Hardware Unboxed - BF 5 benchmarks

5150Joker

5150Joker

Supreme [H]ardness
Joined
Aug 1, 2005
Messages
4,568
6403C197-7B3D-4136-A191-CEBEC7962792.png
A6FF2A5B-78D4-4216-A97C-03EF23E4A1A7.png


This is why I can’t take AMD seriously in the high end CPU market for consumers, they get pummeled in games. Yeah Ryzen has cores but that’s about it really. The game might be in a beta stage but I doubt this result will change much between now and final release.

Video:
 
The 7700K/8600K being 25% faster than the 2700X seems a bit suspicious. I watched the video without audio -- did he mention anything about OC's?
I don't recall Battlefield 1's results looking like this.
 
Last edited:
When the average framerate is near 100 for every chip, and the minimum doesn't go under 60 except for the lowest tier...well there's something to be said for having as much fps as possible, but it's not something I much care about as long as it's not garbage. Guess I'm not an enthusiast. :/
 
I saw a test like this recently, except with a different game. Results looked very similar to this when testing with a 1080ti. But at the end they reran all the test with a 1070. Those results were within a few frames. I’ll try to find it and add it to this.
 
CPU benches are tighter when you impose more GPU bottlenecks... The CPU gap is going to get wider in a few weeks when the 2080 Ti comes out. Even though 99.9% of PC gamers won't be using it.
 
Nobu said:
When the average framerate is near 100 for every chip, and the minimum doesn't go under 60 except for the lowest tier...well there's something to be said for having as much fps as possible, but it's not something I much care about as long as it's not garbage. Guess I'm not an enthusiast. :/
Click to expand...

FPS players these days want 120+ fps so it matches high hz panels. 60 fps just doesn't cut it anymore.


TaintedSquirrel said:
CPU benches are tighter when you impose more GPU bottlenecks... The CPU gap is going to get wider in a few weeks when the 2080 Ti comes out. Even though 99.9% of PC gamers won't be using it.
Click to expand...

Maybe not yet but people tend to keep their CPU for 3-5 years so GPUs being removed from the bottleneck is a good thing as it exposes which CPUs have better longevity. I've never been impressed by Ryzen and I couldn't understand why people where shitting on Intel when they're still clearly the performance leader by a sizable margin.
 
I saw the exact thing when I tested both Ryzen 1600 @ 4Ghz and 8700K @ Stock on my 1080 Ti when I had both systems on-hand.. except game was pubg and the difference was at 1440p. When the battle royale mode for BF5 comes out you're gonna wait as much cpu horsepower as possible if it's more demanding than this benchmark they are using.
 
tangoseal said:
Boy this thread is so dumb. I'm reading for sheer entertainment.
Click to expand...

Nice ninja-edit :)

I will add that that the benchmark carries more weight than usual for a pre-release as the engine used in BFV is the same mainline Frostbyte used in many games recently.

On the flip side, it's still pre-release, and it still isn't [H[.
 
IdiotInCharge said:
Nice ninja-edit :)

I will add that that the benchmark carries more weight than usual for a pre-release as the engine used in BFV is the same mainline Frostbyte used in many games recently.

On the flip side, it's still pre-release, and it still isn't [H[.
Click to expand...

Yeah I ninja edited. I dont want to be controversial lol

I own both AMD and Intel and I just cant be as blindly polarized as some of the folks on the site.

It's almost as bad as single issue voters in American politics. Wheres my shears.... got some wool to collect.
 
5150Joker said:
.....

I've never been impressed by Ryzen and I couldn't understand why people where shitting on Intel when they're still clearly the performance leader by a sizable margin.
Click to expand...

Probably because in games that are finished and released, the difference isn't nearly so big .

BF1, 1080p, dx11 ultra, 1080ti, 8700k @ 5.0, 2700x @ 4.2. 11% faster on the min, 1.2% on the avg.

https://www.techspot.com/amp/review/1655-core-i7-8700k-vs-ryzen-7-2700x/

We all know that there's a lot of polishing to do before release, the idea that this is indicative of final performance is pretty silly.
 
Anyone who follows hardware and gaming has known for a long while now that low res ( yes 1080p is entry level and low gaming res in 2018) high refresh gaming has favored IPC and Intel.
Since moving to 1440p/144hz on my personal rig and using combinations of AMD/Intel/Nvida/Radeon I can get over 100fps on any game I play with dialed settings with any combination of hardware that is mid to high end.

