Decision time i7 vs i9 for new 2020 iMac?

[BuCkDoG]

For me money is not an object and I actually decided on the i7. Multiple reviews have shown that there is basically no performance difference between the two processors, and then it makes no sense to spend $500 on the i9.
This is a typical example of people being scared that they might be missing out by only getting the i7 - But when tests show that there is barely any difference, then that should be enough to decide. Ram and GPU is different in the way that those two are usually the biggest limitations when a computer grows old. Ram can thankfully be upgraded in the iMac so that is not an issue. The GPU though is what I would always go for the biggest option available if it provides a decent performance increase. Tests of the 5700 XT compared to the regular 5700 shows 15-20% performance difference, in which case it's definitely worth it to get.

People tend to greatly overestimate how much CPU power they need. For 99,9% of all use-cases, the CPU is never the limiting factor. I'm definitely what you would call a power user - I do everything from video editing, photography, development, security research running multiple VM's, some 3D modeling and I tend to have chrome running with 500+ tabs opened for months at a time. I also like to play games once in a while. Currently I'm running a 2013 iMac. The big limitation in that machine is not the CPU, but the GPU and partially the SSD speed. Sure a new CPU will be faster, but unless a task really uses all cores for an extended period you will only feel tiny differences.
You have to keep in mind that the CPU rarely runs anywhere close to 100% and rarely uses all cores, which is why a theoretical CPU increase has very little every day effect. And when the i7 gives you more or less identical performance to the i9, then paying $500 for the i9 seems silly. But upgrading the GPU on the other hand, that will definitely make a huge difference for the longevity of the machine.
A lot of things that used to be CPU based is also being moved to the GPU today, which is why it makes a bigger difference. Video rendering, 3D rendering and a lot of other stuff is offloaded to the GPU and also the T2 chip(for HEVC).

People are thinking so much about the i9 because in their mind it has 20% more cores and therefor should be quite a lot faster than the i7. But this is why we have benchmarks and real-world testing - Because performance is not a linear thing.

I couldn’t agree more with your statement and assessment on this.

 

[leifp]

I also got the i9 10 core, 4TB SSD, 5700xt and 10GbE with 128GB of RAM from HyperX (Kingston). Yes, it cost a lot of money. I did the 12 month 0% interest with the Apple Card. I don't want to ever wonder what I could've done or should've done. I want this to last me a long time and have no regrets. To each his own. I plan on splitting that SSD right down the middle to run Windows 10 on Bootcamp, and I also run Parallels with multiple VMs. I intend to do some serious gaming as well. About time there was a Mac that could do this and not cost $10K.

This is exactly what I did, and exactly why I did it. Although I used Samsung RAM and VMWare due to a spat with Parallels years ago. I think Parallels is likely superior... although the new VMWare (12) ought to remove many of my issues/concerns.

 

[BuCkDoG]

This is exactly what I did, and exactly why I did it. Although I used Samsung RAM and VMWare due to a spat with Parallels years ago. I think Parallels is likely superior... although the new VMWare (12) ought to remove many of my issues/concerns.

You also intend to do some serious gaming on your iMac? 😂😂😂

 

[leifp]

You also intend to do some serious gaming on your iMac? 😂😂😂

Yes. I eGPU and Bootcamp. Sometimes I game right in macOS. That seem strange to you? Impossible?

 

[BuCkDoG]

Yes. I eGPU and Bootcamp. Sometimes I game right in macOS. That seem strange to you? Impossible?

It’s just weird why you would spend that kind of money to game on your iMac. You can easily just buy a windows computer and a higher refresh rate monitor and enjoy your serious gaming much better. Especially since your spending money on an eGPU as well which has limited bandwidth naturally. To each it’s own though.

 

[leifp]

It’s just weird why you would spend that kind of money to game on your iMac. You can easily just buy a windows computer and a higher refresh rate monitor and enjoy your serious gaming much better. Especially since your spending money on an eGPU as well which has limited bandwidth naturally. To each it’s own though.

There much of a price difference between a fully loaded iMac and an iMac + gaming PC? Pretty sure there isn't (I have that setup at another location: OG 5K iMac + 1080ti gaming PC). I achieve 80% performance +/- for negligible price difference but >50% space reduction. I'll pay a bit for that privilege. That is very much to each their own...

