The M1 Benchmarks Continue - Emulated performance appears to *still* outperform any intel-based Mac

[LAwLz]

45 minutes ago, 05032-Mendicant-Bias said:

Now, this is at least a believable score, and still very impressive considering TDP.

Half performance of a 3600x at a fraction of the power.

Power efficiency comes from the better process and the weaker simpler core architecture.

It's not half the performance of a 3600X.

 

CInebench R23:

 

Single core score:

3600X - 1286

M1 - 1498

Difference: The M1 is 16-17% faster at single core.

 

 

Multi core score:

3600X - 9588

M1 - 7508

Difference: The 3600X is 27-28% faster at multi core.

 

 

So with 50% more cores, 40% higher clock speed and 850% higher power consumption it is slower at single core, and 27-28% faster at multi core.

I'd say that is VERY impressive for the M1.

[leadeater]

27 minutes ago, LAwLz said:

Also, @leadeateris deliberately trying to make AMD look as good as possible by showing the power consumption on the 5950X which according to Anandtech had some rather odd behavior when it came to power consumption. If we instead look at the 5900X, the 5800X or the 5600X we see a much more expected and linear power consumption increase as the core count increases.

It's not odd, the 5950X simply has better binned CCDs and the higher number of cores means lower clocks so the cores use less power. It's actually not odd at all to see the 5900X, and 5800X using more power per core when they are boosting higher on each core under load.

 

27 minutes ago, LAwLz said:

We don't know if increasing the IO will balloon the power usage like leadeater claims it will.

I never said it will balloon, it's a well known fact that I/O is not power efficient and also does not respond well to node shrinks. M1 I/O design is totally different, it doesn't need to drive Infinity Fabric links and it doesn't need to drive higher power DRAM that is much more distant to the memory controller, both these being a significant portion of the IOD power usage.

 

27 minutes ago, LAwLz said:

A single Zen3 core uses more power than the entire M1 chip uses. This is not something AMD can just optimize away with less IO and a node change.

No this is simply a misrepresentation, you should know it. Using peak single core boost power to state a single Zen 3 core uses more than the M1 is not true. A single Zen 3 core CAN use more than the M1 if you allow it. If AMD were to release a Zen 3 SKU with microcode that caps peak clocks to 3GHz that will not be true at all.

 

I'm not sure what you are trying to get at here with this but you should know better.

 

Edit:

P.S. If I wanted to pick the most favorable viewpoint for AMD I would have showed Ryzen Mobile U series graphs, which are lower power per core again.

 

The entire point of that post of mine was to show how low the power is required for Zen 3 cores at lower clocks. I think you missed the point. Also power usage and frequency is non linear, you know that too.

[LAwLz]

10 minutes ago, leadeater said:

It's not odd, the 5950X simply has better binned CCDs and the higher number of cores means lower clocks so the cores use less power. It's actually not odd at all to see the 5900X, and 5800X using more power per core when they are boosting higher on each core under load.

Here is a quote from the article you yourself referenced:

Quote

There are two significant features of this graph.

 

First is the hump, and a slow decrease in total package power consumption after 8-10 core loading. We saw this when we first tested the previous generation 3950X, and is indicative of how the processor has increased current density as it loads up the cores, and as a result there’s a balance between the frequency it can give, delivering the power, and applying the voltage in a consistent way. We’re seeing the difference between the two values also increasing slightly, as more data is transferred over those off-chiplet communications. We see this effect on the 5900X as well, perhaps indicating this is a feature of the dual chiplet design – we’re not seeing it on the 5800X or 5600X.

 

The second feature is an odd dip in power moving from 4 to 5 cores loaded. Looking into the data, the frequency of the active cores drops from 4725 to 4675, which isn’t a big drop, however the voltage decreases from 1.38 V to 1.31 V, which seems to be more sizeable drop than other voltage readouts as we scale the core-to-core loading. There’s also a bigger increase in non-core power, up from 16 W to 21 W, which perhaps decreases the power to the cores, reducing the voltage.

