Apple vs PC: The war between creatives and engineers
The world of apple and PC lovers, have always been at odds with each other. While there is some unfounded bias to support the ecosystem you use and hate on the other. There is also some valid reasons why this might be the case and what could possibly be an opportunity to close this gap further.
Background: Design and Engineering?
I have always been a rare breed of folks, who dabbles between the creative space and engineering space. I spend weeks on building my own app, reading and implementing the latest in AI, experimenting with 3D applications and writing extensions. You don’t have to take my word, you can check my GitHub profile(clearly see months where I have no commits:P) —
But at the same time, I try and dabble with on the creative space — like trying sketching, music production or writing and more. I tend to see the patterns of how I incline towards my windows PC for technical workflows and how I tend to my iPad for creative workflows. I started to wonder, why this was the case, and maybe come up with a middle ground that could work.
Apple
The Apple ecosystem is great when you have creative workflows. The entire creative industry are inclined to a Mac workflow, one way or the other. It is no surprise that creatives from the music industry, artist and designers all prefer the apple ecosystem. A video from Hardware Canucks interviewing a music producer and a photographer, explains why —
It seems that one of the major consideration for anyone in the creative space is the need for Hardware to not come in the way of production. The entire process of creation of something new and novel seems to a challenge in itself, so any additional layer added to the process needs to be seamless and Apple seems to excel at it.
Take the example of music production, the entire process of recording musical instruments and vocals into a streamlined process. Where it can be further tweaked to add effects and plugins, seems to be possible because the Apple Hardware take the backseat, while being an enabler to consolidate all of this information in it’s platform. This seems to be really valuable for a creative.
Just to drive home the point further, let’s take a look at the big picture. What is the end product for most creatives? It is something that the audience can experience that can transform or give a new perspective to everyday things. Like, a music producer creates a piece that is expected to transport the person listening to a whole new world, or the world that the producer envisions and share the same experience to the audience. Same goes for a photographer, it is to capture some amazing pieces of the world, stopped in time — into a Frame. The expectation is to transport the viewer to the exact time and space of the photographer, so that they can experience the same beauty that they tried to capsule into an image.
Windows PC
The PC ecosystem is great for anyone in the engineering/scientific research space. The entire sector where raw performance numbers matter, shine in the PC world. The people who are into development, into building machine learning(ML) application and models, into building scientific and engineering applications all benefit from the PC ecosystem. But why is this the case?
The story for anyone in development, seems to be that raw PC performance is where most of their work is. The major difference between a creative and a researcher/scientist is that. A creative is building new experiences with existing pieces of technology or tools. But engineers are building the next set of tools to enable future research and possibilities. The general space where they tend to lie is in the corner of the hardware capabilities. Let me explain.
I was working on a diffusion model to generate images. While I was able to run an inference(use it out of the box) of a really optimized diffusion model(image generation tool). Once, I started to explore more non-optimized models like say DreamBooth, which could take input data and re-train the data to match input specifications. It required a lot more GPU requirements, which was not available to me. Some of the options I had though was to buy some more RAM or Nvme drive to offset performance from the GPU with new supporting libraries like HuggingFace accelerate. But something like this would almost be impossible with an Apple device.
CPU v/s GPU conflict
Based on what I have concluded thus far, I think the pattern lies in the space that you are more likely to work on — is it the CPU/RAM space or GPU/vRAM space. Apple basically took the crown with the invention of the Apple silicon with the M1s and are just pushing the boundaries with M2s, the Max’s and the Pro’s. The clear integration of RAM to the CPU — which is basically why you cannot extend the RAM in the future. You need to known exactly how much hardware you would need before, and even if you get the maximum available hardware at the time of purchase, you cannot extend it to the latest and greatest in the future. Which why you see many production companies, still using Macbooks/Macs from 2015 and it still gets the job done. You don’t really need the latest and greatest hardware, so long as it doesn’t hinder the production process.
But, the case for Windows PC is so much different. Many of the latest in scientific and engineering research, mostly first and foremost tend to have a really great GPU. Nvidia GPUs have paved the way to democratize GPU hardwares to the general public with their more affordable 60 series GPU and keeping the ceiling as high as possible. Allowing the enthusiasts to explore it based on their needs. But again, the main advantage here is that, many applications of development can make use of the excess hardware available. Let’s say a ML model requires a minimum of 12Gb vRAM to run an inference, you can buy hardware capable of running it in your local machine. Say you need additional RAM to support the GPU functionalities, you can always swap in additonal RAMs/NVMe drive to speedup your work.
Conclusion
At the end of the day, it all really depends on your workflow. And currently, clearly one ecosystem is more suited to you compared to the other. But, all of this can change in the future. With two key developments-
- Further integration of GPU and CPU —
The entire power of hardware is capped with the inter-play among separate components. Like the saying goes, you are as fast as your slowest member or something like that. Same goes to performance, the performance is capped by the bottleneck of the slowest hardware. So companies are investing further into owning the entire stack to themselves, that is why Apple moved from Intel CPUs to their custom silicons, which gave them competitive advantage with the rest. Similarly you can see that Nvidia announced their custom CPUs at Computex, 2023 which means they can optimize the entire pipeline. Also other companies are trying their shot at the same, Intel announced to get into the GPU market with affordable GPUs and AMD doing great both in the GPU and CPU space with the Radeons and Ryzens respectively. The pattern is the same — own the full stack and optimize it to perform well with one another and free bottlenecks. - Cloud GPUs —
I also believe there is another niche growing market for cloud computations. While all of the things covered in this article is for on-premise computation. There is a story to lease all computation to the cloud and just maybe not have the entire Apple vs PC conflict to begin with. Maybe all we need is a good web interface to allow us to interact with applications seamlessly and maybe we wouldn’t care so much for the ecosystem and can concentrate our efforts on what we build.
But all of these are still under active development. And what we have today are two great ecosystems — Apple and PC. While they are both good at what they do. The solution for someone dabbling into both creative and technical space maybe just need to have two ecosystem to support their needs. But with really affordable devices like iPad and Mac mini for the Apple ecosystem and really performant laptops like Lenovo legions with RTX GPUs, maybe you aren’t that bad off. What a time to be alive, with the power of years of hardware innovation at your finger tips at fraction of the price!
