the most reason i like m1 is battery, it may be not superb like icore 7 or more ram.. But i got imac on whatever missing on m1
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
8GB RAM is a disappointment?
- Thread starter MacBH928
- Start date
- Sort by reaction score
how much do you allocate RAM to Windows VM? Which virtualization software you use? I used 4GB RAM for a linux distro and it was moving in slow motion.
I don't understand computers too much but it will be cool if they can do SSD's like a slip-in card for easy replacement. Kind of like Wifi cards back in the day.
I recently designed a product myself for retail sales and designed it so people can actually replace parts or add parts. I hate it when manufacturers sell you something that can’t be modified or allows end users to replace/modify parts themselves.
That comes from Steve’s appliance design strategy. He viewed computers like appliances. You get what you need now, and then a few years later if you want something better you just get a new one.
In Steve’s defense, it wasn’t just about profits. It was about making sure that every component that went into the machine was curated by Apple to ensure the best experience. Steve wanted customers to know that no matter what Apple product you use, there is a baseline standard that you know will be met because you can’t throw in a cheap after market piece. It also helps with tech support because it eliminates many of the variables. Of course we all know the drawbacks to this design method, but Steve always wanted his products to be sold like appliances.
Out of pure curiosity - what is Google.ads website and why would someone use it... or have 20 tabs with it?
what are you talking about ?
ipad os is build on the same base as mac os , sounds like you just pretend to know what your talking about and set expectations what weren't met. do some research.
It is for creating ads in Google search.
You can easily have 20 tabs on it if you are managing multiple websites and ad accounts as an agency or even as someone working internally as it is a Pro tool for ad managers.
Google Ads - Get Customers and Sell More with Online Advertising
Discover how online advertising with Google Ads can help grow your business. Get customers and sell more with our digital advertising platform.
ads.google.com
As a feedback after a few months at this. 8gb is enough, since it is still so fast and not choking while managing many tabs, but you definitely need to be aware of how many tabs you are using when using heavy work-related web apps (and there are lots of them now). But it has not become a problem so far.
Intel with 16gb choke on Google Ads too.
More and more people will be using Web Apps rather than desktop apps with the trend of remote work.
And with these heavy webs app, you cannot just block them because they use more ram. Plus you can easily replicate their ram usage since you can have 10 tabs of the same web app running. Which makes it heavier than pro Desktop apps since their window is their browser where anything can essentially run.
Web apps being in the browser can ingest more data quite quickly and browsers like Safari have no choice but to accept it and just say "The website is using significant memory".
So yeah go for more ram. (even though 8gb is fine)
Last edited:
It is for creating ads in Google search.
You can easily have 20 tabs on it if you are managing multiple websites and ad accounts as an agency or even as someone working internally as it is a Pro tool for ad managers.
Google Ads - Get Customers and Sell More with Online Advertising
Discover how online advertising with Google Ads can help grow your business. Get customers and sell more with our digital advertising platform.
As a feedback after a few months at this. 8gb is enough, since it is still so fast and not choking while managing many tabs, but you definitely need to be aware of how many tabs you are using when using heavy work-related web apps (and there are lots of them now). But it has not become a problem so far.
Intel with 16gb choke on Google Ads too.
More and more people will be using Web Apps rather than desktop apps with the trend of remote work.
And with these heavy webs app, you cannot just block them because they use more ram. Plus you can easily replicate their ram usage since you can have 10 tabs of the same web app running. Which makes it heavier than pro Desktop apps since their window is their browser where anything can essentially run.
Web apps being in the browser can ingest more data quite quickly and browsers like Safari have no choice but to accept it and just say "The website is using significant memory".
So yeah go for more ram. (even though 8gb is fine)
Thanks.. Sorry I will be for skipping your ads.
UBlock Origin all the way....
I have an i9 and i7 MBP for work, my personal M1 shreds them in performance.
Even basic things like number factoring a large semi-prime: 46154213179399394382363686716564958816430153059052298683641522409028243219453.
Bash:
Intel i7-8850H (12) @ 2.60GHz:
CPU times: user 4min 59s,
Wall time: 5min
206424722777657687138653749918290748733 * 223588592288495400433214196823678395841
Apple M1:
CPU times: user 3min 41s,
Wall time: 3min 42s
206424722777657687138653749918290748733 * 223588592288495400433214196823678395841
Last edited:
performance when i do Xcode not quite yet crash frequently even simple project compare to iMac 2017. But it's just 3 month appear and 1 month i used it so hope more better in 6 month performance.
8GB is fine for consumer use, however I'm a Cloud Engineer and I got the Air base model for now because I wanted to sell my 16" before the new ones.. I find it that it works however if I have 10 tabs open on 2 of my browsers plus Visual Code, Mail, Outlook and rest of Office 365 + Teams (which is still using rosetta and consuming tons of memory) the swap moving will start to happen around 8GB per day.
That's also debatable. You don't have to be an engineer to "need" a lot of RAM.
"Thanks" to tabbed browsing, even a lay person can have multitudes of tabs opened at once on their browser. Add on the behavior of "closing" apps on MacOS is not the same as quitting the app, even a lay person can have a bunch of stuff opened without realizing it.
Even if you have a million of tabs open there is no reason why they have to be resident in the RAM. Swapping in contents of a tab from an SSD is faster than the tab switch animation anyway.
Sure, but with how the aggressive swap mechanism works in MacOS with the M1 it's usually more than enough again for "casual consumer" if you are someone who work with Photoshop or say a lot of web apps like Google Ads where you need multiple tabs or even say something like Notion or a mood board you already in to get 16GB since you know you gonna have tons of tabs open.
