
Intel's ability to upgrade from 14 nm to 12 cores next year depends on their manufacturing roadmap and market demands.
Intel's latest desktop CPU announcement has been released, suggesting they are nearing their maximum capabilities. What are your thoughts on their plans for next year? It seems likely they aim to pack more cores into their 14 nm chips, though it’s unclear if that’s feasible without evolving them into a Threadripper-style chiplet. It’s possible they won’t be able to continuously boost clock speeds indefinitely. The most reasonable approach might be to reduce prices, but based on their current strategy this year, they appear determined to avoid that.
Their ten essential components are already reaching 250W at 5.0 or 5.1; further improvements in cores and clock speed aren't possible. One option is to reduce it.
They might eliminate the iGPU and build a 12-core special version. I’m hoping so. This seems to be the unusual final push tied to their 14nm manufacturing.
this won’t provide sufficient power. Intel won’t take any action. no matter how hard they try with Threadripper and EPYC, they essentially stop doing anything.
They are definitely capable, as demonstrated by their existing X99 (up to 10 cores) and X299 (up to 18) systems. These are all monolithic CPUs without chiplets. As core counts rise, they face challenges: matching the speed of fewer-core models, but benefiting from single-core performance boosts and significantly lower base frequencies—similar to laptop processors or specialized cooling chips.
Producing these at affordable prices remains a hurdle, requiring cost reductions that justify their value. Sometimes innovation demands compromise; investing in alternative designs (like chiplets) or architectural advances (such as the Sandy Bridge era) becomes necessary. AMD experienced this shift, especially when launching Bulldozer while Intel phased out Sandy Bridge.
They adapted by focusing on strengths—high multithreaded performance, power efficiency, and cost-effectiveness—while maintaining their core strategy. Despite setbacks, they continued refining their approach, sometimes with mixed results. AMD released models like the FX-9370 and FX-9590 to capitalize on clock advantages, especially in challenging markets like HEDT. They also improved pricing for consumer chips, though they still hold a niche for high-performance gaming.
Price reductions aren’t always the solution; sometimes limiting production and maximizing margins on profitable products works better. Intel once faced similar pressures, prioritizing high-margin segments and letting less profitable areas to competitors. This path can be effective if executed wisely.
It seems like they're questioning how they'll compete in the consumer desktop space now that AMD has achieved 5nm optimization. The upcoming upgrade would have put them at a disadvantage unless they had already improved their die shrink technology. The price-performance ratio is clearly stronger, especially where I live, but it's mainly due to their ability to keep single-core advantages. If that wasn't the case, it would be tough to make a case for investing in Intel.
The 10 core setup is the final push. Check if a willow cove version fits Z490 boards. Which game needs 24 threads? A 12 core 10900k would consume way too much power for full-core operation without direct die mode. Trying Realbench 2.56 on a 5 GHz 10940X will strain most fans except the top three. See how much heat the 10940X produces even when all cores run in Cinebench R20 at 5 GHz—it won’t be cooled by any current fan.