Talk about 10 nm Intel desktop CPUs and their trade-offs between Zen 2 and Zen 3 performance
Talk about 10 nm Intel desktop CPUs and their trade-offs between Zen 2 and Zen 3 performance
Intel's upcoming 10 nm desktop CPUs are expected to launch early next year. How do you see their performance compared to the newest AMD offerings? While these chips likely won't be top-tier yet, I wonder how the team's parts will match similarly priced AMD models. The Ice Lake mobile line already shows around an 18% improvement in IPC over the previous generation. I believe Intel will quickly regain its advantage across all price ranges once they efficiently integrate more cores into their chips. Their current strengths in gaming, clock speeds, and productivity suggest they’re well-positioned to solidify their lead.
Irrelevant to guess since we lack performance data for Intel's 10nm desktop models... unlike what we see with Ryzen.
What do you believe will occur once Intel releases a 10 nm architecture for desktop CPUs? I expect concrete performance figures to emerge. While we can discuss ideas freely, the reality is that precise data remains elusive.
Yes, it's available on mobile for both options. What are your thoughts on how it might look on a desktop?
our estimate points to a 10nm mobile chip, though this is limited by thermal factors and not fully reflective of real-world performance. It's probably not representative of early silicon. The term "desktop" likely refers to NUC-style devices intended for desktop use. The main challenge isn't the packaging of cores but rather achieving reliable core yield. Intel's mature 28-core server CPU is already difficult to manufacture reliably.
The main concern isn't the appearance of 10 nm but the timeline for its arrival. At this pace, it might still be another year or two away, and any promises from Intel would likely be met with skepticism. Over the past four years they've repeatedly underestimated delivery dates, making exaggerated timelines a common tactic to avoid concern among investors. It's also likely they'll stretch their statements to obscure the reality of their challenges.
Technically, improvements from 14nm+ seem limited, possibly even reducing performance instead of boosting it, which is already visible in mobile devices. Intel currently holds only a minor advantage in single-threaded tasks. If they lose that edge to Zen 3, they risk falling behind entirely. They'll likely continue increasing core counts while accepting slower clock speeds, which would be a significant setback.
AMD, on the other hand, excels in multi-threaded performance, efficiency, cost-effectiveness, memory and PCIe bandwidth, and security—areas where Intel lags. Their products are more advanced overall, especially for mobile users who are now at a disadvantage compared to AMD.
Intel's strongest position remains in mobile markets, where AMD's offerings are outdated and focus on mid-range segments. It will take years for Intel to regain serious ground in high-end desktops and servers, unless they develop broader solutions like Ryzen or a next-gen architecture (such as Sandy Bridge 2.0) that address their current weaknesses.
It seems Intel is increasingly focusing on mobile computing, which appears to be their most profitable area. At least for consumers, that's the case. While desktop powerhouses remain popular with enthusiasts, mobile devices seem to be where people invest their money. Most consumers aren't interested in high-performance PCs; they prefer powerful SoCs and move on from there. Personally, I believe the era of large, consumer-owned powerful PCs is coming to an end.
It's a smart decision right now. The long-term vision is unclear, but the requirements and limitations of the mobile setup fit perfectly to leverage their strengths while minimizing weaknesses. Just keep in mind, this seems like the best choice at the moment.