Well - in your mind it may be about…
BUT for me it is about years of experience with Intel SUCCESS vs AMD FAILURES in my personal experiences…
1 Like
Totally ridiculous funny argument
I’m not pro-intel, maybe I’m here to try and tempt myself the idea of an AMD build.
I am also not pro-intel (for other users) either, and I looked at this thread to discover where the performance strength currently are with CPU that are available. It is just that from my really bad past experiences with AMD systems that I bought, and with the lack of any testing evidence with the main program that my PC is built to perform with specifically Cubase. I am just going to stay with what I-me-personally-my-own-self have had better experiences with. I just don’t see why this thread should be labeled and “AMD” only thread. When the OP opened with the very first sentence referencing an Intel. And in the mean time I am still interested in the market trends.
I was looking for some overwhelming evidence that AMD is just way better with Cubase, and I just don’t see the evidence. And I am not going to say “again” what the evidence is that I can find.
I am still hoping that someone will quantify these performance differences instead of piling on anecdotal references
What would you consider to be a valid source?
Just anyone who could do a head to head compare on similar priced high-end AMD vs Intel at clock boost with the same Cubase 11 heavily loaded project with same audio interface and setup like same video card and settings - I would personally like to see this done with the i9-10900k vs Ryzen 9 5900x - I “think” those would be considered comparable. If this was in a you-tube well wouldn’t that be cool. I like what Dom did where he pushed till they broke - but those cpus were different generation and price, so that is not helpful to discern what to buy today.
Wouldn’t you be interested in that too?
Since we can quite well compare the CPU which Dom used with the present Intel generation (showing anything but breathtaking progress on Intel’s side), we DO have a rough guess about the result. As someone being able to count to three, I see no problem getting the basic facts for my AMD decision at all.
I understand acting on former bad experiences. On the other hand, I have had bad experiences all over the place with computer and music gear, and have learned to re-evaluate situations much faster in our fast changing age of technology.
On a side note, I also heavily dislike the power abusing monopolistic misbehavior of dominant companies ruling markets worldwide. So whenever two companies deliver roughly the same, at roughly comparable prices, as customer I would ALWAYS support the smaller one, to help kicking the bigger one out of bad monopolistic business opportunities, which they rarely resist to abuse.
And on another side note, in our times, I would of course also ALWAYS support more energy efficient solutions as much as I can.
1 Like
No you can’t. It’s two completely different computers. One is Doms own rig he has had for some amount of time, the other is likely a fully spec’d demo unit from a DAW builder. Different mobos, different bios tweaks, different GPUs, different cooling, different CPU gens, one could be overclocked and the other not, different RAM, could even be different Windows 10 versions.
Now that’s a funny argument! Maybe you’re not American, but I’d rather support an American company with American workers and investing in American factories and foundries, rather than a company using foreign labor with unregulated working conditions and safety and product safety.
The AMD computer is from SCAN which is a UK company who build audio PCs. The one he got is one if their offerings and they don’t overclock them for audio workstations. They are reasonably priced as well. Since they test their machines for audio you are unlikely to hit problems experienced by off the shelf or self build. I have used them a few times as well as self build. I’m from the UK so buying from USA doesn’t apply to me. I just want a stable machine built for audio.
If you take a look at scan.co.uk you can see exactly what they offer (I don’t work for them)
1 Like
Ok but think about what you (plural) are saying now. You’re saying “Oh, I had bad experiences in the past with Intel and so AMD needs to prove itself with the current generation before I choose their CPUs”. But then you look at a video where the latest most powerful Zen 3 CPU is compared to Intel and you have an objection - your objection is that the Intel CPU that’s compared is a previous generation.
Ok, so what though? Aren’t you judging AMD for it’s past failures (many years ago btw)? So if you can take the past into account when judging AMD why not do the same for Intel?
Dom says he’s seen at most a 40% increase in power generationally for Intel, and because Intel has messed far less with their architecture in ways that affected internal latency I think it’s actually a pretty fair bet to assume that the improvement is likely predictable. In other words if we can see IPC improvements in other tests that quite likely carries over to audio - whereas it didn’t necessarily do that to the same degree with AMD.
And if that’s the case then if we assume an absolute best case scenario then a 10900K would do 1.4x29= 40 @ 32 samples which is still only 70% of the 5950X from AMD. See what I’m saying? The AMD CPU is still clearly in the lead when assuming a massive 40% uplift over the 9900K.
But wait: I’m sure you’re thinking “But how do you know it’s only a 40% uplift from the 9900K to the 10900K - what if Dom’s experience isn’t true in reality in this case?” Fair enough. If we look at SoundOnSound they have the 10700K compared to the 3950X, and if you take the 64-sample buffer score of the 10700K and add two cores (assuming it scales linearly, which often isn’t the case) you’d get 1050 as a score, whereas the 3950x scores 1540. So that’s the 10xxx gen Intel architecture compared to the previous gen Zen 2+, and you already had the Zen 3 outscore the preceding Intel architecture.
Yeah, it’d be neat to see a new comparison, but even at that point you’d (plural) have objections. I’ve seen these discussions before - there’s always something wrong with “their” product.
Now, we can scale the difference down from 5950x to any other cpu you want and compare. Currently due to pricing it seems to me that Intel is the better buy at a certain price range. But if we’re discussing who’s currently producing the better chips it really seems to be AMD.
By definition the motherboards, bios tweaks, cooling and RAM would have to be different when comparing, so those objections always are there.
amount of differentiation is a variable - same brand mobo? same year/generations of release? both firmwares updated? hyperthreading/CMT enabled or disabled? etc
What were the setups for the Scanproaudio and Sound On Sound tests?
The latest Intel i9-11900K is even performing worse than it’s predecessor, in several tests on youtube and elsewhere, and where it is better, it’s not significantly. So conceding a guess of 40% performance increase over two generations, based on former years increase, was rather in benefit of doubt in favor of Intel.
Trying to throw nebula at well known relevant basic performance and energy efficiency facts, by talking about mobo variations and trying other kinds of hair splitting, doesn’t change much either.
So the result is: some of us will just move on to the significantly better AMD performance in 2021, and others will cling to Intel, for whatever (from my view in part really weird) reasons. Not much to discuss further.
Simple as that.
3 Likes
I’m curious, those of you running AMD processors with Win10, can you load the results of the LatencyMon at idle, no Wifi connected? 1/2 hour will satisfy me, here’s mine, i7-8700 CPU@ 3.2GHz, 32B RAM, all SSD Drives, Win10 Version 21H1, OS Build 19043.1165.
Here’s mine for you. Ryzen 3950x 3.6Hz, 64gb ram, Asus Prime X570-P motherboard, 2 x XPG SX8200 Pro 2TB M.2 drives.
Thank you for doing this. That’s 16 cores, right? My i7 is 6. I will assume there are areas where your AMD processor with so many cores will outperform mine but my attention is on Cubase. Do you have any iZotope programs? I’d be curious to see what the Cubase CPU meter can offer as a comparison because these programs are typically heavy CPU users.
I use Izotope plugins, mostly Ozone and RX. This computer is very powerful and stable, I’m happy (except for some Cubase 11 bugs that force me to stay on Cubas 10.5.20)!