i hope intel comes back to beat the m1

[cmaier]

We’ll find out for sure when AMD releases their earnings soon. If AMD’s margins also took a hit, took less of one, or none at all that will tell us a lot about Intel’s design wins vs AMD. But the biggest data point for now would be Intel’s much weaker margins in data center as well as consumer. Every analysis is pointing to AMD’s success as the primary culprit rather than pandemic related which makes sense given that market.

Also analysis has indicated Apple’s share of the premium market grew substantially given the M1 CPU advantage. IDC pegged it at twice overall market growth and while estimates and comparing Mac to PC sales is not always easy, at least it shows that there was growth in the premium consumer market to be had as well during the pandemic and Intel doesn’t look like it got any.

Btw I agree even if Intel is to fall eventually (which is far from guaranteed), Intel isn’t going anywhere for awhile. I’m just being somewhat more negative about where they are already from a business perspective. There are worrying signs (for Intel) that their engineering troubles are starting to come home to roost financially. This isn’t destiny (yet) but I think I peg them as being in more trouble now than you. It’s a matter of degree rather a qualitative disagreement.

P.S. not a business person at all, just random internet guy’s opinion, but others who do seem to know their stuff seem to share it (and informed it) so that’s why I share it

I have gotten pretty far in life by always betting against IDC’s numbers ;-)

 

[Krevnik]

This isn’t destiny (yet) but I think I peg them as being in more trouble now than you.

I think you might be reading more optimism in my responses then there actually is. I'm mostly not discounting that Intel has an opportunity to turn things around in the 3-5 year timeframe. We will see if they can do it.

It's more the folks saying Intel is already doomed that I don't really believe.

Intel isn’t going anywhere for awhile.

Exactly my point. The decline will be slow enough as it is, and the ability to capitalize on Intel's position will be limited for the next couple years. Intel could capitalize on the extra breathing room if they have a path forward.

Honestly their plan to open fabs to others sounds a bit like a "Plan B" to me. It's a way to make up lost CPU revenue in the coming years.

 

[AHDuke99]

A lot of these things seesaw back and forth. Remember Intel in the late 1990s and early 2000s with the Pentium III/IV? Those chips were terribly inefficient and the PPC would kick their asses. Then comes the Core Duo and it begins kicking the G4 when it came to efficiency.

Intel has been around too long and has too much of the market to not figure out how to fix their chips. I'd like to see more ARM adoption on the PC side of things because the one thing I am going to miss when I do upgrade to an ARM Mac will be the ability to dual boot when I need to do something.

 

[cmaier]

Intel has been around too long and has too much of the market to not figure out how to fix their chips.

They said the same thing about many other companies in the past.

 

[crazy dave]

I think you might be reading more optimism in my responses then there actually is. I'm mostly not discounting that Intel has an opportunity to turn things around in the 3-5 year timeframe. We will see if they can do it.

It's more the folks saying Intel is already doomed that I don't really believe.



Exactly my point. The decline will be slow enough as it is, and the ability to capitalize on Intel's position will be limited for the next couple years. Intel could capitalize on the extra breathing room if they have a path forward.

Honestly their plan to open fabs to others sounds a bit like a "Plan B" to me. It's a way to make up lost CPU revenue in the coming years.

Another possibility of course is that Intel doesn’t “die” but transforms the nature of its primary business (ie does an IBM). Very hard to pull off but possible. It would for all intents and purposes (relative to the consumer x86 cpu market) mean that Intel “died” even if the company is still around and even still big. So you can argue about distinction without a difference, but it is another possible outcome to the list of a) Intel dies completely or b) Intel thrives much as it is now. Intel is already bigger than just CPUs though that is by far their primary business now and for the near future. Some transformation of course is a given with IDM 2.0, but I’m talking about a hard pivot away from x86 CPUs. Again, very unlikely in the near future, but possible.

 

[crazy dave]

They said the same thing about many other companies in the past.

Aye and it’s a truism mentioned many times already but such declines can at first be glacial ... until they’re not.

 

[Digital_Sousaphone]

What year did this thread start in? 2019? 2018? Why are people still talking about Intel when they have been number 2 for a while now? M1 chips are great if you don't know what else is available other than the laughable chips Apple has gone with for the last 10 years. Apple kept the core duo market alive with their garbage low end offerings. You can make some absolutely fantastic graphs showing increases in performance when you've only been offering the bare minimum in Intel chips...

 

[cmaier]

What year did this thread start in? 2019? 2018? Why are people still talking about Intel when they have been number 2 for a while now? M1 chips are great if you don't know what else is available other than the laughable chips Apple has gone with for the last 10 years. Apple kept the core duo market alive with their garbage low end offerings. You can make some absolutely fantastic graphs showing increases in performance when you've only been offering the bare minimum in Intel chips...

Sure, sure.

 

[Digital_Sousaphone]

Sure, sure.

View attachment 1764552

That's precisely the type of graph I'm talking about. Now put in chips not provided by Apple and add the peformance leaders by AMD and you'll have a real comparison of the current state chip peformance. Thanks for proving my point. Much appreciated.