2600 and a 1080? no problem
4770k and that 1080? no problem
2600 and now Vega 64? no problem


This whole thread seems to be a search for some justification that the OP has for buying Intel vs Ryzen. Or just stoking a flame war.

Buy what you like and enjoy.

Damn.
 
TheHig said:
Anyone who follows hardware and gaming has known for a long while now that low res ( yes 1080p is entry level and low gaming res in 2018) high refresh gaming has favored IPC and Intel.
Since moving to 1440p/144hz on my personal rig and using combinations of AMD/Intel/Nvida/Radeon I can get over 100fps on any game I play with dialed settings with any combination of hardware that is mid to high end.

2600 and a 1080? no problem
4770k and that 1080? no problem
2600 and now Vega 64? no problem


This whole thread seems to be a search for some justification that the OP has for buying Intel vs Ryzen. Or just stoking a flame war.

Buy what you like and enjoy.

Damn.
Click to expand...

Yeah, none of these seem like crippling performances, and as stated the standard as become 1440p+ where things become more GPU Dependant it will matter less

I wonder how my X5670 with GTX 1080 will run this at 1080p. I have a 1440p monitor but it is only 60hz.
 
thebufenator said:
You know its a great engine when I5's perform better than I7's in this day and age.
Click to expand...

Looking again, I see the 7600k running at what I presume is stock keeping up with the Ryzen CPUs... that's only 4C|4T!

I have two lying around, one running a server, one running Linux on metal. No idea that they were doing so well up to the "1%" level, but of course that doesn't show you minimum frametimes and lets occasional hitching go unreported, so I would expect the Ryzen CPUs to be smoother (as well as anything with 6+ threads).

Oh, and at least one of those 7600k CPUs can run at 4.7GHz :D
 
Those results are all over the place. Look at i3 8350k, performance is better when paired with a vega than a 1080 ti, same with 7600k.
 
TurboGLH said:
Probably because in games that are finished and released, the difference isn't nearly so big .

BF1, 1080p, dx11 ultra, 1080ti, 8700k @ 5.0, 2700x @ 4.2. 11% faster on the min, 1.2% on the avg.

https://www.techspot.com/amp/review/1655-core-i7-8700k-vs-ryzen-7-2700x/

We all know that there's a lot of polishing to do before release, the idea that this is indicative of final performance is pretty silly.
Click to expand...


Right from the link posted by TurboGLH and I couldn't agree more. Either way anything halfway decent will get you a nice gaming experience with a good GPU nowadays. Horray for choices!

So it stands to reason that BFV will get more parity once they polish it.

In a nicely optimised BF1 we see much closer results. Other games still favor that 800Mhz clock advantage Intel has. Assassins Creed was still heavily favoring the 8700k at 5.0 for instance.


BF1.png
 
TurboGLH said:
Probably because in games that are finished and released, the difference isn't nearly so big .

BF1, 1080p, dx11 ultra, 1080ti, 8700k @ 5.0, 2700x @ 4.2. 11% faster on the min, 1.2% on the avg.

https://www.techspot.com/amp/review/1655-core-i7-8700k-vs-ryzen-7-2700x/

We all know that there's a lot of polishing to do before release, the idea that this is indicative of final performance is pretty silly.
Click to expand...


Considering BF5 is due out in November, I doubt Ryzen will close this gap by then.
 
It's more likely an issue with his setup/bench method rather than the game itself. Wait for other sites to bench it.
 
BF5 is running on the same engine as BF1. There shouldn’t be this large a discrepancy between the two titles for performance. I’m not buying these benches.
 
5150Joker said:
Considering BF5 is due out in November, I doubt Ryzen will close this gap by then.
Click to expand...

You also seem to think that this is indicative of ryzen game performance, and the evidence clearly shows that it's not. So I'm comfortable waiting until then before deciding which is better
 
TaintedSquirrel said:
It's more likely an issue with his setup/bench method rather than the game itself. Wait for other sites to bench it.
Click to expand...

Wasn't able to view the video, but I tend to agree it is setup dependent, we all know ram bandwidth and latency matters more for Ryzen platform than Intel.
 
Endurance-vm said:
BF5 is running on the same engine as BF1. There shouldn’t be this large a discrepancy between the two titles for performance. I’m not buying these benches.
Click to expand...

