The 5900X has better silicon.
The 5900X has better silicon.
I've been analyzing why 5900X CPUs tend to stay cooler than 5800X despite having four extra cores. The main ideas I've come across are: the 5800X uses a single CCD with two active cores, making heat more concentrated. The 5800X employs lower binned silicon, which demands more power and leads to higher temperatures. You mentioned the 5800X uses a 5950X CCD while the 5900X has two CCDs with one or two failed cores each. Those CCDs are actually better binned than those used for the 5600X. This suggests the 5900X might have worse silicon quality, but since it's built from chiplets with failed cores, it could still perform better due to the surviving cores.
This seems more than just a random thought. Are you curious about checking CPU performance? Ryzen chips usually don’t benefit much from overclocking unless you significantly increase clock speeds and voltage. The 5800X tends to get hotter due to its single-core design, while the 5900X might offer better performance in certain scenarios.
I believe you're overanalyzing things. The 5800X features its eight cores much more densely than the 5900X's twelve, which increases thermal density and reduces efficient heat transfer between the CPU and cooler—essentially your main concern. I question the idea of significant 'binning' here, since these are standard parts and the current chip shortage makes such practices unlikely.
I'm just curious. I already purchased a 5900X and I've noticed that my temps are lower than those looking at temps for the 5800X. Those aren't my hypotheses. Those are the two primary reasons I've seen given on sites like here and reddit. I agree that the first one is more likely, but perhaps there's something I don't know. There's certainly a degree of binning happening, otherwise the 5900X and 5950X wouldn't have higher boost speeds than the 5600X and 5800X. If all the silicon was identical, they would all be able to boost to the same level. I'm pretty sure all AMD silicon is binned, because there's no difference between the silicon for servers and the silicon for consumers. They all use the same chiplets, so you can't bin for only the server-grade ones, because the only difference is the binning.
It doesn't matter whether the chips are really outdated. Usually the binning serves a practical purpose for core count, and results remain consistent enough that yields between similar models are nearly identical. Binning outside the manufacturer's range is another challenge. Occasionally, top overclockers receive entire batches for testing, sometimes with special cooling solutions, leading to impressive world records through trial and error. The manufacturing process has improved significantly over the past decade, making chips more uniform. Today's Ryzen chips tend to perform very similarly at comparable voltages, so trying to gain a small speed edge is often pointless.
Many 5800X chips running at lower loads or idle tend to run hotter than the 5900X, but under full load they draw more power. The issue arises because the 5800X is so thermally dense with its full CCD that heat dissipation becomes challenging. This affects binning decisions, though it doesn’t matter much to me. There’s a big range—ranging from -8°C to +8°C—among users reporting similar cooling setups. Aris from Hardware Busters shared a 5800X reaching 91°C on a D15S, while others using a Frost Spirit 140 or Peerless Assassin 120 in a NR200 with the side panel off achieved about 73°C. Of course, these differences sound different, but they’re not just about PWM percentages.
I don't think that's the right approach. Your cooling system could be more effective. From what I saw with my 5900X, it performs better and stays cooler than most AIO users. Does that mean my air cooler is actually cooler than an AIO? Not really. My cooler gets direct access to a lot of fresh, forced air outside. Many people just use a heatsink with minimal fan action. You could even try yourself and improve it yourself. When they switch to an AIO, their setup suddenly improves because the cooler is getting real air. I usually take reviews with a pinch of salt, but I still consider them when making decisions.
The final outcome of using ambient coolers is the room temperature. Clearly, the cooler with the lower ambient reading can provide better performance than a liquid cooler. The overall difference might be around 50mhx PBO. Self-adjusting overclocking renders manual overclocking unnecessary.
There are many factors to consider, such as the case, fan size, ambient temperature, fan speed, heatsink placement, PC location, and more.
Based on what I've seen online, the 5800X often shows higher temps under similar cooling setups. I explored this after reading a forum discussion, where someone with a better setup reported similar or lower temperatures. My current cooling system works well for me, though I'm curious about the trade-offs with a custom loop. The Fuma 2 + P300A + extra fans seems solid, but I'm skeptical about its edge over a more tailored solution.