I'm very happy with this base model but can't wait to upgrade to a M1X or M2 because I do need dual displays and more RAM.
Not really the point.
My point was, it's not just engineers or professional photographer/videographer who need more than 8GB of RAM.
And I don’t disagree. Just pointing out that one doesn’t need a lot of RAM to have many browser tabs or apps open - just an agile memory management.
I have north of 100 tabs on safar in 1 window, on 8Gb, not a hitch
Looks like you want to do pro work on consumer device. the Pro M1s are not out yet.
yea and the current M1 are not for them, the 8GB is for the average school teenager, and general office stuff. Its not for creating 3D models in Maya
Even my MSI laptop has 8GB of RAM and it has been able to handle everything that I throw at it. Any game or Visual Studio.
Same with Macs.
I remember back in 2014 my classmate said that 2GB of RAM isn't enough for phones too, but iPhone 5s had only 1GB and I never had any issues.
Everybody's so obsessed with RAM.
From PC side I'd say that processor matters way more. My MSI has 8GB of RAM and Ryzen processor and my work HP has some Intel processor, but my MSI is way faster.
Same with Macs.
I remember back in 2014 my classmate said that 2GB of RAM isn't enough for phones too, but iPhone 5s had only 1GB and I never had any issues.
Everybody's so obsessed with RAM.
From PC side I'd say that processor matters way more. My MSI has 8GB of RAM and Ryzen processor and my work HP has some Intel processor, but my MSI is way faster.
I don't know how unified memory works on the AS machines, but wouldn't you need more conventional RAM to make up for the fact that you don't have any dedicated GPU RAM? I.e., consider a current 16" MBP with 16 GB of (CPU) RAM, plus a GPU with 4 GB or 8 GB of dedicated RAM. That's a total of 20 or 24 GB RAM for the GPU and CPU. Wouldn't an AS MBP, whose RAM needs to support both the CPU and GPU, need more than 16 GB RAM to perform equivalently (by equivalently I mean handle as many tasks before experiencing memory pressure)?
For instance, imagine you're inverting a matrix that, on a 16" Intel MBP with 16 GB (CPU) RAM, takes up most of the available RAM (available = what's left of the 16 GB after you take care of the OS and whatever app you're using to do the inversion). Now suppose you want to do that same matrix inversion on an AS MBP with 16 GB RAM. Won't you run into a problem because now that 16 GB of RAM, in addition to containing the OS, your matrix inversion app, and the matrix itself, also needs to support the RAM needs of the GPU?
[Granted, a matrix inversion done in the CPU doesn't put much demand on the GPU. But imagine you were doing a visualization that put significant RAM demands on both the CPU and GPU.]
For instance, imagine you're inverting a matrix that, on a 16" Intel MBP with 16 GB (CPU) RAM, takes up most of the available RAM (available = what's left of the 16 GB after you take care of the OS and whatever app you're using to do the inversion). Now suppose you want to do that same matrix inversion on an AS MBP with 16 GB RAM. Won't you run into a problem because now that 16 GB of RAM, in addition to containing the OS, your matrix inversion app, and the matrix itself, also needs to support the RAM needs of the GPU?
[Granted, a matrix inversion done in the CPU doesn't put much demand on the GPU. But imagine you were doing a visualization that put significant RAM demands on both the CPU and GPU.]
Last edited:
The GPU VRAM doesn't add to any usable memory that matters to the CPU.
You can imagine this scenario:
Suppose you have 1 GB of data that you need the GPU to process. In a dGPU scenario, you would have to move that data from disk storage into system RAM. Once the data is in RAM, you then need to move than data to the GPU VRAM via PCIe. At this point, your system RAM is down 1GB and your GPU VRAM is also down 1GB. When the GPU is done processing, you then have to copy the result from the GPU VRAM back to your system RAM. Assuming 1GB of data result in exactly 1GB of results, with no growth or shrinkage, net effect is that your system memory usage is the same regardless of how big your dGPU VRAM. In fact, you are actually limited by the amount of dGPU's VRAM if you want to use the GPU to process your data.
Compare this to unified memory. The copying to and fro from VRAM is skipped as the GPU can read from the unified RAM directly. The GPU in this instance can potentially use all available system RAM as it's data source. So a theoretical AS Mac Pro with 1TB of RAM can dedicate most of it to the GPU for processing. I think we only have 32GB VRAM dGPUs at the moment? So there'll be a lot of copying thru and fro.
The PCIe bus will definitely hobble the thruput, but it should be mitigated by the higher bandwidth of the dGPU VRAM. With Apple going to higher bandwidth bus, it should increase their GPU thruputs as well.
So with the next Mx SOCs that's guaranteed to have better GPUs (which the A15 has shown), Apple will definitely use higher bandwidth memories and busses to keep them fed with data.
Thanks for taking the time to write that explanation. What you wrote makes sense based on the RAM usage scenario you portrayed. Essentially you are saying that, with conventional architecture, any memory usage by the VRAM also requires the same memory usage by RAM. So having separate VRAM doesn't add any additional capacity.
But aren't there use cases that would require significant VRAM without also requiring usage of that same amount of RAM, i.e. where the GPU was using a lot of VRAM for its own internal processing? E.g., suppose you send 1 MB of data to the GPU with a processing request (say, run a simulation on that data and give the result)*. It then processes that internally, requring 4 GB of VRAM to do so. Then once it's done it sends the answer back to the CPU, where the answer is 2 MB in size.
If the CPU needs 16 GB RAM while this is going on, then total memory usage is 16 GB RAM + 4 GB VRAM.
But with UMA (unified memory architecture), total memory usage during this process would be 20 GB RAM.
*Not saying things would actually work this way; just trying to illustrate a concept.
Last edited: