October 18th Apple Event - Unleashed - Apple Silicon, MacBook Pro upgrades, HomePod mini, AirPods 3rd Generation

[LAwLz]

2 hours ago, BillTheThrill said:

I'm keeping a close eye on what comes from Apples contribution to Blender and how it works on their own silicon. Though I'm not expecting anything huge. The M1 was a disappointment there as it even gets beat out by an AMD Ryzen 3 1300x.

[Citation needed]

I find that claim very hard to believe since the M1 wipes the floor with the 1300X in terms of raw CPU performance.

[leadeater]

4 minutes ago, LAwLz said:

Just wait for the machines to end up in the hands of people like Andrei at Anandtech, and then you will get what you want.

Apple's press events are not aimed at people who want 20 minutes of graphs comparing it against the competitors in a bunch of different scenarios. Their events are aimed at the average Joe who wants something quick and easy to digest while still getting the overall picture across.

And if it did come from Apple then you have to apply the as required skepticism to it, so it would be quite the wasted effort on Apple's behalf. It's more useful when AMD, or Intel, does it because they are more typically talking about an architecture not specifically the CPU itself then go on to compare specific models against others. This is also due to that their CPUs go in to many different devices of different configurations and allowed power/boost settings so we kind of need that standardized baseline to compare to and fact check their claims so if they didn't give these more detailed presentations we'd have very little to compare to.

 

Apple SoCs only come in Apple devices and always run as Apple designed/configured them to, not something you can say about an AMD or Intel laptop CPU.

[Laborant]

I am very impressed - also thinking of how well Apples claims for M1 held their truth with the M1 "classic" (or entry-level) lineup. In my social circle, Macs always have been viewed as being a sports-car that runs really well on paved road but struggles in less smooth workflows. The road of video-editing seems to be really smooth and the Apple-Silicon lineup rocks that like a Porsche 911 the Nürburgring. On the less smooth roads (e.g. FPGA design compares more to a hiking path) they are simply put useless. (Only x86-Linux/Windows supported) 
As both FPGA-companies are strongly tied to the two big x86-companies, I do not expect any Apple-Silicon port anytime soon.

But it gives me hope, that PC manufacturers try to catch up. Competition just became fierce. 
I wonder how the GPU in the M1 lineup compares to the mobile nVidia solutions in rasterization and ray tracing (if applicable benchmarks are available). I expect Apple still to be crushed in ray tracing.

[maplepants]

1 hour ago, LAwLz said:

Just wait for the machines to end up in the hands of people like Andrei at Anandtech, and then you will get what you want.

Apple's press events are not aimed at people who want 20 minutes of graphs comparing it against the competitors in a bunch of different scenarios. Their events are aimed at the average Joe who wants something quick and easy to digest while still getting the overall picture across.

You are totally right about waiting for the reviews. Watching the announcements, I'm pretty excited to see reviews. The CPUs are so different and interesting that you get a wide variety of reviews. One of the most interesting M1 reviews I saw was this one comparing machine learning performance on the M1 to the then top of the line 16" MBP. 

 

With 64GB of unified memory and a very beefy SOC, it'll be interested to see people pit these laptops against their Mac Pro towers. I think we all know the new MBP will out perform the Xeons it some single-threaded bursty workloads like javascript, but that's not very exciting because Xeons aren't optimized for that type of workload. I'm looking forward to seeing the MBP and Mac Pro compared across traditional Xeon workloads like machine learning, compiling code, compressing/rendering 3D assets, virtualization, etc.

 

Apples vague M1 charts were vaguely correct about what to expect for performance. So if these are too, then we're going to see/read some very interesting reviews in the next few months.

[porina]

5 hours ago, IAmAndre said:

Considering that the M1 is available and can be compared to x86 CPU's then you can make better approximations.

That doesn't really help either. The M1 might be efficient but wasn't interesting in absolute performance, not that any of the performance comparisons made at the time were really useful to me. Making multiple leaps to estimate performance is best avoided where possible.

[Aidoneus]

The most interesting thing about this launch is that it validates a leak from c. 6 months ago regarding the Jade C-Die chips. 
 

Jade C-Chop = M1 Pro

Jade C-Die = M1 Max

Jade 2C-Die = ?

Jade 4C-Die = ?