 

[BuCkDoG]

There much of a price difference between a fully loaded iMac and an iMac + gaming PC? Pretty sure there isn't (I have that setup at another location: OG 5K iMac + 1080ti gaming PC). I achieve 80% performance +/- for negligible price difference but >50% space reduction. I'll pay a bit for that privilege. That is very much to each their own...

Enjoy your setup. 😂😂😂

 

[leifp]

Enjoy your setup. 😂😂😂

Aside: I prefer the quality of the iMac monitor over the refresh rate of my RoG Swift on the WinPC. I'd love if there were a monitor that hit "best of all worlds" but I haven't even come close to finding it yet...

 

[jasonefmonk]

Aside: I prefer the quality of the iMac monitor over the refresh rate of my RoG Swift on the WinPC. I'd love if there were a monitor that hit "best of all worlds" but I haven't even come close to finding it yet...

Yeah, the wide gamut, large, single panel, at “retina” PPCM is really special and overlooked by many PC-first players. I’m not sure games like Civilization have ever looked better. I know high and variable refresh rate is awesome, but I don’t see it as more desirable than other monitor qualities.

 

[leifp]

Yeah, the wide gamut, large, single panel, at “retina” PPCM is really special and overlooked by many PC-first players. I’m not sure games like Civilization have ever looked better. I know high and variable refresh rate is awesome, but I don’t see it as more desirable than other monitor qualities.

I don't play "twitch"/FPS games, so I do prefer those elements as well. Mainly TP/RPG. I had a 32" 4K external as well but I did not appreciate it anywhere near as much as the internal 5K. Don't think I'm not saddened by that! Much easier to drive 4K than 5K...

 

[cltd]

As I read forums and see youtube tests, base models don't exist. Does anyone consider i5 3.3 ghz?

 

[leifp]

As I read forums and see youtube tests, base models don't exist. Does anyone consider i5 3.3 ghz?

Absolutely! I've even seen reviews on it (although they really are rare). And if I were not attempting to combine a Mac and a WinPC in one device, that would likely be my choice. My quad core i7 is plenty of machine for me! Except the GPU, which really wasn't. Even when using it for non-gaming applications...

I've hedged my bets with this iMac; the 2014 5K has lasted me near six years and I'm still happy with it [except aforementioned caveat]. (Photography got me to buy it, and its gorgeous screen has kept me enraptured)

 

[BuCkDoG]

Aside: I prefer the quality of the iMac monitor over the refresh rate of my RoG Swift on the WinPC. I'd love if there were a monitor that hit "best of all worlds" but I haven't even come close to finding it yet...

I will agree for sure that the 5k panel that does come with the iMac is something special to enjoy. If your playing games at 5K window size and most likely 2560x1440p render scale or depending on the game 5K render scale with most settings probably set to low, the sharpness is just incredible.

 

[RAWvJPG]

As I read forums and see youtube tests, base models don't exist. Does anyone consider i5 3.3 ghz?

I have the i5 3.3Ghz. Runs well, but just a hot and loud as the top tier models it seems.

 

[Mizouse]

As I read forums and see youtube tests, base models don't exist. Does anyone consider i5 3.3 ghz?

The annoying thing about the base i5 3.3 is you cant upgrade the video card should you want to.

Also with the even more base i5 3.1ghz you cant even upgrade the storage. its stuck at 256GB

 

[cltd]

I have the i5 3.3Ghz. Runs well, but just a hot and loud as the top tier models it seems.

How so? Shouldn't be quiet and colder? As I understood forums, thermal footprint is main advantage of the i5 iMac.

 

[pldelisle]

How so? Shouldn't be quiet and colder? As I understood forums, thermal footprint is main advantage of the i5 iMac.

No. The temperature is the independent variable, while the clock speed is the dependant variable. The CPU has a target temperature of 100 degrees. It will adjust its clock speed all the way up until it reaches the max temperature. Once it reaches it, it lowers the clock speed. The i9 will just lower it just a bit further than the i7 but will still have 2 cores more running full blown. Pretty much the same for the i5, but with fewer number of cores.

