Cinebench R23 with support for Apple M1 SoC is now out. Geekbench scores up too.

[leman]

1. Ah ok. So are people predicting the M1 chip just released will be a faster than the A12Z, and by how much?

By a very healthy margin. Wait for official reviews.

2. Is the M1 likely to exceed performance of my 6-core i7-8700B (in particular for Logic)

For logic? Most certainly, since it will be able to take advantage of M1’s unique features

3. Do you think they will released a higher end SoC for the MacMini also?

Most likely, since they still left the higher end Mini in the shop. Next year maybe?

 

[ADGrant]

Yes but I was thinking the 2020 i3 might be similar performance to my 2018 i7 6-core. In which it would be 3x faster than my current machine (i7). Maybe the wrong assumption..

Is the 2020 i3 similar performace to the 2018 i3? I can only find benchmarks for the 2018 i3; https://www.macworld.com/article/3318501/799-mac-mini-review.html

If so, these figures would suggest the 2020 i3 provides approx. half the performance of my 2018 i7. In the above article it says the 2018 i3 scored 588 (i presume this is single core..?) vs my 1160 pts 2018 i7 (7084 pts multi)

Yes, that was the wrong assumption. The 2020 Mac mini has exactly the same CPUs as the 2018 did.

 

[thingstoponder]

Yes, that was the wrong assumption. The 2020 Mac mini has exactly the same CPUs as the 2018 did.

The 2020 Mini wasn’t a real update, they just double the storage configurations when they released new MacBooks with the same doubling of storage. I don’t think Apple officially refers to it as the 2020 Mac Mini, although it did get the “new” tag on the website when it was released.

 

[imdog]

 

[eulslix]

Thats a nonsensical comparison... the Tesla gpu has been built for entirely different purposes, which the M1 to my knowledge isn’t even competing with, and isn’t capable of in the first place (e.g. ECC). Consequentially, that benchmark says nothing about the relationship between those two chips... ?

 

[Sanpete]

Thats a nonsensical comparison... the Tesla gpu has been built for entirely different purposes, which the M1 to my knowledge isn’t even competing with, and isn’t capable of in the first place (e.g. ECC). Consequentially, that benchmark says nothing about the relationship between those two chips... ?

Saved me $3k. I was on the fence.

 

[imdog]

Thats a nonsensical comparison... the Tesla gpu has been built for entirely different purposes, which the M1 to my knowledge isn’t even competing with, and isn’t capable of in the first place (e.g. ECC). Consequentially, that benchmark says nothing about the relationship between those two chips... ?

What are you talking about? It crushes it in every single available test lol, it beats most GPU's in every benchmark. All I'm saying is that its benchmarks are good, an objective fact.

 

[ArPe]

Even the GPU seem to be competitive. At 10-15W, M1 performs the same as AMD/NVIDIA at much higher TDP. There would be no problem to add 32 core GPU for nearly 4X performance corresponding to a 5700XT even if the chips draw perhaps around 50W.

We already knew the M1 CPU is a winner, but the GPU may be as competitive. Also the neural engines will offload some compute from the GPU and there the gains are even larger. Seems good to me also on the GPU/compute department.

I wish benchmarks reporter performance per watt.

Thats a nonsensical comparison... the Tesla gpu has been built for entirely different purposes, which the M1 to my knowledge isn’t even competing with, and isn’t capable of in the first place (e.g. ECC). Consequentially, that benchmark says nothing about the relationship between those two chips... ?

yep, those compute cards don’t do gaming. They don’t do many mainstream things.

In a comparison against a 5700XT that gets 200fps in that benchmark the M1 GPU is about 3 times slower.

It’s pointless to benchmark a integrated graphics chip against a desktop GPU anyway. That GFXBench app itself is ancient.

We aren’t going to see Call of Duty level raytraced 4K gaming any time soon. Use the M1 for being productive on a laptop with reasonable expectations. If you think you will be able to load “thousands of layers” “billions of polygons etc then you set yourself up for disappointment.

 

[imdog]

Saved me $3k. I was on the fence.

They are wrong

 

[imdog]

yep, those compute cards don’t do gaming. They don’t do many mainstream things.

In a comparison against a 5700XT that gets 200fps in that benchmark the M1 GPU is about 3 times slower.

It’s pointless to benchmark a integrated graphics chip against a desktop GPU anyway. That GFXBench app itself is ancient.

