Same Laptop, Different CPU - Part 2: Ryzen 4000 vs. Intel 10th-gen Gaming Battle

[Julio Franco]

This is part two of our investigation into Ryzen Mobile 4000 versus Intel's 10th-gen Core processors in the same laptop design. Before we had covered productivity performance and today we're rounding that off with a look at gaming.

Read the full article here.

https://www.techspot.com/review/2076-same-laptop-different-cpu-gaming/

 

[Irata]

Thanks for the second part.

One thing I‘d be really interested in is a comparison between this laptop‘s gaming scores vs. other similar models (e.g. Asus).

Looking at the results, unless Rocket League and Wolfenstein 2 are your main games, there is a minimal difference in gaming performance using overboost, so not really a saving grace.

Of course, if you only use the laptop for plugged-in gaming, don‘t care about anything else (including battery life), then the Intel based model would still be the slightly better choice, but for most other use case scenarios it looks like a more expensive poor second.

 

[sorten]

Excellent comparison reviews; thanks for the effort. It really helps to have similarly optimized systems for the comparison.

Looking at the two reviews it seems that Intel is a reasonable choice for gaming only scenarios, but AMD would be my choice because gaming is not the main reason I sit down at a computer. The 42% advantage for AMD in the code compile test would be enough of a factor for me.

 

[Shadowboxer]

This is just torturing the fanboys! These new AMD laptops are outstanding in every way except in gaming. Much like their desktop counterparts.

Gaming is Ryzens weakest area yet I constantly see tech reviewers doing gaming only tests etc. At this point it’s beginning to feel a bit cruel as the AMD fans seem to have a real bee in their bonnet about the fact that Ryzen is still clearly second best when running games.

Personally I’d take a Ryzen laptop over an Intel laptop every day of the week (if all other things are the same) and that’s because I don’t really play games on my laptop. Desktop is a different story, I only play games on my desktop so I’d pick Intel.

But still, leave AMDs inferior gaming performance alone!

 

[emgarf]

I'd love to see a comparison of boot times between the two configurations.

That's been the biggest annoyance of my recent swapout of MB/processor from Intel to AMD. Everything else - drives, GPU, case, USB peripherals, etc. remained the same; boot SSD was wiped and Win 10 Pro was clean-installed on the new config.

Old config: i7-7700K / GA-Z270-Gaming K3: 11 second boot to Win 10 login
New config: 3900X / TUF GAMING B550-PLUS: > 90 second boot to Win 10 login

My rendering speeds are great now. Eventually.

 

[Irata]

This is just torturing the fanboys! These new AMD laptops are outstanding in every way except in gaming. Much like their desktop counterparts.

Gaming is Ryzens weakest area yet I constantly see tech reviewers doing gaming only tests etc. At this point it’s beginning to feel a bit cruel as the AMD fans seem to have a real bee in their bonnet about the fact that Ryzen is still clearly second best when running games.
....
But still, leave AMDs inferior gaming performance alone!

Totally agree that Ryzen is still behind Intel when it comes to gaming on average in this review. That said, did we both read the same review ? Ryzen is behind on some games but also ahead on a few.

From the review:

The end result is that on average, the Intel laptop is just 3 percent faster

for average fps and for the 1% lows:

That smaller deficit on the AMD system when using Overboost is reduced to practically zero when viewing 1% low performance.

.

If anyone worries about being behind 3% on average fps and 1% on the min 1% that person really cannot be helped. Same for someone using a rather strong word like ‚inferior‘ for such a rather small difference.

 

[Vrmithrax]

I'd love to see a comparison of boot times between the two configurations.

That's been the biggest annoyance of my recent swapout of MB/processor from Intel to AMD. Everything else - drives, GPU, case, USB peripherals, etc. remained the same; boot SSD was wiped and Win 10 Pro was clean-installed on the new config.

Old config: i7-7700K / GA-Z270-Gaming K3: 11 second boot to Win 10 login
New config: 3900X / TUF GAMING B550-PLUS: > 90 second boot to Win 10 login

My rendering speeds are great now. Eventually.

That discrepancy seems more related to the other factors like a bad OS/driver setup or incorrectly configured motherboard. The CPU has very little to do with boot times, the computer's storage and OS configurations are doing all of the work during that process.

I recently dropped a new 3600 and X570 motherboard in to replace an older i5+mobo configuration on a secondary PC I had lying around, keeping all other components in place, and the Ryzen unit boots almost twice as fast as the old i5 unit did (10 seconds now from POST splash screen to Windows login). I didn't even reinstall Windows, but it was a very clean Windows install to start with, and the only time it lagged was on initial boot up after replacing that hardware, while it was getting updated drivers and optimizing for the new hardware configuration. So based on my anecdotal experience, I could make a generic "AMD boots much faster than Intel" statement. But that's not the case - the infrastructure around the CPU (chipset, drives, OS and programs installed, etc.) typically make or break your boot times.

