I tried running the CGE program written in python. between my laptop and the lecturer's laptop, his is working so much faster. a simulation that took a day in my computer is just working in less than 35 minutes in his. So, I wonder if buying it is worth it? I'm not familiar with mac environments too. But that should be able to be familiarized. The productivity increase I envisioned is not really high though. my heavy compute use case is running models, but my daily task usually is making powerpoints and researching for that and update data. Even for modelling, the model that took a lot of time is just three, four. some of them is also platform specific, that I can't run on mac. @xiazixin
You'd get a better bang for your buck running a PC with a good graphics card. But it sounds like you want something portable. It's going to cost you twice as much compared to a pretty good windows notebook. And you're going to have trouble finding your usual apps for your machine. I personally don't think it's worth it but it makes a great fashion statement since you're still in college and the performance isn't bad. But unless you go all in, you're not going to have a fun experience using it. The philosophy behind Macs is different from Windows and Linux. A walled garden you could say.
No. Only time you need one is to code apps for apple store. You can get a much more powerful laptop for the same price with the only drawback being battery life (but you would leave it plugged in for simulations no?) I saw it for around 2-3k usd online (may not be accurate idk) which also gets you https://www.bhphotovideo.com/c/prod...ector_16_hx_a14vig_693us_16_vector_16_hx.html for under 2k you can get a 4080 laptop with similar cpu specs too. I doubt mac has any real advantages outside of electricity efficiency. If you can’t use gpu power for your simulations then there’s other options for much cheaper. (Why not get a dedicated server for running those though?) I do recommend getting a pc for more intensive tasks vs using a laptop for everything. Laptop bottlenecks you in so many ways. More importantly windows (or even Linux) has much higher compatibility unless your specifically doing apple stuff.
If your primary tasks involve PowerPoint, research, and data updates, a MacBook M4 Pro might not justify the price. The performance boost for occasional modeling might not outweigh the cost, especially with platform-specific software that won’t run on macOS. While the M4 Pro excels in heavy computations, its benefits are more noticeable for large, frequent simulations. If only a few models require heavy lifting, consider alternatives like cloud computing or upgrading your current laptop. The MacBook M4 Pro is great for some tasks, but may not significantly increase productivity for your everyday workload.
Bro how does anyone upgrade their laptop outside of ram and storage??? Chat gpt response ngl cloud computing is a decent option though, it’s what companies tend to use
depends on what you are going to do with your mac. i am working with python and satellite data, my m2 mac is good enough for it. For ai stuff that use gpu (tensorflow / pytorch), i would rather use a workstation with a good gpu instead of laptop anyway (laptop with gpu is not good enough for that kind of stuff). for a entry level ai stuff (i have only tried random forest) it is more than enough. but, if you have a workstation already, just use your old laptop and use jupyter / jupyter remote kernel / dask gateway to process your code instead of doing it in your laptop. Remember that no laptop can substitute a workstation. Or just use remote desktop instead. if you are only going to use it for power point and excel, mac is a little bit overkill. Macbook battery is very good, which is a major point why i buy a mac, portability needs good battery after all, if your workflow need alot of moving with little time to charge, i suggest getting a used mac or a new but older generation mac. m1 to m4 is not that different in performance to justify its price hike.
I'm not in college. I'm working, but my job is economic policy research. so, academia work is not really that different from mine. fashion statement doesn't really matter to me?alright. that's one for not worth it. alright, that's two for not worth it. I'm not looking for gaming laptop to be honest. If it is gaming, I already have ASUS TUF (2020) working just fine. For on board AI? the VRAM is just not sufficient. I don't see any productivity increase that necessitate an upgrade on my GPU on laptop tbh. It's too expensive for no productivity increase. Not that macbook can run a decent sized LLM with that 24GB of RAM. yeah. well... for excel, it's pretty slow on mine though. in my other thread, my use case for excel is heavy on computation too. excel don't have GPU support at all.
thanks for the @ I haven't touched the M-series chips of mac, The last Mac I touched before is the one with 2 Redon 7 GPU. So I can't really say for sure. If you're fine with me, you can send me a copy of your file and tell me what work you're supposed to do, I will analyze the bottlenecks of the system or the python venv or the codes. Also, If I remember correctly, you're using 11th gen intel, right? You can maybe turn on or turn off the avx512 and check if it makes a difference, I remember the avx512 implementation on 11th gen wasn't good, as it give out more heat. It shouldn't be like 20 over times the difference if you got the same venv? Erm, I must say, my frequency of changing computer is really higher, I think when I get 11th gen intel, I got a 5800x after that, then I got a 12650h, and a 12900h. And if I didn't overspend on camera and I didn't blow something up big times, I would have got another if not for the AMD bad chipset and Intel horrible chips. So personally, I would just buy a new laptop and maybe an additional desktop while doing it. The issues with Mac are that people usually still need a Windows PC while having a Mac, Many of times that Mac works well on paper, example photoshop and premiers pro, but many companies still throw them in mass and get back to windows because there were many issues in work cycle, there were many small problem. Kinda like my dad, sure you can run Autodesk on non-qudro cards, but what about the many plugins that have not updated in decades. So he sticks with AMD till he quit his job. After all, your PC is already 5 years old. So it's your call, Personally I think your laptop is still a solid machine for most people. Though I thought I was the only one who buys computer in the crypto mining chip shortage, ouch.
@xiazixin thankfully the model is open source. https://github.com/EAPD-DRB/OG-IDN I don't do any modification to the required packages. base conda + everything in environments.yml is what I need. Running python scripts in the examples folder should be representative of my work tasks. For me, it's a solid machine in majority of my tasks, and it's really struggling doing my CGE models. The maintainer said that it should be able to be run in windows, osx and linux. neither wsl and windows work here.
Bro, I was referring to things like upgrading the storage to an SSD or even improving cooling systems. He can also consider external GPUs or cloud-based services for intensive tasks like modeling. The goal is to make sure his current setup work as efficiently as possible for his needs
Did you use miniconda or did you use the ultra bloated fatty anaconda? I'll try the build on spare time, though I really don't want to get the whole shit of anaconda and all other packages on my computer...
I can't seems to find a pc that don't have tons of stuff in it that I can just dumped the anaconda in it. I thought it would be sufficient with venv...