Geekbench 5 Results - Geekbench

Geekbench 5 Results - Geekbench

Dang, why isn't the M1 topping those charts?

 

[robco74]

Well, once again we are beaten by superior reasoning, well done. I mean, if a low power chip can't beat a server chip, why is Apple even trying? Obviously they should pack it in, sell the company, and return the money to the shareholders.

 

[ADGrant]

That's precisely the type of graph I'm talking about. Now put in chips not provided by Apple and add the peformance leaders by AMD and you'll have a real comparison of the current state chip peformance. Thanks for proving my point. Much appreciated.

Geekbench 5 Results - Geekbench

Geekbench 5 Results - Geekbench

Dang, why isn't the M1 topping those charts?

The 10900K was not provided by Apple (not were any other K variants) and is one of the fastest benchmarking I series CPUs Intel has ever released. It beats its 11 gen successor in several benchmarks.

 

[crazy dave]

That's precisely the type of graph I'm talking about. Now put in chips not provided by Apple and add the peformance leaders by AMD and you'll have a real comparison of the current state chip peformance. Thanks for proving my point. Much appreciated.

Geekbench 5 Results - Geekbench

Geekbench 5 Results - Geekbench

Dang, why isn't the M1 topping those charts?

A perfect post: How to tell someone you don’t know anything about computers while emphatically declaring you know about computers. Oh and the fact that you tried this on for size with @cmaier to boot ...

I love me some overclocked 5950x equipped MB Air ...

 

[cmaier]

That's precisely the type of graph I'm talking about. Now put in chips not provided by Apple and add the peformance leaders by AMD and you'll have a real comparison of the current state chip peformance. Thanks for proving my point. Much appreciated.

Geekbench 5 Results - Geekbench

Geekbench 5 Results - Geekbench

Dang, why isn't the M1 topping those charts?

1) none of the blue squares are provided by apple
2) why are you bringing up geekbench when the graph is specint (ps: spec is what CPU designers actually pay a little attention to when designing cpus. We do not think about geekbench, ever.)

 

[pasamio]

That said, Apple's history is what makes Catalyst and SwiftUI's current state a bit perplexing to me. They feel more like a Microsoft-released API than anything else Apple has done in 20 years.

Catalyst I think was an experiment at getting developers to think about the two platforms as a single code base going from the direction of the mobile devices where input is more constrained. They released a number of their own apps that way but I think many developers would rather the dual sales from iOS and MacOS. SwiftUI I see in a similar thought process but hasn't fully come to fruition.

 

[robco74]

Catalyst I think was an experiment at getting developers to think about the two platforms as a single code base going from the direction of the mobile devices where input is more constrained. They released a number of their own apps that way but I think many developers would rather the dual sales from iOS and MacOS. SwiftUI I see in a similar thought process but hasn't fully come to fruition.

Developers can still opt to distribute via universal app purchase without implementing Catalyst and/or SwiftUI.

 

[jdb8167]

A perfect post: How to tell someone you don’t know anything about computers while emphatically declaring you know about computers. Oh and the fact that you tried this on for size with @cmaier to boot ...

I love me some overclocked 5950x equipped MB Air ...

Apparently the macOS version of Geekbench 5.3.1 Tryout doesn't validate timings (neither does the Linux version). But the Windows version does. Guess what you get when you look at the single core performance of Ryzen 9's? "This benchmark result is invalid due to an issue with the timers on this system." These numbers are all the same as the macOS hackintosh Ryzen single core numbers. This tells me that they are all invalid. So try again.

 

[cmaier]

Apparently the macOS version of Geekbench 5.3.1 Tryout doesn't validate timings (neither does the Linux version). But the Windows version does. Guess what you get when you look at the single core performance of Ryzen 9's? "This benchmark result is invalid due to an issue with the timers on this system." These numbers are all the same as the macOS hackintosh Ryzen single core numbers. This tells me that they are all invalid. So try again.

View attachment 1764624

i feel like responding with some misinformed logic in all caps, but nothing is coming to mind.

 

[crazy dave]

Apparently the macOS version of Geekbench 5.3.1 Tryout doesn't validate timings (neither does the Linux version). But the Windows version does. Guess what you get when you look at the single core performance of Ryzen 9's? "This benchmark result is invalid due to an issue with the timers on this system." These numbers are all the same as the macOS hackintosh Ryzen single core numbers. This tells me that they are all invalid. So try again.

View attachment 1764624

Fascinating, I just assumed they were overclocked hackintoshes which would still be an invalid comparison on so many levels.

 

[jdb8167]

i feel like responding with some misinformed logic in all caps, but nothing is coming to mind.

Not sure what you mean. Geekbench browser has issues of people gaming it or mismeasuring because they overclock. I just went through several pages of top results and I didn't find reliable Ryzen 9 results until about page 35. Not saying all the previous results are bogus but it is a bit odd that only macOS hackintosh and Linux machines get above 1850 single core results but exactly no Windows versions do.

