iMac 2020 i9 with cheapest GPU - am I crazy?

[ADGrant]

This thread compares a couple of sizes : https://forums.macrumors.com/thread...er.2250057/page-2?post=28830312#post-28830312

But blackmagic only gives the sequential read write, which happens nearly *never* in real world scenarios.

To measure impact on software development, IOPS and 4K read/write are more significant I think.

Thanks, so according to the benchmarks in that thread, the performance curve seems fairly flat after 2TB but between 512 and 2TB we are looking at sequential write speed improvements of 400mb to 800mb per second depending on which 512 benchmark you are looking at. I didn't see any 1TB benchmarks in the thread but I would definitely avoid the 512GB SSD.

On a machine with a large amount of RAM, there is no reason a compiler couldn't read the entire source file into memory to parse it and write out the entire object file from memory (I do that when reading files off an HDFS cluster). That said, sequential vs random file access is more of an issue with hard drives than SSDs.

 

[pldelisle]

I would definitely avoid the 512GB SSD.

Me too.

On a machine with a large amount of RAM, there is no reason a compiler couldn't read the entire source file into memory to parse it and write out the entire object file from memory

Of course, it’s always in RAM. But did you ever compiled Tensorflow for instance ? On my MBP 15 2013, it could take 2h to compile on 8 threads And uses easily all available memory. It’s pretty impressive. Even if the entire lib can fit in RAM, the compiler makes use of the available RAM and cores/threads. I think the difference in I/O between 1 TB and 2 TB is negligeable versus the gain 2 additional cores can bring.

 

[pldelisle]

Which SSD Apple uses ? We might be able to find benchmarks of the SSD they use in retail equivalent models.

 

[filmgirl]

So my workflow is very, very similar — with the exception that I also do video and OBS, so I went for the 5700XT. I’ve got the i9/5700XT/2TB SSD and I wound up with 128GB of RAM.

I’d personally upgrade to 1TB SSD just because for VMs and containers you’ll want it — but it looks like it’s a $200 difference between the 5300 and 5500XT.

Since you won’t be gaming, that should be fine. It’s not like we can do CUDA on our Macs anyway (thanks, Apple!), so I’m less concerned about the differences between 4GB and 8GB for compile and compute. Obviously, if you can get $200 more (or whatever the equivalent is in Euros), even better, but you could always get an external GPU down the road too.

I won’t wade in too much on the SSD benchmarks because I just don’t know. Apple used to have some of the fastest NVMe options but that’s not true anymore compared to higher-end PCs/laptops, but they are still fast. I don’t know of the capacity between 512 and 1TB would make more of a difference over more cores — my suggestion for 1TB over 512 is simply about space for your containers and VMs (as well as Boot Camp if you want to use that).

 

[pldelisle]

I’d personally upgrade to 1TB SSD just because for VMs and containers you’ll want it

Definitively.

It’s not like we can do CUDA on our Macs anyway (thanks, Apple!)

Yeah ... I still believe the best Mac ever would have been with Nvidia GPUs. But I’m feeling better knowing that in less than 2 years even AMD will (probably) be history on Mac.

but you could always get an external GPU down the road too.

Only a patch.


Attachment would be my Mac. 5700 to play with Metal and having fun but definitively not necessary. Would go with 5500XT if I would be on a tight budget.

 

[ADGrant]

Attachment would be my Mac. 5700 to play with Metal and having fun but definitively not necessary. Would go with 5500XT if I would be on a tight budget.

I would be interested in seeing how quickly your configuration can build the test project that Max Tech used (and how quickly it could build Tensor flow).

 

[vjson]

I would be interested in seeing how quickly your configuration can build the test project that Max Tech used (and how quickly it could build Tensor flow).

Yes, I agree it would be very interesting to see if your machine also outperforms Max Tech in that benchmark, and if so by how much!

 

[pldelisle]

I would be interested in seeing how quickly your configuration can build the test project that Max Tech used (and how quickly it could build Tensor flow).

Unfortunately I’m still too poor to afford this 😂

Grad school is expansive😅

But soon I will 😎

 

[vjson]

Unfortunately I’m still too poor to afford this 😂

My bad, I didn't read your post carefully enough when you wrote that your mac "would be...".

 

[highlypaid]

Hi everyone,

To begin, I want to make it clear to anyone reading this thread that I have probably watched ALL reviews of the iMac 2020 27" on youtube, and most of the various threads on these forums regarding the upgrade options and performance benchmarks for this computer. Seriously, I've spent so many hours researching my upcoming iMac purchase that it's embarrassing to even admit.

Now that we got that out of the way..!

I'm a software developer using a 2011 27" iMac that I over the past two years have upgraded with a 3.4ghz i7, 500gb ssd, 16gb ram and a K2100M gpu, and the system is honestly working surprisingly well for its age. Thanks to the switched out gpu I've managed to install Catalina and can therefore use XCode 11. However, I have random crashes and reboots when the computer has been asleep for a while, I'll most likely not be able to install Big Sur and the new XCode 12, and the cpu is ridiculously slow compared to intel's 10th gen. In other words, it's time to buy a new computer.

Since I'm a software developer, compile times and being able to run VMs, many docker containers and multiple IDEs with large projects simultaneously is what I want my new iMac to handle flawlessly. Therefore I'm definitely going with the i9 over the i7. I don't care at all that the maximum performance gain I'll get from this probably is no more than 10% (and only when the workload is parallelised and heavy), because that's still a performance gain. So saving money by going i7 and instead upgrading the gpu because it's more bang for the buck is irrelevant in my case, because highly doubt the faster gpu will help with compile times, etc.