 

This might be an odd quirk of our specific chip, our power test, or it might be motherboard or BIOS specific (or a combination of several factors). We might go back in future on other boards to see if this is consistent.

 

 

 

8 minutes ago, leadeater said:

I never said it will balloon, it's a well known fact that I/O is not power efficient and also does not respond well to node shrinks. M1 I/O design is totally different, it doesn't need to drive Infinity Fabric links and it doesn't need to drive higher power DRAM that is much more distant to the memory controller, both these being a significant portion of the IOD power usage.

Then I think you should have phrased it better because we now have people "quoting you" saying that once Apple start adding more ports to their laptops they will no longer be more efficient than AMD or Intel.

That the M1 is only more efficient that Ryzen because it lacks IO ports.

 

 

11 minutes ago, leadeater said:

No this is simply a misrepresentation, you should know it. Using peak single core boost power to state a single Zen 3 core uses more than the M1 is not true. A single Zen 3 core CAN use more than the M1 if you allow it. If AMD were to release a Zen 3 SKU with microcode that caps peak clocks to 3GHz that will not be true at all.

It's not a misrepresentation because in things like the Cinebench R23 test (which we are using as a reference for the performance each chip puts out) the CPU does boost since it only using a single core.

We are comparing apples to apples here. If we start going "okay but we shouldn't look at the boost clock of the Zen chip because that means each core uses more power" then we shouldn't use the boosted clock score in benchmarks either. You can't pick and choose in this way, where you go "oh we're talking about efficiency? Then let's not look at boosted clocks?" and then turn around and go "oh, we're talking about performance? Yeah then let's factor in boost, but not if we are going to talk about efficiency!".

 

Either we don't include boosted clocks but if we don't do that then we should run Cinebench with boosting disabled as well. Or we do include boosted clocks but then we include that in the power consumption figures as well.

[Video Beagle]

16 hours ago, LAwLz said:

Microsoft are trying VERY hard to make ARM PCs a thing.

.....

Apple will not have any of these issues with their transition. On top of that, it seems like Apple are simply better at writing software than Microsoft at this point. Microsoft's x86 emulation layer in Windows for ARM seems to suck ass compared to Rosetta Stone 2. 

I wonder, and keep in mind as I've said in other threads, we might as well be talking fairy dust with how I understand this level of the computer....in fact, I understand how Disney's fairy dust works much better than I do this level...Could Rosetta Stone 2...in the hands of like Parallels, be the secret sauce for getting windows on ARM?

 

16 hours ago, RedRound2 said:

 

The new techquickie actually gives a pretty decent picture about it (Lightning connector), despite the clickbaity title

 

What? A well done LMG article about an apple topic with an immature title designed to appeal to the least among us?  Say it ain't so!

 

10 hours ago, Bombastinator said:

Noting that this is even possible.  People like to work with what they are used to.  I very much doubt vi and bourneagain shell are popular any more for example.

I use vi when I need to edit inside the command window (which on a mac you don't really need to do)....its what I learned on so many years ago, and newer editors confuse me

[A51UK]

I have look at where the source of the data and Benchmarks people are talking about, I would not be talk about M1 right now as the source is a really bad one and have too much bias for apple.  We should not be take a word from site that is that so pro apple and could be pay by apple. We should wait for the non-bias reviews, I am real looking for to them until them let stop waste time and stop give out highly unreliable benchmark information .  The M1 could be really good but could very poor but we only find out from  non-bias reviews and other Benchmarks. I betting it going be may half the performance or less but a really good power efficiency.

[LAwLz]

1 minute ago, A51UK said:

I have look at where the source of the data and Benchmarks people are talking about, I would not be talk about M1 right now as the source is a really bad one and have too much bias for apple.  We should not be take a word from site that is that so pro apple and could be pay by apple. We should wait for the non-bias reviews, I am real looking for to them until them let stop waste time and stop give out highly unreliable benchmark information .  The M1 could be really good but could very poor but we only find out from  non-bias reviews and other Benchmarks. I betting it going be may half the performance or less but a really good power efficiency.