 

That same leak predicted that the Jade 4C (the presumed high-end desktop chip) would be 40 CPU cores and 128 GPU cores. If we’re talking about 32 cores in the M1 Max in the same breath as a mobile 3070/80, just think about 4x that performance…

 

That’s around desktop 3090 performance on Apple’s first go at a desktop GPU. AND the GPU cores in A15 got a huge boost over the cores in A14 which are the base of the M1. Could be an interesting next few years in the GPU space.

[Kronoton]

10 minutes ago, Laborant said:

The road of video-editing seems to be really smooth and the Apple-Silicon lineup rocks that like a Porsche 911 the Nürburgring. On the less smooth roads

 

So you really thing  the Ring is "smooth roads"?

 

Or should I point out that in the olden days they used (with some success) 911s for Paris-Dakar?

 

I do see what you are trying to tell, your analogy is just totally borked. 

[RejZoR]

9 minutes ago, maplepants said:

You are totally right about waiting for the reviews. Watching the announcements, I'm pretty excited to see reviews. The CPUs are so different and interesting that you get a wide variety of reviews. One of the most interesting M1 reviews I saw was this one comparing machine learning performance on the M1 to the then top of the line 16" MBP. 

 

With 64GB of unified memory and a very beefy SOC, it'll be interested to see people pit these laptops against their Mac Pro towers. I think we all know the new MBP will out perform the Xeons it some single-threaded bursty workloads like javascript, but that's not very exciting because Xeons aren't optimized for that type of workload. I'm looking forward to seeing the MBP and Mac Pro compared across traditional Xeon workloads like machine learning, compiling code, compressing/rendering 3D assets, virtualization, etc.

 

Apples vague M1 charts were vaguely correct about what to expect for performance. So if these are too, then we're going to see/read some very interesting reviews in the next few months.

The M1 Max is quite a chungus with 8 big cores and 2 efficiency cores. Which doubles the performance core count of regular M1. Also, the thing with Macs is, they don't need Xeon's approach of "we have many cores/threads". It being a SoC it has a predictable CPU and GPU computational power. Where with Xeon you can pretty much assume it's just CPU and for GPU, it's anyone's guess if it's even there and it can be a GTX 1050Ti or a cluster of NVIDIA Tesla or Radeon Instinct cards. And because OS is tailored specifically for it and so are apps, they'll do tasks that run better on CPU there and run tasks that run better on GPU on GPU. Like there is no point in encoding MP3's with GPU, there is no point handing highly multithreaded tasks to those 8 HP cores if you have a GPU there with predictable feature set and performance to do it.

[RedRound2]

33 minutes ago, Aidoneus said:

The most interesting thing about this launch is that it validates a leak from c. 6 months ago regarding the Jade C-Die chips. 
 

Jade C-Chop = M1 Pro

Jade C-Die = M1 Max

Jade 2C-Die = ?

Jade 4C-Die = ?

 

That same leak predicted that the Jade 4C (the presumed high-end desktop chip) would be 40 CPU cores and 128 GPU cores. If we’re talking about 32 cores in the M1 Max in the same breath as a mobile 3070/80, just think about 4x that performance…

 

That’s around desktop 3090 performance on Apple’s first go at a desktop GPU. AND the GPU cores in A15 got a huge boost over the cores in A14 which are the base of the M1. Could be an interesting next few years in the GPU space.

I still remember the days when detractors in this forum used to swear by themselves that ARM chips would never compete with x86 in terms of raw performance and innvoating in the GPU space was next to impossible task due to the patent minefield. And this was like merely 2-3 years ago

Now, its quite amazing to see Apple single handedly shatter those misconceptions. Ignoring the fact that Apple Silicon will only be available on macOS and Apple devices, their new chips are admittedly more impressive than AMD's return with Ryzen, which by itself was impressive given how much catching up they had to do (and still do in the GPU space)

 

And as you mentioned all the M1 Jade dies are based on  a year old A14. M2 chips looks to be even more impressive

 

Pretty exciting times ahead in the CPU and GPU space after a decade of 10% improvements yoy

[Dracarris]

Hmm an awful lot of radio silence on the main LTT channel. Usually by early European morning a video was up after an Apple event. I wonder what they are baking this time.

[Dracarris]

4 hours ago, leadeater said:

Just FYI they don't actually cost that much, I have prices but you can't see them muahaha (not allowed)

This arrived yesterday to one of our research groups, as a gift from Nvidia. I was told the regular price would be north of 150.000$

[leadeater]

24 minutes ago, Dracarris said:

This arrived yesterday to one of our research groups, as a gift from Nvidia. I was told the regular price would be north of 150.000$

Nice, one of our departments went with this: https://lambdalabs.com/deep-learning/servers/hyperplane-a100 (The 4U model with 8 A100).

