Arrow Lake vs Ryzen 9000
Arrow Lake vs Ryzen 9000
They lack that feature entirely. You require a tailored loop designed for 14900ks, powered without limits, with flexibility for every core up to 6ghz—sometimes outperforming a 7800x3d setup. These are generally not standard options and even then they’ll find it hard to stay ahead. It’s frustrating now.
It hinges heavily on the workload and the typical values will be much lower than expected. That range could fit Cinebench best when running with HT/SMT enabled, especially since gains come from overall throughput rather than individual thread speed which tends to drop. Dropping HT/SMT mainly cuts the risk of Windows making odd scheduling decisions. In the extreme scenario, HT/SMT offers almost no performance improvement while still boosting power use. For now, we can't be sure until the products are released and evaluated. We may learn more about Arrow Lake’s performance as the earlier low-power Lunar Lake is also built on the same core architecture. If rumors hold true and AMD doubles FP execution in Zen 5, it could mark a long-awaited shift where AMD clearly leads in FPU capability across the board.
I’d like to see more Arrow Lake models featuring over eight P cores per tile. Reports suggest this won’t materialize. I’m hoping AMD’s Zen 5 will deliver a single CCX with more than eight cores on one chip, similar to what Zen 3 and 4 offer. It’s hard to judge the real gains without seeing actual benchmarks. Some recent updates seem to disappoint compared to earlier expectations for Zen 5. A 10% improvement in IPC and comparable clock speeds would be encouraging. AMD appears to be prioritizing AI hype over solid performance, which feels misguided. GPUs still shine, but the CPU race seems overblown. The Arrow Lake seems only a slight step forward due to clock speed changes, but better IMC, improved thermal management, and lower power consumption would help. With stability concerns like those seen in previous generations, I’m hoping for more consistent results. I really want Zen 5 to bring at least a 20% IPC boost—especially since we’re still limited to eight cores per CCD for the upcoming RTX 5090. If performance doesn’t improve significantly, sticking with the 7800X3D might be the only viable option, even if it means missing out on much higher gains from Zen 3 and 4. I’m also concerned about rumors of reduced WIN10 support due to this push toward AI acceleration. Hopefully, AMD delivers real progress beyond just marketing buzz.