Half the performance of what?

 

Also, we already have third party independent tests from reliable sites. We have that thanks to the M1 sharing cores with the A14.

We already know the performance of these cores, and it is fantastic. 

[leadeater]

19 minutes ago, LAwLz said:

Here is a quote from the article you yourself referenced:

Yes I know, I have read it, they give explanations as to potentially why however it has no impact at all on what was being discussed. What I was specifically pointing to was the power required on a core at lower clocks, I could have used any of them but I used the 5950X as it had the lowest all core clocks of the SKUs tested which is the most relevant part to what I was talking about.

 

19 minutes ago, LAwLz said:

Then I think you should have phrased it better because we now have people "quoting you" saying that once Apple start adding more ports to their laptops they will no longer be more efficient than AMD or Intel.

Where did someone actually specifically say that? I've seen discussion on power etc but not that statement far as I've seen.

 

19 minutes ago, LAwLz said:

It's not a misrepresentation because in things like the Cinebench R23 test (which we are using as a reference for the performance each chip puts out) the CPU does boost since it only using a single core.

No it very much is, why are the 16 cores not using 20W each then? You now better, please don't.

 

Edit:

My post had nothing to do with what ever you are discussing but you saying a single Zen 3 core use more power than the entire M1 SoC simply is disingenuous, I'm actually rather disappointed you'd throw that out knowing what you know.

[A51UK]

Half the performance of AMD or Intel high end CPU not something like atom, What reliable sites can you give us some names? Also site like MacRumors are not reliable site and what other benchmarks other them Geekbench, 

[RedRound2]

9 minutes ago, A51UK said:

I have look at where the source of the data and Benchmarks people are talking about, I would not be talk about M1 right now as the source is a really bad one and have too much bias for apple.  We should not be take a word from site that is that so pro apple and could be pay by apple. We should wait for the non-bias reviews, I am real looking for to them until them let stop waste time and stop give out highly unreliable benchmark information .  The M1 could be really good but could very poor but we only find out from  non-bias reviews and other Benchmarks. I betting it going be may half the performance or less but a really good power efficiency.

How does the website known for reporting Apple stuff making up positive benchmarks? 

 

These are real people who tested out their machines in the real world. Doesn't matter if the report was initially posted by an Apple website or Android Authority.

 

Right now, most people in PCMR are trying really hard and performing some insane level of false justification and invalidation to spin this in a way to delegitimize the objectively impressive performance out of a 10-15W chips

[hishnash]

9 minutes ago, A51UK said:

Half the performance of AMD or Intel high end CPU not something like atom, What reliable sites can you give us some names? Also site like MacRumors are not reliable site and what other benchmarks other them Geekbench, 

Wait 12 hours for the review embargos to lift. At the moment until that happens all we realy have is benchmarks that trick people into accidentally uploading results. No real apps do that so we need to wait for those numbers.

[A51UK]

Just now, hishnash said:

Wait 12 hours for the review embargos to lift. At the moment until that happens all we realy have is benchmarks that trick people into accidentally uploading results. No real apps do that so we need to wait for those numbers.

I total agree that my point  We need to wait for the review embargos to lift and see what the really number. 

 

 

 

[LAwLz]

1 minute ago, leadeater said:

Yes I know, I have read it, they give explanations as to potentially why however it has no impact at all on what was being discussed. What I was specifically pointing to was the power required on a core at lower clocks, I could have used any of them but I used the 5950X as it had the lowest all core clocks of the SKUs tested which is the most relevant part to what I was talking about.

You never specified "at lower clocks". In fact, you literally said "higher clocks" in your post.