 

Just about to sort out networking and power allocation for it. 

 

Wish ours was a gift hahah.

[HenrySalayne]

3 hours ago, LAwLz said:

Apple's press events are not aimed at people who want 20 minutes of graphs comparing it against the competitors in a bunch of different scenarios. Their events are aimed at the average Joe who wants something quick and easy to digest while still getting the overall picture across.

That's true but the constant "up to", "up to", "up to" triggers me quite hard. It's just fuel for the hype train.

And don't get me wrong - these machines probably are powerful beasts. But just imagine Intel would say their processors are "up to" 4x faster than AMD's because they used an AVX512 benchmark nobody cares about. They would create an inevitable shitstorm of epic proportions.

Cherry-picking the best result feels dishonest. I would much rather prefer something like "this device is 60% faster across a wide variety of applications" than "this device is 2.5x faster in this particular usecase we don't even disclose". It would be much easier for the average Jane and Joe to understand how the new Macbook impacts their everyday work.

 

 

[hishnash]

2 hours ago, Laborant said:

I expect Apple still to be crushed in ray tracing.

This will depend a lot on your tasks, if a large amount for the tree structure is small enough to fit on die in its cache layers (it might well have over 64MB of cache on die) then it could be much faster. And if the scenes (for a professional task) is larger than the VRAM limit of your dedicated GPU 16GB then it again will be much faster.

[Video Beagle]

15 hours ago, RedRound2 said:

And can be charged by Type C as well. So I guess I wouldnt actually be needing to plug out the Magsafe

Could you plug the magsafe cable into the usb-c port of the macbook and have perpetual power?

[leadeater]

10 minutes ago, Video Beagle said:

Could you plug the magsafe cable into the usb-c port of the macbook and have perpetual power?

Go around stealing power from all the Windows laptops lol

[Spindel]

Hmmm laptop CPU at around 50 W rivaling a desktop CPU at 2,5 times the power draw. At least of this is true. 
 

https://browser.geekbench.com/v5/cpu/10478011

[Video Beagle]

1 hour ago, Dracarris said:

Hmm an awful lot of radio silence on the main LTT channel. Usually by early European morning a video was up after an Apple event. I wonder what they are baking this time.

I see a post from Anthony @GabenJr in the new articles list on the side, in the official catagory, so guess you were just a tad early for the post.  (I was shocked the Technews had a postiive headline...I was sure they'd do a headline critcizing the notch)

[leadeater]

3 minutes ago, Video Beagle said:

I see a post from Anthony @GabenJr in the new articles list on the side, in the official catagory, so guess you were just a tad early for the post.  (I was shocked the Technews had a postiive headline...I was sure they'd do a headline critcizing the notch)

Seems both videos, LTT and Mac Address have been up for an hour. Time to start watching

[Video Beagle]

Just now, leadeater said:

Seems both videos, LTT and Mac Address have been up for an hour. Time to start watching

it's 4:45am here.. I'll wait a bit longer

[WolframaticAlpha]

16 hours ago, leadeater said:

We could always play a game of who's looks worse, here is the 16" with the M1 Max with 64GB

 

Waiting for yourself or someone from Japan to crush my number lol

Damn, New Zealand pricing is worse than India's

[WolframaticAlpha]

1 hour ago, Dracarris said:

This arrived yesterday to one of our research groups, as a gift from Nvidia. I was told the regular price would be north of 150.000$

I would kill for this.

[WolframaticAlpha]

I suspect that the M1 Max would give off a ton of heat on a 14 inch form factor. M1 Max is only one sku on their entire product stack. It will still be extremely impressive to have that much power in a notebook.

 

I wish that I had waited for M1 before pulling the trigger on a wintel machine(winamd to be precise) for the ports. With the exception of USB-A, this laptop has everything that I need. Oh well.

[leadeater]

21 minutes ago, WolframaticAlpha said:

Damn, New Zealand pricing is worse than India's

Well they are all the "same price", just different currency and taxes. GST here is 15% so you can take that off before converting back to USD. Still, plenty of currencies out there that would be a bigger number haha

[WolframaticAlpha]

Just now, leadeater said:

Well they are all the "same price", just different currency and taxes. GST here is 15% so you can take that off before converting back to USD. Still, plenty of currencies out there that would be a bigger number haha

GST on electronics is 18% and we have import duties/sanctions on china(iirc).