 

[lipe123]

This is just torturing the fanboys! These new AMD laptops are outstanding in every way except in gaming. Much like their desktop counterparts.

Gaming is Ryzens weakest area yet I constantly see tech reviewers doing gaming only tests etc. At this point it’s beginning to feel a bit cruel as the AMD fans seem to have a real bee in their bonnet about the fact that Ryzen is still clearly second best when running games.

Personally I’d take a Ryzen laptop over an Intel laptop every day of the week (if all other things are the same) and that’s because I don’t really play games on my laptop. Desktop is a different story, I only play games on my desktop so I’d pick Intel.

But still, leave AMDs inferior gaming performance alone!

That statement is very loaded.
One could point out that you pay almost 100% more for a Intel CPU that gives you 3%-5% more fps. Also at that point we are typically talking past 100fps already so it would be impossible really to tell the difference unless you stare at a frame counting number.

It's not so much a "fanboi" argument as it is a value for money argument.

 

[Hamo231]

3% faster in gaming but up to over 40% slower in many work related stuff but about $100 more expensive?

Amd it is then.

 

[0dium]

I'd love to see a comparison of boot times between the two configurations.

That's been the biggest annoyance of my recent swapout of MB/processor from Intel to AMD. Everything else - drives, GPU, case, USB peripherals, etc. remained the same; boot SSD was wiped and Win 10 Pro was clean-installed on the new config.

Old config: i7-7700K / GA-Z270-Gaming K3: 11 second boot to Win 10 login
New config: 3900X / TUF GAMING B550-PLUS: > 90 second boot to Win 10 login

My rendering speeds are great now. Eventually.

1300x + cheapest a320 board + Kingston a series 120gb ssd. Boots in less than 10 seconds.
Must be something with your config.

 

[Someone8MyPosts]

In the first article for this, they said an AMD CPU rose to 90C. That's 194F. Take their results here with a grain of salt.

 

[brucek]

In the first article for this, they said an AMD CPU rose to 90C. That's 194F. Take their results here with a grain of salt.

Not sure I understand your point - are you skeptical GPUs and CPUs can get that hot? Without cooling, they'd reach far higher than that near instantly under load (and then halt from thermal fault, or die or if that didn't work). Ideally a good cooling system will keep them lower, but then again "good cooling" and "laptop" are somewhat exclusive terms. My 2017 first gen Gigabyte Aero laptop Intel CPU will easily reach and stay at 90 at max fan, because that's all the cooling that model has to give.

 

[brucek]

Another fascinating article, thank you again.

As someone who has bought exclusively Intel CPUs since Nehalem in 2008, I'd choose AMD for a laptop today. The productivity upside is much larger than the gaming downside, and I'm starting to believe we'll see more games in the next 2-3 years that do more with more cores. (That said I remain skeptical of gaming on laptops period, I much prefer desktop, my laptop is a last resort and increasingly I find I'll just watch a movie or read on it instead.)

 

[Mithrandir95]

Excellent article.

For your Rocket League numbers, assuming you tested at Medium (you show Medium quality in the combined charts, and talk about it in the text, but on the Rocket League graph says "Render Detail Quality, FXAA High").

In any case, with FPS this high, I'd be nice to see the "High" quality numbers. Then I think the "real world" CPU/GPU mix might favor Intel less.

No argument that the higher boost clock, and larger cache gives Intel the single and lightly-threaded win, but if we're looking real world testing, it would be nice to see the effect of the power-headroom AMD is giving to the GPU has on "real world" mobile gaming. (For that matter the Low, Medium, High numbers might show that well).

 

[Lionvibez]

That statement is very loaded.

That should be expected by now

And to Tech Spot both reviews were excellent thank you for taking the time.

 

[BBoogey]

Prize??? Vs x y?

 

[Evernessince]

I'd love to see a comparison of boot times between the two configurations.

That's been the biggest annoyance of my recent swapout of MB/processor from Intel to AMD. Everything else - drives, GPU, case, USB peripherals, etc. remained the same; boot SSD was wiped and Win 10 Pro was clean-installed on the new config.

Old config: i7-7700K / GA-Z270-Gaming K3: 11 second boot to Win 10 login
New config: 3900X / TUF GAMING B550-PLUS: > 90 second boot to Win 10 login

My rendering speeds are great now. Eventually.

That's an ASUS thing. They have been doing terrible is the BIOS department lately.

 

[Bp968]

All I own is intel equipment. That said, I really don't see "inferior gaming performance here", I see tests that seem designed to find the absolute outlier scenarios that no one in the "real world" would actually see. Lets look at the rocket league numbers: 31% looks like a huge difference until you see that its well past 100fps above the monitors top refresh rate. Lets see what the difference is on 1080p max settings? I'm going to guess both systems peg 144fps.