 

21 hours ago, leadeater said:

In relation to Zen 2 and Zen 3 the CCD's are actually very power efficient themselves, each core does not require that much power to achieve high performance and higher clocks. The clock scaling and power is still poor on the upper extremes but everything is, that's not an Intel/AMD/Apple/TSMC/x86/ARM thing that's just electrical and atomic properties at play.

  

That's a vague term so you got some protection there, but the fact remains that people are now using your post as a source for completely and utterly ridiculous claims like the M1 CPU not being more efficient than let's say Intel or AMD. That it's just IO die and 5nm that makes the difference, which is stupid.

 

I'm sorry to have to bring you into this conversation but it seems like people are misunderstanding your post and using it as an argument. You might have picked the 5950X as a point of reference because it had the lowest clocks, but people are now using your numbers to try and say the M1 isn't good in Cinebench R23, where half the test is done using a single boosted core.

People like CarlBar are literrally saying AMD's CPU cores don't use more than 2 watts each when we were talking about Cinebench scores.
Maybe you should tell @CarlBar "hey, stop referencing my numbers because you are using them incorrectly" so that we can avoid your post leading to a bunch of incorrect conclusions?

 

  

10 minutes ago, leadeater said:

Where did someone actually specifically say that? I've seen discussion on power etc but not that statement far as I've seen.

Someone posted Cinebench scores.

CarlBar replied that he was not impressed with the M1 scores. When asked why he said that Zen3 on the same process node and with a smaller/more efficient IO die would get the same power efficiency, using you as a source. Here is the full post.

 

14 minutes ago, leadeater said:

No it very much is, why are the 16 cores not using 20W each then? You now better, please don't.

Because they clock the cores way down when all of them are active. But that's not what we are talking about. I get that the conversation can be a bit confusing since you were dragged into it halfway through but here is a short summary:

Person1: Posts CInebench scores

CarlBar: These are not super impressive. 

Me: The 3600 is only 13% faster despite having 50% more cores, 40% higher clock speed and 850% higher power consumption and desktop cooling, and you're not impressed? Holy crap...

CarlBar: leadeater says the IO die uses a lot of power and each Zen2 core only uses 2 watts, so if AMD went to 5nm with Zen3, with a smaller IO die they would be just as efficient.

 

People are using your post with the 5950X's per core numbers and thinking they apply to chips like the 5600X too. 

 

 

 

19 minutes ago, A51UK said:

Half the performance of AMD or Intel high end CPU not something like atom, What reliable sites can you give us some names? Also site like MacRumors are not reliable site and what other benchmarks other them Geekbench, 

Anandtech's A14 analysis.

[Blademaster91]

13 minutes ago, RedRound2 said:

How does the website known for reporting Apple stuff making up positive benchmarks? 

 

These are real people who tested out their machines in the real world. Doesn't matter if the report was initially posted by an Apple website or Android Authority.

 

Right now, most people in PCMR are trying really hard and performing some insane level of false justification and invalidation to spin this in a way to delegitimize the objectively impressive performance out of a 10-15W chips

PCMR? Thats an interesting way to accuse people of being fanboys.

It doesn't matter if people are testing real hardware, people are getting overhyped over synthetic benchmarks and comparing completely different CPU's and operating systems. Wait for reviews.

[LAwLz]

Just now, Blademaster91 said:

PCMR? Thats an interesting way to accuse people of being fanboys.

It doesn't matter if people are testing real hardware, people are getting overhyped over synthetic benchmarks and comparing completely different CPU's and operating systems. Wait for reviews.

Still sticking to that "it's just synthetic benchmarks!" are we?

What's the next straw you're going to grasp at once "real world" benchmarks (because apparently things like compiling code isn't real world to you) are released? 

Are you really going to start saying you can't compare benchmarks running on different OSes? That's a neat way of invalidating any comparison to Zen that case.

[RedRound2]

5 minutes ago, Blademaster91 said:

PCMR? Thats an interesting way to accuse people of being fanboys.

It doesn't matter if people are testing real hardware, people are getting overhyped over synthetic benchmarks and comparing completely different CPU's and operating systems. Wait for reviews.