Furthermore, I do zero (!) video editing, zero gaming (might try some games in boot camp for fun though), and after selling my pro camera equipment I only shoot with my iphone nowadays, so no RAW editing in LR/PS either.

Based on all these parameters, wouldn't the smartest config for me to order be the mid tier iMac with the Radeon 5300 gpu, and just upgrade the cpu, storage and third-party ram? I have yet to see a single person go for this kind of config, with the most expensive cpu and the cheapest gpu, as every single person who owns an iMac seems to be a pro video editor and only care about final cut render times🤨

Correct me if I'm wrong, but the Radeon 5300 should be powerful enough to drive the built-in display, playback 4k videos, with zero problems, right? And in that case, why even spend money for the pricier top tier iMac, where the only difference from my config would be a mandatory 5500XT over the 5300?

The only downside I can imagine might be a lower resale value due to most people wanting those beefier gpus, but other than that, am I missing something? Will the 5300 in any way bottleneck the cpu for the type of work I do? Or might it even be the other way around, that the weaker gpu gives more "room" for the i9 when it's being pushed to its limits?

I'm happy that more software developers such as myself are getting into the debate. People like Max Tech are extremely awesome, but they do video editing. I am thinking of running tests running Docker, Kubernetes clusters, compiling the Linux kernel, etc., on the i9 and uploading to Youtube.

That said, I wouldn't necessarily recommend getting the lowest end gpu. I have the i9, 5700XT and I'm a software engineer + UI designer. I'm also getting into 3d modeling. You never know when you might want to do some hardcore graphics stuff, so I'd at least throw down for the 5700 $300 upgrade ($270 with edu discount).

 

[highlypaid]

Sequential speed is nothing when compiling. Yes, there are a couple of hundred of MB/s more in 1 TB vs 512 GB, and probably even less difference between 1 and 2 TB, but this is in sequential speed. The difference is very minimal when comparing real world scenarios. Code files are small, very small, and has nothing to do with sequential speeds.

But compiling with two more cores can reduce compile time by a lot and can, at the end of the week, maybe save an hour or two.
[automerge]1599268137[/automerge]

I‘m machine learning research engineer. Python is my main language.

Sorry I thought you were quoting me ahah

At $100/hour, that's $400-$800/month saved. So definitely worth it. Even at $50/hour, that's $200-$400/month.

 

[wardie]

At $100/hour, that's $400-$800/month saved. So definitely worth it. Even at $50/hour, that's $200-$400/month.

Only if you can’t productively use the time otherwise. When I were a lad, you could lose a whole day through messing up your overnight batch recompilation... but I don’t remember getting paid less overall. It’s really hard to accurately price efficiencies like that. But if it helps people reduce their cognitive dissonance to justify a purchase they want to make anyway, so be it. I’ve done that a lot and seem to end up with lots of toys .

 

[cltd]

The only downside I can imagine might be a lower resale value due to most people wanting those beefier gpus, but other than that, am I missing something?

Forget that factor. When ARM Macs hit the market, Intel Macs resale value will be near to none, as it was with G4/G5 Maca during MacIntel era.

 

[wavesm]

Forget that factor. When ARM Macs hit the market, Intel Macs resale value will be near to none, as it was with G4/G5 Maca during MacIntel era.

I don’t think comparison with G4/G5 is a proper one.

There was nothing new that could be ran with those.

Intels will still run latest Windows or Linux.

 

[r6mile]

I don’t think comparison with G4/G5 is a proper one.

There was nothing new that could be ran with those.

Intels will still run latest Windows or Linux.

You could always run Linux on PPCs, still can. Didn't stop them from losing value. And surely someone intending to just run Windows will just buy a PC?

 

[cltd]

Intels will still run latest Windows or Linux.

People rather buy PC to run Windows, for a friction of Mac price. And they can change CPU, GPU, fan etc etc...

 

[ADGrant]

You could always run Linux on PPCs, still can. Didn't stop them from losing value. And surely someone intending to just run Windows will just buy a PC?

Who said anything about just running Windows?
[automerge]1599322336[/automerge]

I don’t think comparison with G4/G5 is a proper one.

There was nothing new that could be ran with those.

Intels will still run latest Windows or Linux.

Agreed, the G4/G5 Macs weren’t very useful outside the Apple Ecosystem.

 

[ADGrant]

Only if you can’t productively use the time otherwise. When I were a lad, you could lose a whole day through messing up your overnight batch recompilation... but I don’t remember getting paid less overall. It’s really hard to accurately price efficiencies like that. But if it helps people reduce their cognitive dissonance to justify a purchase they want to make anyway, so be it. I’ve done that a lot and seem to end up with lots of toys .

It comes down to how much you value your time and how much money you have. If you have a lot of money you might hire a house keeper to clean for you and dine out so you don’t have to cook and wash the dishes. You might pay someone else to paint your living room. You may not wash your own car (unless you really love your car I guess)

Similarly someone using their computer to “get things done” presumably wants to get those things done sooner rather than later irrespective of how much they are being paid to get those things done. If they have the money to afford the faster CPU or larger SSD, they should buy the faster CPU or SSD.

As for software build times, sooner is always better than later. The Go programming language was developed because the creators got sick of waiting for C++ builds to complete. If you are debugging an issue and want to test a fix, you don’t want to wait a couple of hours for the fix to build. Then there is always the risk that your concentration will wander and you will find yourself on a internet forum...

 

[rkuo]

Faster GPU also just makes the UI a bit snappier, and you feel that every second you use the computer. It’s not a terribly hard sell. I would for sure go to the 5700 at least since according to https://support.apple.com/en-us/HT211281 it’s better at driving more external monitors.