Same base engine, but DICE upgrades the graphics every time, and they try to add more background stuff like destruction as well. So while it should still run well, newer games and newer versions of Frostbyte should also be more CPU intensive. Of course, that doesn't mean that higher demands from the engine are causing the delta we're seeing above.

TurboGLH said:
You also seem to think that this is indicative of ryzen game performance, and the evidence clearly shows that it's not. So I'm comfortable waiting until then before deciding which is better
Click to expand...

Honestly it really is indicative of Ryzen game performance, just in a fairly CPU-bound scenario. Ryzen presents lower IPC and lower clockspeeds which means lower per-core performance, while at the same time being more sensitive to internal and external latencies. Simply put, Ryzen is slower at the type of work that games resemble most.
 
IdiotInCharge said:
Same base engine, but DICE upgrades the graphics every time, and they try to add more background stuff like destruction as well. So while it should still run well, newer games and newer versions of Frostbyte should also be more CPU intensive. Of course, that doesn't mean that higher demands from the engine are causing the delta we're seeing above.



Honestly it really is indicative of Ryzen game performance, just in a fairly CPU-bound scenario. Ryzen presents lower IPC and lower clockspeeds which means lower per-core performance, while at the same time being more sensitive to internal and external latencies. Simply put, Ryzen is slower at the type of work that games resemble most.
Click to expand...

Honestly, it's not, claiming something is so doesn't make it so.

The link I posted above shows that quite clearly. Average at at 1080p is 9% slower, not the 22.5% shown here. It's clearly an outlier, and almost certainly will be corrected before launch.

Summary, in case you don't feel like reading the whole article .

https://www.techspot.com/amp/review/1655-core-i7-8700k-vs-ryzen-7-2700x/page8.html
 
TurboGLH said:
Honestly, it's not, claiming something is so doesn't make it so.

Average at at 1080p is 9% slower, not the 22.5% shown here. It's clearly an outlier, and almost certainly will be corrected before launch.
Click to expand...


I'm saying that Ryzen is slower in CPU bound gaming workloads- not that this is necessarily an example of this, or that this delta matters for various use cases when GPU limitations are thrown in.
 
TurboGLH said:
Honestly, it's not, claiming something is so doesn't make it so.

The link I posted above shows that quite clearly. Average at at 1080p is 9% slower, not the 22.5% shown here. It's clearly an outlier, and almost certainly will be corrected before launch.

Summary, in case you don't feel like reading the whole article .

https://www.techspot.com/amp/review/1655-core-i7-8700k-vs-ryzen-7-2700x/page8.html
Click to expand...

Let's look at some CPU dependent games like ARMA 3, TW, SC, FC Primal and others that clearly show why Ryzen falls behind when the GPU bottleneck is removed:

1.png FCP.png hitman.png jc3.png PUBG.png SC2.png TWSaga.png Wreckfest.png

Now as time passes and people upgrade to better GPUs (keep in mind most people are still at 1080P), the CPU bottleneck will be further exposed and it will show Intel leading Ryzen almost every single time by a sizable margin. Ryzen may be acceptable for 1440p and above where the GPU is typically the bottleneck but considering the vast majority of gamers are on 1080p, it shows it's weakness even today. Bottomline, I wouldn't go near Ryzen as a gamer.
 
Last edited:
5150Joker said:
Let's look at some CPU dependent games like ARMA 3, TW, SC, FC Primal and others that clearly show why Ryzen falls behind when the GPU bottleneck is removed:

View attachment 103330 View attachment 103331 View attachment 103332 View attachment 103333 View attachment 103334 View attachment 103335 View attachment 103336 View attachment 103337

Now as time passes and people upgrade to better GPUs (keep in mind most people are still at 1080P), the CPU bottleneck will be further exposed and it will show Intel leading Ryzen almost every single time by a sizable margin. Ryzen may be acceptable for 1440p and above where the GPU is typically the bottleneck but considering the vast majority of gamers are on 1080p, it shows it's weakness even today. Bottomline, I wouldn't go near Ryzen as a gamer.
Click to expand...

You picked 8 out of 35, selectively choosing info to support your position and ignoring the rest. Even at 720p the avg is only 11% slower. In addition, there's a 16% clock deficiency that accounts for some of the loss. You'll never be able to do anything about that, but it's there none the less.