Where did this notion arise that synthetic benchmarks mean nothing? And a new twist that is suddenly doesn't mean anything when you're comparing different CPUs and operating system? So does that mean all Intel and Ryzen comparisons are invalid and it absolutely means nothing?

 

These "real reviews" you so call say, if you didn't know (including LTT) uses these "syntehtic benchmarks" to compare performance.

[Belgarathian]

2 hours ago, Spindel said:

FCPX performance on M1 mac compared to iMac Pro:

 

Exporting H.264 Sony 10 bit 422 footage with just one rec709 lut takes:

 

- 11 minutes and 30 seconds on iMac Pro with Vega 56 and 128gigs of ram.
- 10 minutes and 20 seconds on M1 MacBook Pro with 8gigs of ram.

 

30 seconds long H.265 canon 10-bit 422 footage at 100fps takes 

- 80 seconds on iMac Pro 
- 45 seconds on new MacBook Pro

 

 

https://forums.macrumors.com/threads/fcpx-performance-on-m1-chip.2268919/

Well okay then.

[A51UK]

30 minutes ago, RedRound2 said:

Right now, most people in PCMR are trying really hard and performing some insane level of false justification and invalidation to spin this in a way to delegitimize the objectively impressive performance out of a 10-15W chips

Saying people are PCMR (fanboy) is wrong most of us just like computers, I may not like apple myself more to do been overprice, right to repair, T1 chip, customer right problem, no care for developer, keep remove ports (e.g. audio jack, USD), monopoly but I do own three IPods.  Do you know that AMD was working on ARM CPU for years and there Windows for ARM.  I do own a ARM PC that run Linux and can run Windows (something I planning to try).  I will do hope that M1 chips does well and if hit half the performance  of top x86-64 CPU them it have do really well and would be a good thing for ARM CPU.  

[leadeater]

44 minutes ago, LAwLz said:

You never specified "at lower clocks". In fact, you literally said "higher clocks" in your post.

Is 3.775Ghz not higher than 3GHz of the M1? Does the table not show exactly what I said? Reasonably low per core power even as the frequency increases but the increase is non linear and at the extremes gets very high. There is a difference of ~20W vs ~6.2W looking at a single core, I think that shows there is a pretty extreme difference in power as clocks go up towards the extremes i.e. the highest possible point.

 

The M1 at 3.8GHz wouldn't be a 10W part, we both know that. No idea what it would be as I have no idea how the frequency scales on Apple's design. Doesn't really matter however. Neither AMD's Zen 2/3 cores or Intel's have to be a power hog and terrible power efficiency as it is on desktop designed parts. 

 

44 minutes ago, LAwLz said:

People like CarlBar are literrally saying AMD's CPU cores don't use more than 2 watts each when we were talking about Cinebench scores.

Well that isn't actually wrong though, the Ryzen 4700U (Zen 2) sustained power is 15W which gets 3GHz all core over the 8 cores so that is already 1.875W per core without taking off any I/O power of the CPU. The 4700U also scores 2500 in CB R20 Multi, the 4800U (also 8 cores 15W) scores 3200, there is more 4700U data which is why I'm more using that.

 

So Zen 2 & 3 on 7nm can already use as little as 2 watts per core, add on say 15% more performance for the Zen 2 vs Zen 3 if you want to look at performance. Won't matter the M1 is much faster here at these package power. Going to need some R23 4700U or 4800U results to talk more about this.

 

So Carlbar's hypothetical post about Zen on 5nm etc is actually not unrealistic, it's already true now on 7nm, only about per core power though.

[CarlBar]

@LAwLz leadeater has done a pretty good job of going over everything you quoted. But also, your extra charts aren't as decisive as you'd like. They don't go as low down in frequency. If you check the 5950X's power per core at the lowest frequencies on the  5900X and 5600X charts you'll notice that the 5900X is about the same, (taking a quick eyeball average on the 4475 and 4425 as there's no exact 4450 numbers for the 5950x), and the 5600X actually has lower values, it's only the 5800X thats performing worse.

