Review shows i7 2020 iMac outperforming i9

[jobinhosyntax]

This review from mac4ever shows the i7 outperforming the i9 on Logic benchmarks. They show the core average on the i9 is around 1.4 while the i7 clocks around 4, suggesting sever throttling, although I find these results hard to believe. They also show the i9 scoring only 8000 on Geekbench 5 while most i9s score 10’000. Either there's a huge disparity between i9s or their unit is faulty or they installed the RAM incorrectly. What do you think?

Test de l'iMac 27" 2020 : le dernier des Mohicans ! (+ vidéo)

Pressentis pour la WWDC, Apple aura finalement attendu le milieu de l'été pour renouveler ses iMac !

www.mac4ever.com

 

[Whiteman007]

i just ran geekbench 5 without closing any apps and i got 9800. so they have most likely Ram issue either not duel channel or its downcloaked when using 3rd party with the apple Ram

Could also just be a lemon too

 

[swamyg1]

i just ran geekbench 5 without closing any apps and i got 9800. so they have most likely Ram issue either not duel channel or its downcloaked when using 3rd party with the apple Ram

Could also just be a lemon too

I really hope this is the case. I’ll be mighty upset if I regret purchasing my i9 over the i7. Anybody else care to comment?

 

[chill991]

I’ll have mine in a week, just shipping now. and will be able to test then.

 

[KrazyKanuck]

I really hope this is the case. I’ll be mighty upset if I regret purchasing my i9 over the i7. Anybody else care to comment?

I agree with the others, there's something wonky with their configuration to get those Geekbench scores. 9700-10050 seems to be where most i9 are clustering as I briefly scan through the results. I mean there is such a thing as winning or losing the silicon lottery, but it's probably the RAM configuration or a background application, or something else.

 

[jobinhosyntax]

I agree with the others, there's something wonky with their configuration to get those Geekbench scores. 9700-10050 seems to be where most i9 are clustering as I briefly scan through the results. I mean there is such a thing as winning or losing the silicon lottery, but it's probably the RAM configuration or a background application, or something else.

A 20% drop in performance looks like a fault somewhere in the equation

 

[KrazyKanuck]

A 20% drop in performance looks like a fault somewhere in the equation

Yeah, I suspect improper RAM configuration.

 

[Whiteman007]

I really hope this is the case. I’ll be mighty upset if I regret purchasing my i9 over the i7. Anybody else care to comment?

oh **** forgot to say mine is an i9 but the i7 should be at 8900 while the i9 is around 9900.

 

[jtylee]

I REGRET purchasing the i9 in my 2018 Macbook Pro. The i5 in my new 2020 iMac is a night and day difference. (It should NOT be a night / day difference).

Apple is throttling these i9s still and not telling us.

 

[CWallace]

I REGRET purchasing the i9 in my 2018 Macbook Pro. The i5 in my new 2020 iMac is a night and day difference. (It should NOT be a night / day difference).

Apple is throttling these i9s still and not telling us.

Intel CPUs throttle themselves when they reach a certain temperature to prevent damage. The iMac has more cooling capacity than a MacBook Pro so the iMac CPU can maintain higher turbo clock speeds.

 

[Cookie18]

For the score to be that off there’s something up but I wouldn’t expect a colossal difference between them either.

 

[quagmire]

Intel CPUs throttle themselves when they reach a certain temperature to prevent damage. The iMac has more cooling capacity than a MacBook Pro so the iMac CPU can maintain higher turbo clock speeds.

Issue is Apple will allow the CPU to thermal throttle in order to keep the machine quiet. Thermal throttling should not be the primary method of preventing a CPU from overheating. Increasing the fan speed should be or a better cooling system. Thermal throttle should be last resort for the system before the forced shutdown.

 

[CWallace]

Issue is Apple will allow the CPU to thermal throttle in order to keep the machine quiet. Thermal throttling should not be the primary method of preventing a CPU from overheating. Increasing the fan speed should be or a better cooling system. Thermal throttle should be last resort for the system before the forced shutdown.

Considering how many people complain about the fan noise on Macs when the CPU is under load...

 

[SWolfe]

Here is the 2020 i7 and i9 compared to 2017 iMac Pro from Bare feats

2020 iMac 5K 8-core versus 2017 iMac Pro 8-core

real world speed test results for performance minded Macintosh users

barefeats.com

 

[swamyg1]

Here is the 2020 i7 and i9 compared to 2017 iMac Pro from Bare feats

2020 iMac 5K 8-core versus 2017 iMac Pro 8-core

real world speed test results for performance minded Macintosh users

barefeats.com

Do you know why they are benching an i9 with 128gb of ram vs. the i7’s with 32gb??

That makes no sense to me.

 

[CWallace]

Do you know why they are benching an i9 with 128gb of ram vs. the i7’s with 32gb??

The original comparison was between the two 8-core models, both with 32GB of RAM.

They subsequently added the results for the i9 once that configuration arrived. I would not be surprised if that model was purchased to (eventually) compare with the 2019 Mac Pro or a higher-spec iMac Pro and as such was configured with so much RAM. And since it would be 4x32GB modules, to knock it down to 32GB would mean only one slot would be filled and that would hurt the scores since dual-channel would not be present.

 

[KrazyKanuck]

Do you know why they are benching an i9 with 128gb of ram vs. the i7’s with 32gb??

That makes no sense to me.

I may be mistaken, but I think Geekbench utilizes a set amount of RAM, and that 32GB vs 128GB of RAM shouldn't result in a noticeable difference between scores.

 

[gilby101]

Issue is Apple will allow the CPU to thermal throttle in order to keep the machine quiet. Thermal throttling should not be the primary method of preventing a CPU from overheating. Increasing the fan speed should be or a better cooling system. Thermal throttle should be last resort for the system before the forced shutdown.

Based on past experience, I am sure that Apple will tweak the Turbo Boost (what you call throttling) in future macOS releases. My experience with a 2019 i9 iMac (nominally 3.6GHz) is that, for the first ~3 months, the Turbo Boost was turned down to 3.9GHz long before the fan was noticeable. But gradually, over the past year, my iMac has been much more inclined to use the fan, and so keeping the Turbo Boost at about 4.7 GHz as long as possible.

My 2019 i9 has never been thermally throttled (I don't live in the tropics), though it does go down to about 1.9GHz when lightly loaded.

 

[mlykke]

I may be mistaken, but I think Geekbench utilizes a set amount of RAM, and that 32GB vs 128GB of RAM shouldn't result in a noticeable difference between scores.

You are correct. Rob did test the i7 with both 32GB and 128GB and the results were more or less identical. The 32GB gave 8924 and the 128GB gave 8865 - So both results within the regular fluctuation between multiple tests.