The latest results show Zen 3 performing just a bit better than Intel's 10th generation on tests.
The latest results show Zen 3 performing just a bit better than Intel's 10th generation on tests.
it’ll be fascinating to observe how userbenchmark adjusts its expectations for Intel this time (what keeps their top pick, the 9100f, leading the pack?), perhaps they’ll add more exaggerated criticism about “too many colors bad” to push their point. That’ll reveal how mature and impartial they truly are. By the way, how does their odd scoring system make 2080tis over 10% better than the 3070s? Can’t wait to see their performance on large navigation benchmarks too!
Nothing changed, we understood the 5000 series would struggle against Comet Lake in games. With the Ashes of Singularity Benchmark leak and AMD's announcement, it became clear they offered about 7% better performance than a non-overclocked 10900k paired with slow RAM. If you ask anyone, nobody came out on top—AMD didn't win, Intel didn't, and we as consumers definitely didn't either. (especially when factoring in price increases). It's been three years since Coffeelake launched, and gaming performance hasn't improved more than a single digit since then.
from a gaming standpoint, the difference is minimal. For creators and similar roles, performance is still solid. Ryzen has already shown strong computational capabilities compared to Intel. However, the 8700K launched at around $370 feels less appealing when you compare it to a 5600X for about $300. I actually anticipate Intel might become the more cost-effective option soon, especially if they continue to lag in power efficiency and workstation applications. Zen 2 could very well represent the best value for AMD in the near term, while Intel struggles to match pricing on other fronts. It seems AMD is catching up or surpassing Intel in value, whereas both brands have fallen short for budget-conscious buyers. Not everyone can afford a $300 CPU with a 500+ GPU, so many will likely stick with older models at full price until now.
The value of a 16-core processor is limited to a tiny segment of users. Even though these CPUs are not brand new, moving them to a mainstream platform at a reduced cost because of higher production efficiency won’t significantly impact most buyers. The term "per-core speeds" isn’t well defined—do you mean performance per individual core or overall multi-core throughput? Clarifying this would help explain the concept better.
They secured the win by delivering more than two generations without requiring a new MB.
They actually did. I've noticed folks moving from the r5 1600 to r5 2600 and then to r5 3600. Over that period I haven't purchased anything priced around the Intel 8700k, and it still performs better.
AMD is clearly the better choice for me, regardless of minor speed differences. Building a new machine now would be easier with AMD parts. I’m tired of Intel and frustrated with Nvidia’s restrictions. Intel’s overclocking support isn’t available on non-K models, and ECC RAM is now affordable—used 32GB of it for a 4th gen i7 could cost around €40. That’s significantly cheaper than buying similar memory elsewhere. Nvidia limits virtualization support and blocks open-source drivers, making their cards on Linux mostly unusable. AMD’s recent improvements make it a much more appealing option, both for performance and features.