 

@leadeater Found this list of R23 numbers: https://www.cpu-monkey.com/en/cpu_benchmark-cinebench_r23_multi_core-16

 

Looks acurratte to the results examples i've seen listed in R23 itself. It puts the 4800U at 10k points.

[leadeater]

14 minutes ago, CarlBar said:

@leadeater Found this list of R23 numbers: https://www.cpu-monkey.com/en/cpu_benchmark-cinebench_r23_multi_core-16

 

Looks acurratte to the results examples i've seen listed in R23 itself. It puts the 4800U at 10k points.

Careful with that though, Ryzen Mobile U is 15W, with TDP up 25W mode and TDP down 10W mode. I suspect the 10K points is on a laptop set to 25W.

[exhaustedbrit]

20 minutes ago, leadeater said:

Careful with that though, Ryzen Mobile U is 15W, with TDP up 25W mode and TDP down 10W mode. I suspect the 10K points is on a laptop set to 25W.

This is just me hypothesising, but I actually have a suspicion that the bulk of the thermal budget on the M1 chip is not reserved for the CPU. I've run stuff pretty intensely on my M1 MacBook Air and it's barely getting warm, I haven't launched any GPU intensive tasks yet, I'm installing Shadow of the Tomb Raider and some other games. But given there's a supposed 2.6 TFLOP measure of GPU silicon in there I'd expect that's where this magical 10W envelope actually gets hit.

I had one of the Iris Plus MacBook Airs when they were released earlier this year (and swiftly returned it) and remember how quickly the temperatures shot up and barely sustained 3GHz without any sort of GPU load, when these equivalent mobile chips from Intel or Ryzen are performing in tasks which also use the GPU they have to fight for thermal budget. What that also meant that actually playing games was a toss up between whether I wanted that dual core 3GHz or whether I wanted some semblance of GPU performance in that MacBook Air and the result was an inconsistent game with framerates all over the place. My guess is that we aren't even hitting that 10W power budget with a multicore CB23 run because it only stresses the CPU, Apple releases products not hardware (for now anyway) and that's why these comparisons are difficult.

I'm not suggesting it's meaningless to compare the CB23 results with Tiger Lake and Ryzen mobile chips - but I do think the way in which power budget and thermals are approached with this Apple chip is not directly comparable, and a focus on balancing the overall constituent functions (CPU, GPU, ISP etc) for consistent performance has been the goal here.

Best to wait for the reviews - but for now here are my numbers.

M1 MacBook Air CB23 - Multi 7179*, Single 1476

 

Edit: *Re-ran after iCloud syncing and background stuff subsided and scored 7573 in multi.

Edited November 17, 2020 by KuroNanashi
Reran multicore test after settling.

[RedRound2]

52 minutes ago, A51UK said:

Saying people are PCMR (fanboy) is wrong most of us just like computers,

No, a lot of people just religiously hate Apple. It's a cool thing to do. I never said the same people didn't like computers. But the intersection set of enthusiasts and apple haters isnt insignificant to point out. We have seen it time adn time again online and in this forums (and how LTT likes using Apple clickbaity titles)

Quote

I may not like apple myself more to do been overprice, right to repair, T1 chip, customer right problem, no care for developer, keep remove ports (e.g. audio jack, USD), monopoly but I do own three IPods. 

Overpriced? The iPhones, iPads, entry Macs, AW, homePod have very well been documented on how it's actually reasonably priced, especailly for the unparalleled support, ecosystem and in house OS and software they provide. Acessories and upgrades are what I would categorize as overpriced and that definitely exists to position Apple as premium brand (even though iPhones are one fo the most common devices found)

 

Right to repair - I agree

 

Customer right problem? Same point as above?

