No chiplet design inherently restricts clock speed, but careful integration can optimize performance.
No chiplet design inherently restricts clock speed, but careful integration can optimize performance.
Just another topic for a fun discussion. I've noticed that AMD Ryzen series with chiplet design seems to have a very limited overclock potential and also using quite a high voltate power out of the box. Is it possible that AMD chiplet, by its very nature, limit its potenial to be clock way higher and deliver substantial more performance at the cost of heat and power just like how Intel did with their Alder Lake? Since both Intel and Nvidia seems to prefer craking power to deliver consumer the maximum performance, it's only logical that AMD did the same at least to some of their product to improve their competitive posittion but they had never done it even now. So I wonder if they are limited by the physical nature of their chip in some form.
The CPU OC has been inactive since the 9th generation, leading to a significant drop in potential on both sides. AMD is currently the least capable in terms of OC performance, typically offering only 5.1 to 5.3 GHz with normal cooling, which is quite disappointing given the boost clocks.
OC potential is now extremely low across all locations. Producer companies have discovered a way to avoid setting aside 30% of unused performance capacity, instead delivering parts near the maximum limit.
That's clear, focusing on clock speed rather than maximum output potential makes sense given the conversation. Intel Alder Lake is really pushing its limits, easily surpassing Ryzen 500. It seems AMD hasn't reached the same level of performance boost with their latest chips, which explains why we're asking this question. It might just be that they can't achieve the same efficiency at 180 watts.
The design stands apart entirely, making direct comparisons with simple clock counts meaningless. It’s not just about chiplet size—it involves deeper principles beyond that. While smaller chiplets might seem to support faster clocks, the overall system complexity introduces factors that go far beyond a straightforward approach.
Zen 3 reaches speeds up to 5.1GHz, showing clockspeed isn't restricted by chiplets. The focus is on refining the Zen design.
We can't pinpoint the exact reasons behind AMD's lower base clock speeds. Clock speed by itself isn't the whole story. I don't focus on technical details like cores or specs; instead, I rely on reviews to understand performance trends and where efficiency starts to drop.
Power remains important, though you could argue that a Ryzen 5 at 5 GHz isn't quite the same as an Intel 5 GHz. Still, up to now they haven't managed to boost performance beyond the window limits like Intel has on the CPU side. It seems more logical to focus on whether chiplet is a barrier for stable operation at 200 watts and if it offers meaningful scaling compared to just a hundred watts. It makes me curious—if AMD loses the TSMC advantage because Intel closes the gap, would sticking to chiplet actually put them at a disadvantage? After all, monolithic designs let you run the clock higher as long as cooling works. I don’t see a clear reason yet, but it looks like that’s probably the situation so far.
It seems the discussion isn't really about chiplets, but rather other distinctions. The node size is likely the biggest factor. AMD's chips use a smaller node, which means they're physically smaller but also more challenging to cool. This results in lower overall power consumption, no matter whether they're built from chiplets or monolithic designs. High power isn't ideal—it's only necessary in specific situations.