Or the many other tests running the system past the 144hz max of the panel or running games at low settings. Yes, we know higher frequencies give intel a single threaded advantage, but who actually plays games that way? Most of us target a specific framerate/resolution and crank settings until we hit that fps or settings are maxed out. Who plays a modern game at low settings above 120fps? Who even does it above 90fps except try-hards?

And thats the problem with so many of these "tests". The sites/magazines know seeing a bunch of charts with all the bar graphs being identical won't pull views so we get this stuff.

Sadly too many people look at these charts and see something that really isn't true (as demonstrated in the comments). They see "gamer=intel, productivity=AMD" when that's not at all what the take away should be. Who plays at low settings? Who plays with a fresh install and no extra software running (chrome, discord, antivirus, silly rgb software, etc)? The reality is if you take a system and set it up like the average gamer or enthusiast gamer you'd see performance that's either identical (in gaming scenarios) or significantly in AMDs favor if you have stuff in the background or if your running productivity software.

Tiger Lake might be a help for intel in the really ultralight category but it doesn't seem like something that will help them in this category or desktops, but they have some significant catch-up to do right now, and AMD is the better buy. It something they have most definitely earned. We wouldn't be settings here now with inexpensive 8 core CPUs and 6 core i5s if they hadn't come in guns blazing with Ryzen like they did.

 

[brucek]

Re: framerates above 60/90/target being irrelevant -- I care because I am using them as proxies for overall potential. Systems that can exceed target today on today's games are not guaranteed to be in the same position next year, which is why I'd want my purchase to have some headroom.

As to the conclusion that this article's data in particular do not indicate any massive advantage for Intel, I agree.

 

[Bp968]

Re: framerates above 60/90/target being irrelevant -- I care because I am using them as proxies for overall potential. Systems that can exceed target today on today's games are not guaranteed to be in the same position next year, which is why I'd want my purchase to have some headroom.

As to the conclusion that this article's data in particular do not indicate any massive advantage for Intel, I agree.

The problem is that advantage isn't really there, especially in a laptop where you can't upgrade a video card. So the higher gfx settings actually approximate the future games better because tomorrows low is todays high setting, so that gap will get smaller not larger. Plus as another guy said newer games seem to leverage core counts more and more so the AMD advantage will increase and Intels will decrease.

I long ago stopped "future proofing" systems. I remember a friend of mine back in 2000-2002 buying an 800$ CPU and a 700$ GPU and I bought a 300$ CPU and 350$ GPU. His was about 20% faster for 2-3 years. Then I upgraded to a newer 350$ CPU and newer 350$ GPU. So my total was 100$ less then his and my performance ended up being almost 50% faster and I had a second PC (though thats before buying more ram since I used ram from the previous system, so building a 2nd system brought the 2 system cost up to slightly more than his total for the first system).

Long story short, I buy for what im playing today in the settings I want, plus maybe a small margin for the future. Your just better off selling the old GPU and buying a new one usually. And with AMD you dont even need a new MB (though I expect the current socket to end by zen 4 if it isnt killed in zen 3)

 

[yeeeeman]

And, as expected Intel games better even with fewer cores.

 

[krizby]

Wow, so Intel is still able to pull ahead, this is surprising considering the 4800H is not even using its full 45W power limit in game, so there is a bottleneck of some sort in between. I guess this is down to Renoir CPU only support PCIe gen3 x8 for dGPU.

Would be interesting to see match up between Intel 10875H + 2080 Super vs AMD 4800H + 2080 Super that would alleviate 4800H's lower PCIe bandwidth with higher efficiency. If 4800H come within 3% of the Intel machine then of course it's AMD win with their superior productivity performance.

And I don't think it would be fair to compare a Max Q GPU on Intel Machine vs Max P on the AMD one. The Overboost result on the AMD machine show the GPU is already running at 86C, that is too high for comfort since the RTX GPU have max operating temp of just 88C. Using higher fan speeds or better GPU coolings on the AMD machine invalidate the results.

 

[Markoni35]

Old config: i7-7700K / GA-Z270-Gaming K3: 11 second boot to Win 10 login
New config: 3900X / TUF GAMING B550-PLUS: > 90 second boot to Win 10 login

There's something wrong with your drivers. Try upgrading/changing them, obviously they are waiting for something unnecessarily. Can't be anything else than a driver timeout.

 

[emgarf]

There's something wrong with your drivers. Try upgrading/changing them, obviously they are waiting for something unnecessarily. Can't be anything else than a driver timeout.

Thanks for the reply. The drives / peripherals in the system are identical; only the MB and CPU were changed, so not sure why there weren't driver issues before if there are issues now...
EDIT: ...the disk controller driver (or anything else related to MB functions) I guess would be different.

 

[Markoni35]

I suspect the MB driver. It's the most complex one.