After months of silence, Intel has officially fired its first benchmark shots at Apple’s M1 processors.
It’s about time. I’m all for a knock down, drag out fight between Apple Silicon and X86. Everyone wins if these two can really push each other.
Can you even imagine how hot that intel was. Interesting how they left out the battery performance
Jason Snell sums up the article:
Inconsistent test platforms, shifting arguments, omitted data, and the not-so-faint whiff of desperation. Today’s M1 processor is a low-end chip for low-end systems, so Intel only has a small window to compare itself favorably to these systems before higher-end Apple silicon Macs ship and make its job that much harder.
The standout Excel benchmarks are interesting; would they mean that MS and Intel are working together to optimize certain workloads for Office?
That’s some impressive cherry picking.
Switching computers for different tests - yeah, you go Intel.
The benchmarks they show are relative as well. Do they matter? If Excel opens a spreadsheet in 250ms vs. 200ms, that’s 25% faster, which can look impressive on the right bar graph, but does it matter?
From my own experience, I can say my M1 MBP is a lot peppier and more responsive than my iMac Pro. I haven’t done any serious data analysis on the MBP, but in using Emacs, Obsidian, Safari, Brave, etc. it’s very responsive.
Excel has always been slow on macOS, this has nothing to do with the processor. Just bad MS code I guess?
It’s more than that though. Which their benchmarks aren’t showing. Excel is just not as good on macOS, nothing to do with the M1.
I use Excel daily on a PC and a Mac. It’s not just opening files faster, but on the Mac it doesn’t feel as responsive. On a PC, I can enter an easy sum formula and it feels quicker, more responsive. Just moving navigating a workbook feels better on a Windows PC.
And then you have charts. I have one spreadsheet that has about 6 charts at the bottom. I have to scroll down to see them. On Windows, they are already drawn and ready to go when I scroll down. On my Mac, they are not there, and it takes a second or more sometimes for them to pop in.
It is perfectly usable on a Mac, but if I am going to be building a new spreadsheet, I use my Windows machine, not my Mac. I just do data entry and maintenance on my Mac.
And another thing
If they wanted to (literally) compare Apples to Apples, they could. Top of the line Intel MBP vs. M1 MBP - same software, same benchmarks. But that isn’t what they wanted.
The x86 architecture is 43 years old. I think they’ve milked that cow for all it’s worth and they’re getting worried.
Surely you’re not in denial?
Just like PowerPC?
Just like PowerPC?
And? PowerPC has been discontinued some 15 years, it seems.
What a load of claptrap this comparison is. They’ve taken Chrome - which is one of the worst performing app on the Mac in terms of resource usage (if not the worst!) - and tried to make this out as a good benchmark! Most Mac users I know (in the real world, not online) don’t even have Chrome installed.
Office is also super-optimized for Windows and therefore is bound to perform better.
And what difference would it make to us anyway?
Most Mac users I know use Chrome as their main browser. Of course that is mainly because most of the Mac users I know are my family who use pass me down Macs I give them. They refuse to listen to me to just use Safari. Plus, they still use PCs as well, so they want one browser for everywhere.
It does seem popular though. Especially for average users.
That wasn’t my point. My point was that PowerPC began to stagnate due to a number of limitations (most of which forced Apple to switch to Intel in the first place back in 2006). Intel is experiencing the same issue now.
We don’t actually know for certain that higher end Apple Silicon chips will exist, nor whether the iteration will be, ahem, “fast and furious”.
But, I certainly expect so - based on zero Apple announcements.
I don’t care. I bought an M1 and the computer is beyond awesome. Yes, some benchmark will show that this or that application or task has better performance on another processor architecture. So what?
Essentially it’s bad for us if there’s just one architecture. Imagine what was going on in Intel`s board rooms after the announcement of Silicon? They need to up their game. It’s bad for customers overall if there’s just one processor architecture, one OS, etc.
AS for M1: yes, there may be some performance gains and faster benchmarks. But consider MIPS/W. For the first week I thought the battery indicator on my M1 was broken or stuck…
What are they going to put in the Mac Pro then? I see no reason why Apple wouldn’t create an M1X chip and an M2 chip and so on and so forth…
Speculating, I would imagine they saw the need to do damage control, more so than lost revenue. I would imagine apple was a small percentage (5%?) of their x86 and Xeon sales. Whereas releasing a processor that is fast and efficient undermines their perceived reign as king of processors.
Hence the “benchmarks.”
Addressing the problem as a percentage of revenue would be one possibility. If a random PC laptop maker decided to go with ARM, they would have cared less. Apple is, even if not substantial by revenue, substantial by “image”.
As for further developments…no idea where we are heading to. ARM (ARMv8-A) has already shown it’s potential for high-performance computing. A constraint might not be the architecture (x64/amd64 vs. arm64) but the SoC design. While SoC makes sense for laptops and deskops in the same performance category, a new “Mac Pro” might be ARM but not SoC.