 

You seem to have a wrong understanding of T1 chip, if you actually think it's a problem and we've been through this in the forum

 

No care for developers? Where does this stem from? Need more sources for this. The 30% thing is an industry standard, but of course only Apple exists in negative light

 

Keep removing ports is a tradeoff they take to move forward with industry. It's becasue of Apple we have cheap and decent bluetooth headsets today, and that USB C is more common in newer devices today

Quote

Do you know that AMD was working on ARM CPU for years and there Windows for ARM.

The point being? Is there a law that states that if AMD can't figure it out, Apple also can't

Windows on ARM is a mess for multitude of reasons. None of the issues which Apple have

Quote

  I do own a ARM PC that run Linux and can run Windows (something I planning to try).  I will do hope that M1 chips does well and if hit half the performance  of top x86-64 CPU them it have do really well and would be a good thing for ARM CPU.  

Good for you? Again point?

[LAwLz]

38 minutes ago, leadeater said:

Is 3.775Ghz not higher than 3GHz of the M1? Does the table not show exactly what I said? Reasonably low per core power even as the frequency increases but the increase is non linear and at the extremes gets very high. There is a difference of ~20W vs ~6.2W looking at a single core, I think that shows there is a pretty extreme difference in power as clocks go up towards the extremes i.e. the highest possible point.

The important piece of info you are missing is that the 5950X NEEDS to be clocked that high in order to be able to perform as well as the M1 does.

The 5950X NEEDS to be clocked at 4.9GHz in order to be ~9% faster than the M1 in single core cores. That is my point and that is very impressive for the M1.

 

Once you start lowering the clocks for Zen3 then sure it becomes more efficient, but it also lowers performance. With the M1 you get both performance and power efficiency. That's what's impressive.

 

 

41 minutes ago, leadeater said:

Neither AMD's Zen 2/3 cores or Intel's have to be a power hog and terrible power efficiency as it is on desktop designed parts. 

True, but they need to be tuned to be power hogs if they want to keep up with the M1.

 

 

42 minutes ago, leadeater said:

So Zen 2 & 3 on 7nm can already use as little as 2 watts per core, add on say 15% more performance for the Zen 2 vs Zen 3 if you want to look at performance. Won't matter the M1 is much faster here at these package power. Going to need some R23 4700U or 4800U results to talk more about this.

We already have that.

At 4.1GHz the 4700U gets a single core score of 1184 in C23.

The M1 gets 1493 at 3GHz.

 

For single core stuff, the M1 is 26% faster, and I suspect that the 4700U is not at its 10 watt tuning for that test either.

 

The multithreaded scores look like this:

4700U - 6874

M1 - 7566

The M1 is 10% faster.

 

In a best case scenario for the 4700U, that 6874 score is at the 10 watt limit. In a worst case scenario, it's at the 25 watt limit. We don't know that yet but even in the base case scenario the M1 is between 10% to 26% faster while using the same or maybe less than half the power.

 

 

It is mind blowing that people aren't impressed by this.

[RedRound2]

1 minute ago, KuroNanashi said:

This is just me hypothesising, but I actually have a suspicion that the bulk of the thermal budget on the M1 chip is not reserved for the CPU. I've run stuff pretty intensely on my M1 MacBook Air and it's barely getting warm, I haven't launched any GPU intensive tasks yet, I'm installing Shadow of the Tomb Raider and some other games. But given there's a supposed 2.6 TFLOP measure of GPU silicon in there I'd expect that's where this magical 10W envelope actually gets hit.

M1 MacBook Air CB23 - Multi 7179, Single 1476

You already got yours? Damn, that's nice

 

Mind if you posts some tests and results on the forum itself 

[exhaustedbrit]

2 minutes ago, RedRound2 said:

You already got yours? Damn, that's nice

 

Mind if you posts some tests and results on the forum itself 

Absolutely, I could start a thread. Any tests in particular you'd like? I re-ran CB23 again after things have settled down, I think I had iCloud stuff syncing and scored 7573 points this time on multi.