Intel vs. AMD

[SlowBurningEcho]

I remember back in the days it was all about Intel for Music Production. Is that still a thing?

[Freaqpeak]

In short: no, it's the other way around now.

Latest AMD Ryzen 5000-series is faster both single core and multicore compared to Intel.

Also, Apple is moving towards ARM-based CPU's, they might be leaving Intel entirely after next generation of Apple Silicon, if they can compete with the best Intel/AMD CPU's.

[Freaqpeak]

To back that up, here's some results:

Results for 32 threads (16 cores):
https://valid.x86.fr/bench/32

Results for 1 thread:
https://valid.x86.fr/bench/1

[Burillo]

didn't AMD have higher system latency due to specifics of their cache architecture?

[Passing Bye]

They all perform great for audio, at least most recent incarnations of all 3 (Intel, AMD and Apple Silicon), saw in your other thread that you are looking for laptop, so if you budget is over 1k, you can't go wrong with any of them (CPU wise), personally for me it would be Apple Silicon>AMD>Intel for laptop.

[ghettosynth]

With laptops I think that the CPU brand takes a back seat to other factors. Apple Silicon is great, but, still has a lot of compatibility issues. Also, don't forget the apple tax on hard drive space.

[SlowBurningEcho]

They all perform great for audio, at least most recent incarnations of all 3 (Intel, AMD and Apple Silicon), saw in your other thread that you are looking for laptop, so if you budget is over 1k, you can't go wrong with any of them (CPU wise), personally for me it would be Apple Silicon>AMD>Intel for laptop.

I am looking for a laptop. Multimedia; Video, Audio, DJing, and etc.

I was looking at this
https://www.dell.com/en-us/member/shop/ ... n5510erets

But I like this better
https://www.dell.com/en-us/member/shop/ ... d4bf1b86c0

But now I am looking at this
https://www.dell.com/en-us/member/shop/ ... db0c9c0ffb

I am undecided because I don't know which will be a better system for me.

What do you think??

Should I go will dell laptop or another??

[Pictus]

didn't AMD have higher system latency due to specifics of their cache architecture?

AMD 5950X vs Intel 7980xe vs Apple M1
https://gearspace.com/board/showpost.ph ... ount=14307
https://gearspace.com/board/showpost.ph ... ount=14262
https://gearspace.com/board/showpost.ph ... ount=14256

[Passing Bye]

But I like this better
https://www.dell.com/en-us/member/shop/ ... d4bf1b86c0

This one is quite nice, I would pick this one over the other two most definitely, price is nice and it looks gorgeous. But wonder about DPC latency and how it performs there...

Here's something to explore further
https://www.notebookcheck.net/DPC-Laten ... 376.0.html

Seems like this one is safest bet concerning DPC latency, if you are going to DJ with it, than it's quite important aspect for you

https://www.dell.com/en-us/shop/dell-la ... n9305epfns

[SlowBurningEcho]

Seems like this one is safest bet concerning DPC latency, if you are going to DJ with it, than it's quite important aspect for you

https://www.dell.com/en-us/shop/dell-la ... n9305epfns

[/quote]

I was all about it until I saw the screen size.

I prefer +15.6"

[vstDOG]

measuring DPC latency on non tuned machines - especially laptops will yield unreliable results in the majority of cases.
These computers must be individually installed with a debloated fresh OS, hand picked drivers, (often some should be deliberately left not installed at all), BIOS related modifications are sometimes vital as well as Windows tweaks with an emphasis on power management and discrete graphics drivers.
PCs that come with firmware locked power settings are a nightmare to tame... Often impossible without the manufacturer's assistance in form of BIOS/firmware fixes.
And lastly, latency tests must be performed while the machine is offline with its LAN/WLAN drivers disabled... Or better yet totally disabled in BIOS.

[Serhii Kot]

didn't AMD have higher system latency due to specifics of their cache architecture?

AMD 5950X vs Intel 7980xe vs Apple M1
https://gearspace.com/board/showpost.ph ... ount=14307
https://gearspace.com/board/showpost.ph ... ount=14262
https://gearspace.com/board/showpost.ph ... ount=14256

This AMD CPU costs more than the entire mac mini with an m1 chip, not quite a fair comparison.

How about a more modern performance benchmark, I mean performance per watt.

[cleverr1]

And lastly, latency tests must be performed while the machine is offline with its LAN/WLAN drivers disabled... Or better yet totally disabled in BIOS.

What an impractical phaff that would be I wouldn't even consider a laptop for a DAW that has to have its network disabled to run decently as a DAW unless there was absolutely no alternative.

There are specialist builders/vendors that take all the pain away from buying kit to be used as DAWs.

[Passing Bye]

measuring DPC latency on non tuned machines - especially laptops will yield unreliable results in the majority of cases.
These computers must be individually installed with a debloated fresh OS, hand picked drivers, (often some should be deliberately left not installed at all), BIOS related modifications are sometimes vital as well as Windows tweaks with an emphasis on power management and discrete graphics drivers.
PCs that come with firmware locked power settings are a nightmare to tame... Often impossible without the manufacturer's assistance in form of BIOS/firmware fixes.
And lastly, latency tests must be performed while the machine is offline with its LAN/WLAN drivers disabled... Or better yet totally disabled in BIOS.

So best is to buy already tested and dedicated audio laptops or ones that are proven to have acceptable DPC OTB, everything else seems like too much work and lottery, what you are writing doesn't really encourage anything else than that.

[vstDOG]

What an impractical phaff that would be I wouldn't even consider a laptop for a DAW that has to have its network disabled to run decently as a DAW unless there was absolutely no alternative.

That's not what I meant...
Disabling the NICs while testing latency is a measure that would provide a base result which eliminates otherwise unreliable ones due to the OS performing networking related activities in the background hence throwing that measurement sample totally out of whack and by that potentially disqualifying a perfectly good, low latency machine all for the wrong reasons.
If a test comes out good with NICs disabled you could then conclude you may potentially have a perfectly worthy real time audio PC candidate in your hands.
And from there certain optimisation tunings can be implemented in taming down the OS's networking related background activities, as well as tweaking the NIC drivers themselves (disabling power saving features for example), then retesting with LAN/WLAN enabled to reliably determine actual real world working scenarios.

So best is to buy already tested and dedicated audio laptops or ones that are proven to have acceptable DPC OTB, everything else seems like too much work and lottery, what you are writing doesn't really encourage anything else than that.

I totally hear ya...
It's indeed a tough call...
Based on my own personal experience, for real time audio, one should generally avoid laptops designed for low power usage in mind for the purpose of prioritising battery runtime usage over raw performance.
The less 'exotic' functionality in a unit the better. Less special drivers usually translates to lower latency.
I've encountered much better results with AMD's 4000H/5000H chipsets than team blue's offerings when it comes to DPC latency.
But with all of that said, one should do their own due diligence in the way of research... Reading reviews and scouring the internets for relevant comments made by people who've already had first hand experiences with any particular unit that may have caught your interest.

Cheers,
vstDOG