We aren’t going to see Call of Duty level raytraced 4K gaming any time soon. Use the M1 for being productive on a laptop with reasonable expectations. If you think you will be able to load “thousands of layers” “billions of polygons etc then you set yourself up for disappointment.

The M1 outperforms many higher priced GPU's in benchmarks that have nothing to do with gaming

 

[ArPe]

The M1 outperforms many higher priced GPU's in benchmarks that have nothing to do with gaming

Until you try to load a triple A title from Steam or a Blender scene, run out of VRAM and reality hits home.

 

[joshallegro]

No results for M1 were posted yet, so I no, I don’t think you got this right The screenshot in the first page is not M1

I just found benchmark result for the M1. How has this occurred per-release?
Also, is a Geekbench figures comparable to cinebench figures or is it a completely different scoring system?;

Mac mini (Late 2020) Benchmarks - Geekbench

 

[imdog]

Until you try to load a triple A title from Steam or a Blender scene, run out of VRAM and reality hits home.

Yea I am not arguing it is an excellent gaming computer but graphics cards are important for a lot more than gaming. I am a video editor and for my use case I am very optimistic about these chips

 

[Ael_MR]

Until you try to load a triple A title from Steam or a Blender scene, run out of VRAM and reality hits home.

I didn't know a MacBook was a gamer machine or a blender render farm.

 

[the8thark]

I wish benchmarks reporter performance per watt.

Just devide the benched number by the TDP of the chip. Simple math.

 

[thunng8]

Just devide the benched number by the TDP of the chip. Simple math.

It is not that simple. Intel and AMD processor regularly turbo boost to achieve that level of performance. For example the 28w tiger lake can draw up to 50w when benchmarking. Also the AMD Renoir 4800u which is a 15w tdp chip can draw up to 55w for a short amount of time and can sustain 40w for up to 2min - enough time to complete some benchmarks.

We don’t know what the m1 does at the moment, but given the MacBook Air is not actively cooled, I suspect power draw will be much much less than intel and AMD systems which cannot run without a fan.

note that even the 2 core 2020 intel MacBook Air which uses a 9w tdp intel chip cannot run without a fan either.

 

[Boil]

CineBench R23 does not use AVX-512.

 

[leman]

Until you try to load a triple A title from Steam or a Blender scene, run out of VRAM and reality hits home.

Is this a kind of work you could have done on the replaced Intel two-port MBP? Not? Than what is the complain? Sorry, I just don’t get it. Need more RAM? Get an Intel machine that supports more RAM or wait until summer when next batch of Apple Silicon Macs is out. M1 is not made fir your use case.

 

[leman]

CineBench R23 does not use AVX-512.

Thanks for bringing that to our attention!

 

[mr_roboto]

Until you try to load a triple A title from Steam or a Blender scene, run out of VRAM and reality hits home.

I bet you thought that was clever. Here is some reality for you: the M1 doesn't have VRAM to run out of.

 

[ArPe]

I bet you thought that was clever. Here is some reality for you: the M1 doesn't have VRAM to run out of.

Bet you thought your answer was clever. Intel iGPU uses shared system memory too and AGP graphics cards grabbed system memory also. Eventually you hit a bottleneck (apps need the memory), so you have to keep your expectations realistic when relying on graphics chips that use system memory.

 

[eulslix]

What are you talking about? It crushes it in every single available test lol, it beats most GPU's in every benchmark. All I'm saying is that its benchmarks are good, an objective fact.

You can also compare a toaster against a M1 in those tests, I’m sure it’s gonna crush it there as well. Makes about as much sense...

 

[acidfast7_redux]

On a busy day, I read around 1000 pages in Word or PDF or write around 20-50 pages in Word.

Can someone provide a benchmark for me?

With the MBA be enough for my demanding tasks?

 

[leman]

You can also compare a toaster against a M1 in those tests, I’m sure it’s gonna crush it there as well. Makes about as much sense...

In GFXBench, the M1 is comparable to a 1650-Ti Max-Q, which is a 35Watt Nvidia Turing GPU. Frankly, I am doubting that result, since it is a bit too much (it would mean that Apple now has a 3x perf-per-watt lead over Nvidia which would be crazy), but it can be used as an indicator. Even if the "real" M1 performance is closer to MX450, it would be a great win for Apple and Mac users.

 

[leman]

On a busy day, I read around 1000 pages in Word or PDF or write around 20-50 pages in Word.

Can someone provide a benchmark for me?

With the MBA be enough for my demanding tasks?

You will want at least 32GB of RAM for that...