 

[Spinnetti]

I've been trying decide between those two as well. I've now seen multiple reviews that show the i7 to perform better. Apps that only use single core benefit from the higher frequency. The i9 seems to slow down more than the i7 when running heavy loads over time, causing the i7 to run at a higher frequency. So I think only very specific situations would cause the i9 to be a better choice.
So for me it's definitely the i7, 5700 XT Pro, 1TB SSD, 10Gbit Ethernet and then getting 64 or 128 GB ram on the side. I'm also passing on nano-texture since reviews seems to indicate the image is less sharp.

This is where I landed too... typed on that machine I just added 32mb ram to the existing 8. I sincerely doubt I'll ever use even that, even in coding or CAD. The i9 def. not worth it over the i7 on a rational basis unless you are literally in a production condition 24x7 and fully billing each of those hours.

 

[iDomDotCom]

Enjoy your setup. 😂😂😂

Don't be passive-aggressive. Most of us use our Macs for work (because we prefer the OS) but would like to game without buying a separate device. Just because you buy your computer for solely pleasure, doesn't mean everyone does.

 

[rkuo]

How much the CPU is used greatly depends on the application and how well it's optimized. Most rendering processes are being offloaded to the GPU and to the T2, which is also why the actual CPU usually doesn't make much difference in most situations. The i9 is a maximum of around 10% faster which has been shown in many benchmarks. That means that in a purely CPU dependent workload with all cores running at 100%, you will see no more than a 10% difference(In most situations it would probably be no more than a few percent).
But when you look at most video editing software they use the GPU much more than the actual CPU, meaning that in those situations there will be absolutely no performance difference between the i7 and the i9. And if you export HEVC then the limitation is the T2 chip which handles that part - And that one is identical between the two machines.

When people are discussing the i7 vs the i9, there is a lot of focus on stuff like video rendering and similar. I'm guessing that 99% of all people in this forum don't ever do video editing or similar tasks. But people tend to forget that when you actually do video editing, 3D modelling and similar heavy tasks - You spend 95% or your time on editing, finding clips, deciding on the story etc. All things where the CPU is only used lightly. So you spend 8 hours editing a video but then people worry about saving 15 seconds on a video export which takes 8 minutes. So unless you do non-stop video export or you export super hi-res formats such as 6K or 8K RED footage, then you won't see any difference and even with those formats you will barely be able to see a difference.

So the additional $500 is for peace of mind and not actual performance. So if you feel better by spending the $500, then do so. But don't do it because you think you're going to get a good boost to performance.

I agree most people aren't going to see the benefit of the 10 core. But I think the benchmarks have actually been pretty favorable for it.

It looks like the power limits have been raised quite a bit in the 2020 imac and thus it's quite good in bursty multi core workloads. Additionally, if not hammering both the CPU and GPU at once, you can still get a lot of the benefit out of the 10 core even though it runs pretty hot.

Since I'm working from home a lot, I have to turn on my business brain and realize that even tho 400 dollars is a typical overpriced upgrade from Apple, it only takes a couple of hours shaved off compile times off of the course of the machine to pay that back. (one of the more obvious areas where more cores scales well)

 

[mlykke]

I agree most people aren't going to see the benefit of the 10 core. But I think the benchmarks have actually been pretty favorable for it.

What kind of benchmarks have you seen? Of all the benchmarks and reviews I've seen so far, in just about every single test the i7 is equal in performance to the i9. Especially when you look at real world tests and not the raw Geekbench scores.
Only one that comes to mind that had a slight difference was a test for Logic, where the i9 handled a few tracks more than the i7 - Something around 127 vs 131 tracks - An increase of only 3%.
The only tests I've seen where there is a difference is in GPU related tests where the i9 sometimes have been compared to an i7 with a smaller GPU.

As an example check out this short benchmarking video. The only tests showing a difference are the ones involving the GPU - And the i7 he's testing only has the 5500 compared to the 5700 XT in the i9, which explains the difference. A lot of the other tests are completely identical.

 

[thermalthrottle]

What kind of benchmarks have you seen? Of all the benchmarks and reviews I've seen so far, in just about every single test the i7 is equal in performance to the i9. Especially when you look at real world tests and not the raw Geekbench scores.
Only one that comes to mind that had a slight difference was a test for Logic, where the i9 handled a few tracks more than the i7 - Something around 127 vs 131 tracks - An increase of only 3%.
The only tests I've seen where there is a difference is in GPU related tests where the i9 sometimes have been compared to an i7 with a smaller GPU.

