AMD Ryzen post-release thoughts and explanations

Ok, I’ve talked about few things regarding AMD Ryzen processors in my last article, I’m going to expand it a bit further with this one, explaining few things that people are concerned over or are raging about, be it justly or unjustly…

Games performance

I kinda forgot about this since Intel was dominating the market for so long, but Youtuber and hardware geek JayZTwoCents reminded me of this. It’s the processor specific optimization of games and why current games perform worse on AMD Ryzen processors even though it clearly has identical IPC (Instructions Per Clock) capability. Partially it’s clock fault because Intel’s quad cores simply come clocked way higher which is favored in games, but mostly, it’s processor specific optimization. Intel pretty much dominated gaming segment for 5 years. That’s eternity in PC segment. And with that, all game studios kinda focused on optimizing games for Intel processors only. Now that AMD is back in the game, things will change again. I don’t expect AMD to dominate the field, but you can be assured they’ll at least get on the same fair level as Intel with upcoming games. Some studios might even optimize current games to better support Ryzen.

Memory (RAM) issues

People complaining about memory issues a lot and complaining how AMD dares to release new platform with such issues, not realizing it’s not AMD’s fault. At least not entirely. Sure, they need to work with motherboard makers to ensure everything is in check with their memory controller inside CPU, but from there on, it’s up to motherboard makers to add RAM profiles, enhance compatibility and deliver all that in form of BIOS updates. Considering AMD Ryzen is an all new architecture with all new memory controller, expecting such monumental release to be problem free is really silly thing to do. When Intel released triple and quad channel boards after years of having dual channels around, they were problematic as well. And they still are today and I know that from first hand experience as I owned both, triple (X58) and now quad channel (X99) setup. With BIOS updates, AMD and motherboard makers will solve compatibility issues when it comes to memory. Also, be aware that if you want absolute compatibility, you have to strictly follow QVL lists provided by board makers. They can only assure rock solid performance and stability with memory modules listed there.

Limited AMD Ryzen overclock capability

I’ve seen quite a lot of people whining how bad AMD overclocks. But all these people are leaving out one super important difference. They are taking overclocking capability of freaking QUAD cores and applying it to EIGHT core processors. That’s not how things work and they never will.

If we look at Intel Core i7 6900k, same core configuration as AMD Ryzen R7 1800X, it also peaks at around 4GHz. Anything over that and you need huge amounts of extra voltage and you’ll also get huge thermal footprint from it because of that. You have to understand it has 4 more physical cores and 8 more threads. This essentially means it’ll require twice as much power and output twice as much heat. It’s not that simple and linear, but for better understanding, that’s what it is. So, you can’t compare a overclocking capability of a freaking quad core to an actual octa core. It would just make no sense.

We can however judge AMD when they release hexa and quad cores. But there is also one more factor. AMD Ryzen was designed on manufacturing process that is called LPP. And LPP stands for Low Power Process. CPU’s designed in such way are bound to be very power efficient, but very stubborn when it comes to high clocks. And that’s the way AMD designed Ryzen. Things may change in the future as they will refine and adapt the manufacturing processes.

Power consumption

I’ve heard about complaints on Amazon about AMD Ryzen power consumption and how it’s clearly not just 95W…

Well, we have to first establish two things. How is TDP (Thermal Design Power) measured and more importantly, where (or more precisely, at which processor clock).

Intel for example measures TDP at processor base clock (which for 6900k is at 3.2GHz). And they measure it as average value and not maximum value. It is yet uncertain how AMD measures it for new Ryzen processors. At least I wasn’t able to find any info on that where Intel clearly states how they measure it on their ARC page.

Now, lets take a look at this chart. I’ll focus on two processors to make an example, measured wattages are at wall socket, so understand that (this is not just CPU; this is power draw for whole system). One is Intel Core i7 6700k, a quad core processor with 8 threads (4c/8t configuration) and the other one is AMD Ryzen R7 1800X. The power draw is almost the same, but R7 1800X features twice as many cores and threads. And it’s not just games where it might depend based on actual cores utilization where games usually use just 4. It’s the same in AIDA64 and Handbrake, which both use 100% of all cores. AMD Ryzen has basically done twice as much work on twice as many cores and still delivered same power consumption at the wall socket. This means, regardless of TDP numbers, it’s a pretty damn efficient CPU. It even draws significantly less power than very similarly configured Core i7 6900k. For the number of cores and threads, Ryzen are pretty damn power efficient processors.