As far as specific games at 720p, FC5 14% (clock speeds. Is on the same engine and doesn't suffer the same as primal), JC3 14% (clock speeds), PUBG 15% (clock speeds), ARMA 3 12% (clock speeds), Hitman 14% (clock speeds),

Total war 22%
FC Primal 20%
Wreckfest 21%
SC2 22%

So, of the games you listed, four have a disadvantage that can't be explained by clock speeds, one is a single threaded game thats 8 years old, another (FC Primal) clearly suffers from something related to the game, as it shares an engine with FC5, came out before FC5 and still performs worse on both intel and amd.

That's not a lot of info to support a pretty bold claim, that as the current generations go on, AMD cpus will age significantly worse than their same gen intel counterparts.
 
IdiotInCharge said:
I'm saying that Ryzen is slower in CPU bound gaming workloads- not that this is necessarily an example of this, or that this delta matters for various use cases when GPU limitations are thrown in.
Click to expand...

"Honestly it really is indicative of Ryzen game performance"

Your words, seems pretty clear that you consider this a good indicator of Ryzen gaming performance in CPU bound instances. Ryzen suffers a clock speed limitation that definitely hurts it in some CPU bound instances, but I reject the claim that it's it will manifest in an increasingly larger and larger delta as time goes on. Almost all of the cases can be explained by simple clock speed, game age, or bad programming. I gave more detail in my reply to Joker.
 
TurboGLH said:
You picked 8 out of 35, selectively choosing info to support your position and ignoring the rest. Even at 720p the avg is only 11% slower. In addition, there's a 16% clock deficiency that accounts for some of the loss. You'll never be able to do anything about that, but it's there none the less.

As far as specific games at 720p, FC5 14% (clock speeds. Is on the same engine and doesn't suffer the same as primal), JC3 14% (clock speeds), PUBG 15% (clock speeds), ARMA 3 12% (clock speeds), Hitman 14% (clock speeds),

Total war 22%
FC Primal 20%
Wreckfest 21%
SC2 22%

So, of the games you listed, four have a disadvantage that can't be explained by clock speeds, one is a single threaded game thats 8 years old, another (FC Primal) clearly suffers from something related to the game, as it shares an engine with FC5, came out before FC5 and still performs worse on both intel and amd.

That's not a lot of info to support a pretty bold claim, that as the current generations go on, AMD cpus will age significantly worse than their same gen intel counterparts.
Click to expand...


Selective or not, it shows an inherent weakness with Ryzen. Why settle for an underperforming CPU at all? Clock speed deficiency is an architecture limitation with Ryzen so that's another plus in Intel's favor and there's also IPC where Intel leads. You can try to claim the lack of performance is due to bad programming but I disagree, I think titles like the ones I showed highlight a fundamental design deficiency in Ryzen.
 
And what happens when streaming is added?

Oh right, Intel falls on its face due to a lack of cores.
 
thebufenator said:
And what happens when streaming is added?

Oh right, Intel falls on its face due to a lack of cores.
Click to expand...

Yeah because the majority of players are "streaming". And I am glad they aren't. Its bad enough looking at the ugly motherfuckers steaming now, the rest of you lot????:wtf:
 
Bankie said:
The amount of Streamers and Cinebench users that came out of the woodwork when Ryzen was released is quite amazing to say the least.
Click to expand...

Especially since streaming off of the CPU is an 'in between' solution. You need to be taking yourself seriously and just doing it for fun for that to be an appropriate solution.

Personally I'd have a separate box running Ryzen that would be doing the streaming :D.
 
TurboGLH said:
"Honestly it really is indicative of Ryzen game performance"

Your words, seems pretty clear that you consider this a good indicator of Ryzen gaming performance in CPU bound instances. Ryzen suffers a clock speed limitation that definitely hurts it in some CPU bound instances, but I reject the claim that it's it will manifest in an increasingly larger and larger delta as time goes on. Almost all of the cases can be explained by simple clock speed, game age, or bad programming. I gave more detail in my reply to Joker.
Click to expand...

So I'll be more clear: the performance presented above is indicative of Ryzen's gaming performance, but this benchmark of this build of BFV may not be indicative of Ryzen's gaming performance.

As in, this is the performance delta we should see when a game is CPU limited, but that BFV may just be unoptimized. Or maybe not! It's in beta!
 
You must log in or register to reply here.
Back
Top