As an example check out this short benchmarking video. The only tests showing a difference are the ones involving the GPU - And the i7 he's testing only has the 5500 compared to the 5700 XT in the i9, which explains the difference. A lot of the other tests are completely identical.

That guy really needs to learn how to present data correctly. The graph he made at 2:20 showing the red raw export times looks like the difference between the i7 and i9 is around 20/25% going by the size of the bars and the scale along the horizontal axis in 0.25 increments. But upon closer inspection the data is measured in minutes and seconds so the difference is 1 minute 25 seconds vs 1 minute 1 second, which on a graph should look like the i7 bar is 42% longer then the i9 seeing as 25 seconds is approximately an extra 42% percent of 1 minute. He's presented 1 minute as 4 x 0.25 segments and 25 seconds as another 0.25 segment when it should be almost 0.5 on that scale of his.

In other words, the i7 5500xt will be 42% slower at exporting red raw 4k to pro res compared to the i9 5700 xt. Good to know and I would have preferred he ran similar tests using other codecs such as canon raw rather then post loads of graphs of h264 and h265 exports which are both handled by the intel integrated graphics and t2 chip respectively therefore negligible difference as both systems have those chips.

 

[BuCkDoG]

Don't be passive-aggressive. Most of us use our Macs for work (because we prefer the OS) but would like to game without buying a separate device. Just because you buy your computer for solely pleasure, doesn't mean everyone does.

Don't tell me what and how to act. I can state my case and opinion as much as I want and If you don't like it, too bad.

 

[ADGrant]

What kind of benchmarks have you seen? Of all the benchmarks and reviews I've seen so far, in just about every single test the i7 is equal in performance to the i9. Especially when you look at real world tests and not the raw Geekbench scores.
Only one that comes to mind that had a slight difference was a test for Logic, where the i9 handled a few tracks more than the i7 - Something around 127 vs 131 tracks - An increase of only 3%.
The only tests I've seen where there is a difference is in GPU related tests where the i9 sometimes have been compared to an i7 with a smaller GPU.

As an example check out this short benchmarking video. The only tests showing a difference are the ones involving the GPU - And the i7 he's testing only has the 5500 compared to the 5700 XT in the i9, which explains the difference. A lot of the other tests are completely identical.

That was a pointless video. He was benchmarking video encoding workloads that don't use the GPU or the CPU heavily.

 

[ADGrant]

How much the CPU is used greatly depends on the application and how well it's optimized. Most rendering processes are being offloaded to the GPU and to the T2, which is also why the actual CPU usually doesn't make much difference in most situations. The i9 is a maximum of around 10% faster which has been shown in many benchmarks. That means that in a purely CPU dependent workload with all cores running at 100%, you will see no more than a 10% difference(In most situations it would probably be no more than a few percent).
But when you look at most video editing software they use the GPU much more than the actual CPU, meaning that in those situations there will be absolutely no performance difference between the i7 and the i9. And if you export HEVC then the limitation is the T2 chip which handles that part - And that one is identical between the two machines.

When people are discussing the i7 vs the i9, there is a lot of focus on stuff like video rendering and similar. I'm guessing that 99% of all people in this forum don't ever do video editing or similar tasks. But people tend to forget that when you actually do video editing, 3D modelling and similar heavy tasks - You spend 95% or your time on editing, finding clips, deciding on the story etc. All things where the CPU is only used lightly. So you spend 8 hours editing a video but then people worry about saving 15 seconds on a video export which takes 8 minutes. So unless you do non-stop video export or you export super hi-res formats such as 6K or 8K RED footage, then you won't see any difference and even with those formats you will barely be able to see a difference.

So the additional $500 is for peace of mind and not actual performance. So if you feel better by spending the $500, then do so. But don't do it because you think you're going to get a good boost to performance.

The reason everyone focuses on video editing is that is what the YouTube reviews focus on. Personally, I don't care what the performance difference is between the i7 and i9 is when editing videos because I don't edit videos.

However, I have had no problem maxing out all 10 cores when compiling code or indexing a project in a JetBrains IDE. I am happy to pay another $400 for a 10% performance increase. Why limit myself over $400. If I was looking to cut costs I would have probably ordered a smaller SSD or perhaps less RAM.