X86 Vs Arm | Gaming & Emulation What’s The Best Solution?

7 thoughts on “X86 Vs Arm | Gaming & Emulation What’s The Best Solution?”

• MacBooks these days run on ARM and they're blazing fast

• I KNOW that x86 does emulation very well. I have computers on my workbench that are probably older than a lot of people watching this video. I used them for emulation back in the day. Raspberry Pi and Android devices just took over when they came out. They were little, cheap and easy. It wasn't fun trying to get a PC hooked up to an old CRT TV. Gaming PCs doubled as heaters back then. They were great in a cold apartment and not so great in the summer. There were a lot of cables, frequently strung across a room and they frequently broke. Cats seemed to like the taste of S-video and rca cables for some reason. I had one that chewed through them every few months. There were a lot of reasons why we got away from PCs years ago. It's time to go back. Modern PCs are so much better.

• A pretty good summary. It has often been difficult to make direct comparisons between ARM and X86 based upon benchmarks, especially not knowing whether the metrics used are different. Antutu is another one that can potentially skew opinions based on results.

  In terms of making comparisons between an SBC and an Mini PC, whilst factually correct that they are different based upon the exposed I/O. when removing the GPIO and similar, what remains at the base ore are computers that can have some degree of comparison. I personally believe that most people likely purchase a Raspberry PI and similar for the purpose of gaming on cheap, low power consumption device, so often reflected in the number of review videos that have heavy focus on gaming.

  The RISC vs. CISC aspect is one that I suspect many (including me) forget about when making any comparisons between the processors and probably counts for most of the key differences in real world performance between the competing architectures.

  It is going to be interesting to see what the Radxa 2XL will do (it looked like it is still sealed?), especially if you do have the base 2Gb model.

  Nearly 14 months since people first began receiving their Orange Pi 5's, development is still looking slow and I fear the RK3588 becoming a lemon like the RK3399.

• You only left out two things that I could tell. RISC needs more memory when compared to X86_64 namely in cache this is because of the repeated instructions. Second it power draw I will say you may not beat them in raw performance for emulation, but power draw does matter as well. I also wanted to point out that the core clock you reported for the OPI 5 in the single core score was that of the little endian core vs the big endian care, this is a issue that doesn't seem like even the emulators can pick up and we don't have a way of telling them to use the big cores to use for emulation and having looked through the code to compile them they aren't telling the compiler to use the big cores on arm setups, I honestly blame RPI Foundation for this as they only one type of cores vs the two that you get in many other systems and they are leaving it up to schedular to determine this.

• You're doing awesome work informing the audience and I look forward to more comparison videos as the competition to RaspberryPi has grown with better value to performance!

• Excellent video man def touches on things I was curious about

• Nice hardware collection. I missed a chart of an emulation benchmark. So like choosing at least 1 (or a few) game(s) and measure the frames per seconds of that game(s).

Comments are closed.