I did find one person who clearly knocked the Ryzen down to 1 core to get slightly higher single core numbers. Of course, any macOS numbers are suspect too since the macOS version also doesn't record timing validity. But I know my numbers aren't overclocked or gamed and I get 1730 or so on M1 MBA single core. So a Ryzen 9 reliably gets about a 6% boost over an M1 on single core.

Edit: I should also note that plenty of Intel Rocket Lake Core-9 results are valid at over 1950 single core.

 

[crazy dave]

Not sure what you mean. Geekbench browser has issues of people gaming it or mismeasuring because they overclock. I just went through several pages of top results and I didn't find reliable Ryzen 9 results until about page 35. Not saying all the previous results are bogus but it is a bit odd that only macOS hackintosh and Linux machines get above 1850 single core results but exactly no Windows versions do.

I did find one person who clearly knocked the Ryzen down to 1 core to get slightly higher single core numbers. Of course, any macOS numbers are suspect too since the macOS version also doesn't record timing validity. But I know my numbers aren't overclocked or gamed and I get 1730 or so on M1 MBA single core. So a Ryzen 9 reliably gets about a 6% boost over an M1 on single core.

His sarcasm wasn’t directed at you but rather posters like @Digital_Sousaphone who post benchmarks without understanding what they mean.

I think stock 5950x CPUs have about the same single thread performance as the M1 - a little more, a little less depending on the test and setup. Anandtech has the 5950x Geekbench 5 score at 1655 vs the M1 at 1745. Some other tests it beats the M1. It’s a wash. Of course a 5950x costs as much as an entire M1 mini (if you can get one at all) and the 5950x single core uses about as much power as the entire M1 SOC (exaggerated, but not by as much as one might think ?).

 

[jdb8167]

His sarcasm wasn’t directed at you but rather posters like @Digital_Sousaphone who post benchmarks without understanding what they mean.

I think stock 5950x have about the same single thread performance as the M1 - a little more, a little less depending on the test and setup. Anandtech has the 5950x Geekbench 5 score at 1655 vs the M1 at 1745. Some other tests it beats the M1. It’s a wash. Of course a 5950x costs as much as an entire M1 mini (if you can get one at all) and the 5950x single core uses about as much power as the entire M1 SOC (exaggerating but you know not by as much as you might think ?).

Well I'm sure you can overclock and still manage to keep your timers adjusted correctly. I don't really know because I'm not really into the whole overclocking PC thing but I assume that's how a few uploads seem to be valid at around 1950-2000 single core. I also found one or two Rocket Lake i9s that get in the 2000 range. But the Rocket Lake core i9s are consistently in the 1950-1980 range. It would be great if Geekbench could figure out how to validate macOS and and Linux tests. Right now many pages of their top results are suspect.

 

[crazy dave]

Well I'm sure you can overclock and still manage to keep your timers adjusted correctly. I don't really know because I'm not really into the whole overclocking PC thing but I assume that's how a few uploads seem to be valid at around 1950-2000 single core. I also found one or two Rocket Lake i9s that get that range. But the Rocket Lake core i9s are consistently in the 1950-1980 range. It would be great if Geekbench could figure out how to validate macOS and and Linux tests. Right now many pages of their top results are suspect.

Aye but even so, as mentioned by @robco74, way outside the bounds of reasonable comparisons to the M1 - overclocked or not, validated or not. The fact that the M1 CPU performs so well that it can be done so under certain circumstances is remarkable.

 

[Digital_Sousaphone]

1) none of the blue squares are provided by apple
2) why are you bringing up geekbench when the graph is specint (ps: spec is what CPU designers actually pay a little attention to when designing cpus. We do not think about geekbench, ever.)

That's fine. Can you show me the same data with AMD included? How does that look? I don't even care where the benchmarks come from as long as it shows the entire market. The one you included did not and is a perfect example of what I'm talking about. People tend to only focus on the chips Apple has sold and not talked much about the entirety of the market.

 

[Digital_Sousaphone]

A perfect post: How to tell someone you don’t know anything about computers while emphatically declaring you know about computers. Oh and the fact that you tried this on for size with @cmaier to boot ...

I love me some overclocked 5950x equipped MB Air ...

I didn't say anything about knowing a bunch about computers. Why are you even responding to me? Do you have anything to add or just blather? Maybe you two should get a room if that person means so much to you?

 

[cmaier]

That's fine. Can you show me the same data with AMD included? How does that look? I don't even care where the benchmarks come from as long as it shows the entire market. The one you included did not and is a perfect example of what I'm talking about. People tend to only focus on the chips Apple has sold and not talked much about the entirety of the market.

This included Intel, the number one in the market, and their super high end chips. This was not cherry-picking apple. Jeez.

Also, the graphs are intended to show the rate of change, year after year, of performance of apple vs. intel. It is not intended to show whether one chip is faster than another particular chip.

Nonetheless, Ryzen performs about as well, slightly faster, on some benchmarks than the equivalent Intel chips on this chart. I can’t find specint2006 for AMD’s recent chips. So what do you think that proves, precisely?

By the way - I actually designed CPUs at AMD for many years, (i assume you did not), so accusing me of somehow trying to hide AMD or something is pretty strange.