Hacker News Reader

Comment by hu3 | Hacker News Reader

M1 is a beast.

But Intel 11th and 12th gen laptop processors beat M1 in single-core processing. See chart: https://i.imgur.com/KoNjGq5.png

Here's an article from 3 months ago comparing them: https://www.pcworld.com/article/608419/laptop-brawl-apples-m...

(that intel laptop beat M1 Pro in multi-core too)

(and M1's integrated GPU got destroyed by RTX 3080 laptops in OpenCL but that could be due to architecture/emulation)

I think it's not a fair comparison since M1 runs much more efficiently. Almost different purposes.

View on HN

You might be right but your citations of cinebench are misleading, geekbench has much different performance numbers, the m1 is only beaten by a brand new and much more power hungry/thermal cpu.

It’s also not exactly an indicator of real world performance which is why I avoided linking to benchmarks.

It's possible the M1 has been dethroned for single core performance, but what you link is not a good comparison.

Claims of future performance are going to be met with a clear shrug because ultimately I've been able to put my hands on an M1 for over a year, theoretical future CPUs are absolutely no consequence.

Additionally what is failed to mention here is how much battery life you get. Which, Apple is easily leading on.

I'm quite envious of the M1 to be perfectly honest, for the thermal envelope it's unbeaten, and its barely being challenged one year on. But you'll take my Dell Precision and my Linux installation from my cold dead hands.

I have Cinebench R23 installed and I can tell you 1T means 1 thread, thus single core as the article mentioned.

You can set how many threads the renderer will use. https://i.imgur.com/kBNvLe1.png

And it is pure CPU render. No GPU is used. It would be borderline comical if the article used a GPU benchmarking tool to measure CPU performance.

> Claims of future performance are going to be met with a clear shrug because ultimately I've been able to put my hands on an M1 for over a year, theoretical future CPUs are absolutely no consequence.

It is not theoretical. They ran their tests on a laptop that is already available for purchase.

When I replied there was commentary about AMDs upcoming 5nm CPU which seems to be gone now.

In fact, the entire tone of the comment has changed, not it's much more balanced.

Yes, the article shows a Geekbench win of 3% in single core performance. It fails to detail the enormous power consumption penalty Intel is paying to get that performance. That makes a huge difference in laptop usability.

Indeed the Geekbench was much closer. And I agree, the M1 does it with much better efficiency.

I wonder that gives Cinebench an advantage of almost 24% to Intel.

Reply on news.ycombinator.com