LevelOneTechs identified a problem with the Ryzen 9000-Series Performance.
LevelOneTechs identified a problem with the Ryzen 9000-Series Performance.
This topic seems confusing at first. Researching a problem can feel like a witch hunt. I wonder if similar discussions about Intel 13th/14th Gen issues are also being scrutinized. Or perhaps you missed the main goal of the forum? At this stage, many confirm the situation through various sources and experts, including AMD’s own team. DerBauer has already reported some performance quirks. It raises questions about whether a public update will come soon or if they’ll wait for fixes. Personally, I don’t see much value in this thread since I already have an AM5 setup with an X3D chip for gaming. Even with better performance than previous Ryzen 9000 models, I wouldn’t upgrade unless it’s necessary. This conversation is mainly about investigating a known issue, which is all there is. The main challenge is the numerous factors affecting new chip performance—drivers, motherboards, RAM speeds, OS choice, and even admin accounts can influence results.
The situation isn't about a bug itself, but how it was shared internally at AMD, causing confusion outside the company. This raises concerns about accountability, suggesting that someone in a single department might have misled others about what was happening.
In short, if the fix comes, 7000 should improve slightly, but it won’t really change how reviews turn out. It’s unclear if the 5000 series will be impacted too. Most Ryzen chips might see a minor boost, though the 9000 still doesn’t offer much more value than before the issue was found. The 7800X3D could actually improve slightly and become even stronger, making the 9000 lineup look less impressive. It turns out my guess about the Ryzen 9000 was correct—if non-X3D chips can’t beat the 7000X3D, this generation might struggle. People don’t mind whether one CPU has X3D cache while another doesn’t. The 9600X should perform similarly to a 7800X3D in games, and other CPUs should be better. Now AMD seems to have plenty of excuses for why the 9000 series isn’t keeping up. It’s strange how Intel is making such a mess right now, and the next big step from AMD appears to stall completely, just like during Intel’s previous struggles. You can’t keep this going. Luckily, I just got my 5800X3D, so I should be fine until AM5 reaches its end of life.
It appears 7000 might improve speed, yet 9000 should also see gains. There appear to be several underlying problems, mostly related to Zen 5. The Windows admin effect impacted both, though not equally. If they work well on Linux, companies could deploy them in LSF grids.
I mean there's a small possibility that older Ryzen generations will also see a small improvement, but the main issue is that Ryzen 9000 clearly wasn't ready for launch. The inconsistencies in performance found with 9000 are pretty severe and the other generations don't really have that same issue. So while fixes could also provide a boost to older generations, I'm fairly certain at this point that the fixes will give 9000 a larger gap in performance over 7000 And the effect on 9000 Well certainly be the most drastic.
I'm not attempting to be rude, but Linux plays only a minor role in PC gaming, likely even missing out on the performance gains that AMD achieved for that platform. Their upcoming releases should deliver much stronger results on every device, particularly on Windows, where most users rely on them.
This discussion centers on PC gaming, yet the majority of these components are used elsewhere. I completely agree that it isn't ideal, even in Linux, for the needs of most users on forums. I didn't mention gaming when talking about LSF. If I started a large Synopsis regression test, those results wouldn't go into Xeons or Epics. They would be sent to dozens of i9 6GHz Raptor Lakes, taking several days to complete. AMD likely recognized their performance improvements on Linux. Synopsis/cadence doesn't run on Windows, as far as I know. It's hard to picture them neglecting their internal components for testing purposes.
The numbers don’t add up like that. A single 1% improvement across all five games means overall performance improves by 1%, not 5%. Regarding the Windows admin issue, it impacts every processor, including older Intel models. It’s not a glitch but a fundamental part of how the system operates. I wouldn’t describe it as a bug, honestly—it’s just how Windows functions. Even if it gets fixed, it probably won’t change the final outcome.