Huch? Are you saying that Intel is shipping consumer products with reconfigurable codecs? Where?
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Intel Alder Lake vs. Apple M1
- Thread starter leman
- Start date
- Sort by reaction score
Huch? Are you saying that Intel is shipping consumer products with reconfigurable codecs? Where?
You...are...very...confused. Intel's answer for codec acceleration is quick sync. It's a fixed function part of their chips (asic), not a fpga, and it's slower than that on the M1 Max.
Apple sells a fpga: the AfterBurner card for the Mac Pro. It's $2000 and slower than the M1 Max.
Please try and do at least some basic research before you vomit nonsense here.
Jim Keller would say...
Apple needs to own an FPGA capability because of bitcoin miners?
I hope Apple does anything and everything in their power to make sure their products are horrible at crypto mining.
In any case, Intel's product line is mature. Apple OTOH, has only release one generation of their own silicon for the Mac. If the progress they've made on the A-series is any indication, there are still plenty of areas where they can improve. It's more embarrassing for Intel that Apple has been able to do so well right out of the gate. Especially considering that much of the high performance in modern PCs isn't because of Intel, but Nvidia.
In any case, Intel's product line is mature. Apple OTOH, has only release one generation of their own silicon for the Mac. If the progress they've made on the A-series is any indication, there are still plenty of areas where they can improve. It's more embarrassing for Intel that Apple has been able to do so well right out of the gate. Especially considering that much of the high performance in modern PCs isn't because of Intel, but Nvidia.
Wrong on so many levels.
GPUs are not ipso facto faster. They are only faster for very specific comptational use cases: if the task at hand is massively parallelizable.
Everywhere else they are much much slower.
Last edited:
So now it’s “Apple bad because they haven’t made an fpga acquisition like intel or amd?”
I gotta admit, that’s creative and fresh.
I gotta admit, that’s creative and fresh.
Exactly and frankly as much as I love Linux I would see no point to bootcamp it.
Intel has joined the RISC-V group.
Does Intel believe that its next ISA will be RISC-V instead of ARM?
Does Intel believe that its next ISA will be RISC-V instead of ARM?
VP of PA Semi so forefather of AS.
Currently, CTO of Tenstorrent which looks like a competitor to Google AI Tensor Processing Unit.
https://tenstorrent.com/grayskull/
LINUX DISTROS SUPPORTED
Currently, CTO of Tenstorrent which looks like a competitor to Google AI Tensor Processing Unit.
https://tenstorrent.com/grayskull/
LINUX DISTROS SUPPORTED
- OS: Linux, Ubuntu or Centos
- ML Frameworks: Pytorch native, Tensorflow and other frameworks via ONNX
- Buda: C++ based low level programming model for adding new ML ops, or non-ML applications
- Open: our graph compiler and programming model will be made open source
Last edited:
He left PA semi before apple bought them. And of course, PA semi didn’t design those early chips for apple by themselves - apple also bought Intrinsity and hired lots of folks from AMD and other places.
after he designed the M72 and A103 he left. All apple has to do is pop them out
I mean definitely didn't work on M1, but if he was their in 2012 he probably worked on its predecessor.
Sure, he probably had something to do with A4 and/or A5. Of course, Apple Silicon has very little to do with those chips. Not sure how long he was there or what, specifically, he worked on.
They have a lot of openings and AI is the space to be in that writes its own paycheck. Need CentOS/Redhat/Ubuntu Linux experience so people with x64 Macs have an advantage on getting up to speed.
https://tenstorrent.com/careers/
https://tenstorrent.com/careers/
You are aware of the fact that starting up
a Linux instance on ARM macs is as easy as starting one in an x86 one? With the main difference that the ARM Mac is much faster.
Right. But that’s technically “Apple Silicon” even if the term was never specifically applied to anything prior to M1.
That’s Apple silicon but not Apple Silicon, because, as you note, the latter term has only ever applied to M1-series chips (with an arguable exception for the A12z’s in the development kit?)
(and A4 wasn’t even Apple’s own architecture, so that’s hard to argue it’s even Apple silicon with a lower case ”s.”)
It's a fail on MBA M1. UTM emulation of preferred CentOS 7 x64 minimal is taking hours at post installation and there's a long delay with every mouse click or keyboard input. CentOS ARM64 virtualization doesn't even install. No free VMWare Player on MacOS. Don't have all day. Stopping and purging.
Instead, installed free VMWare Player x64 and CentOS 7 x64 minimal on x64 laptop in like 5 minutes and it just works.
Last edited:
In comparison context, a CPU is an ASIC relative to the FPGA that is emulating it, and FPGAs are slower than ASICs. The CPU interprets code and manipulates data in a specific way. But if you have a bunch of heavily used functions, you can distill the code down to gate patterns and instantiate them into an FPGA, and by eliminating the code interpretation part, the FPGA can do the data manipulation more efficiently than the CPU. Up to a point.
In terms of getting the actual work done, if you code the FPGA for the work, it will generally get it done much faster than a CPU, for a given function. Not as fast as hard-coding, but significantly faster than a CPU, within the limitations of dataflow availability.
My FPGA comment went way out of bounds... I was simply saying that in 10 years from now, users who are AS hesitant due to requiring some ancient Windows-exclusive software that must run 'natively' need not fret; then they can expect perfect emulation at likely higher speeds than the ancient software requires --- whether it's performed via software or hardware is irrelevant.