mdriftmeyer

tipoo said:

Just to add something, GPU core counts are all counted differently and meaningless across architectures. An Apple GPU core is 128 ALUs, say an Intel one is 8. 

Seeing what they did with the 8C M1, the prospect of a 128 core Apple GPU is amazingly tantalizing, that's 16,384 unified shaders.

Tantalizing and absurd because you will be starving most of your shaders sitting their doing nothing and that's not how GPUs work. Apple is investing in their specialty processing to be the heavy lifter--the Neural Engine, ML Engine, Audio/Video Decode/Encode process most of the system's requests already. 

blastdoor said:

ph382 said:

rob53 said:

Why stop at 32 cores. 

I don't see even 32 happening for anything but the Mac Pro. I saw forum comments recently that games don't and can't use more than six cores. How much RAM (and heat) would you need to feed 32 cores?

https://www.anandtech.com/show/15044/the-amd-ryzen-threadripper-3960x-and-3970x-review-24-and-32-cores-on-7nm

The 32 core Threadripper 3970x has a TDP of 280 watts on a 7nm process. It has four DDR4 3200 RAM channels.

Based on comparisons of the M1 to mobile Ryzen, I would expect an ASi 32 core SOC to have a TDP much lower than 280 watts. 

I bet a 32 core ASi SOC on a 5nm process could fit within the thermal envelope of an iMac Pro. 

It has 8 not 4 DDR 4 3200 ECC RAM Channels and those are limited by the OEMs. Zen supports 2TB of DDR4 RAM since Zen 2. And Threadripper is limited to 32 Cores presently because they haven't moved to the Zen 4 5nm process with RDNA 3.0/CDNA 2.0 based solutions that are on a unified memory plane.

Those 32 cores would be a 16/16 Big/Little and then combine their GPU and other co-processors and you have a much larger SoC or very small cores.

TR 3 arrives this January along with EPYC 3 Milan with 64/128 Cores. The next releases as Lisa Su has stated and their software ROCm has shown will be integrating Xilinx co-processors into the Zen 4/RDNA 3.0/CDNA 2.0 based solutions and beyond. 

Both AMD and Xilinx have Architecture licenses to ARM and have been designing and producing ARM processors for years. Xilinx itself has an arsenal of solutions in ARM.

32 Cores would only be in the Mac Pro. 8/8 cores in the iMac and 12/12 in the iMac Pro is pushing it.

In 2022 Jim Keller's CPU designs from Intel hit the market. The upcoming Zen architecture designs will be announced in January 2021 at the CES Virtual conference. AMD has already announced by 2025 its conservative Product sales of Hardware will be over $22 Billion. That's up from this year's just over $8 Billion.

Apple has zero interest in supporting anything beyond their Matrix of hardware options and people believing they want to be all solutions to all people don't understand and never have understood the mission statement of Apple.

A lot of the R&D in M1 is going into their IoT future products and Automobile products.

I would hope Apple offers this in the 2019 Mac Pro

https://www.amd.com/system/files/documents/amd-cdna-whitepaper.pdf

https://www.amd.com/en/products/server-accelerators/instinct-mi100

Anyone complaining about Nvidia should just stop. The MI100 Radeon Instinct is Passively Cooled to boot. I would love to see a Dual Radeon Instinct on a single Card option by Apple allowing 4 of these connected [Infinity Fabric 3.0 on these allow up to 8 connected] but on the Mac you could have a Dual in on slot and the regulard GPU in the other slot, each with Apple's TB3 quad on-board outputs. The W6900X would support a series of the 6K Displays and the CDNA 1.0 dual card with 64 GB and a pure compute monster offers an extremely long term value proposition for the Mac Pro market.

The white paper explains it all.

I can't speak about non-availability, but it is a completely useless update for the next 6-9 months with all my music equipment and software plugins. The same thing happened with Catalina. It never ceases to amaze how badly the past two major system updates have been so piss poor with coordinating efforts with professional software music companies.

22july2013 said:

David H Dennis said:

I will admit to not knowing too much about this, but my impression was that eGPUs were connected to Intel’s PCI standard, and therefore would only function with Intel chipsets.  So of course Apple Silicon would not support EGPUs.

If I understood the keynote correctly, it seemed like the Apple CPUs and graphics chips were designed to work directly together, cutting the overhead of external chipsets and therefore much faster and more efficient.  This means you are counting on Apple’s graphics engineers as your sole source for graphics developments.  

I was expecting to see a 16” MacBook Pro after the 13” MacBook Pro.  So I was disappointed that didn’t happen. But the logical conclusion is that the larger MacBook Pro systems are going to be considerably faster than the lesser models and therefore very much worth looking at.  I think we should pass judgement on Apple’s solutions here when the larger machine is introduced and benchmarked.

However, I’m tempted to buy a 13” MacBook Pro just so I can say I have it and am on the cutting edge ... just the typical programmer’s ego I’m afraid.

I concur with most of your post, although using the logic in your first paragraph we should not have been expecting Thunderbolt support, since Thunderbolt up until now "would only function with Intel chipsets." Somehow Apple managed to make Thunderbolt (v3) work, and (perhaps) they could have managed the same with PCI. Especially since the PCI standard is no longer managed by Intel but by PCI-SIG which has 800 members, but the primary members seem to be: Agilent, AMD, Dell, HP, Intel, Synopsys, NVIDIA, and Qualcomm. I don't see Apple in there.

Every computer vendor in the world supports PCI. the M1 for sure has PCI or it would be Dead on Arrival. PCIe is the defacto standard that DDR memory runs over. The ignorance of some who called this Intel's PCI standard is sad in the year 2020.