AMD Ryzen's First Game Optimization: 'Ashes Of The Singularity: Escalation,' Tested

[InvalidError]

that leaves 2 potential areas for testing: the CPU, & the GPU.

There are two more variables: game and driver optimizations. Would the R7-1700 vs i7-7700k performance gaps be similar when using an RX480 or a pair in CF/DX12 MDA? Some benchmarks that I have seen show Ryzen/RX480x2 performing a whole lot better than Ryzen/GTX1080, which seems to indicate that Nvidia has a driver bottleneck on Ryzen.

 

[falchard]

AMDs x370 chipset was also updated on the day these tests were performed.
I don't think AMD CPUs will be poorly optimized with a nVidia gpu. AMD used nVidias internally for testing since they didn't have their own high end gpu at the time and you generally want to test a wide variety of hardware.

 

[InvalidError]

I don't think AMD CPUs will be poorly optimized with a nVidia gpu.

Since the CPU runs x86 code, there is nothing for AMD to "optimize for nVidia" here. Nvidia's drivers on the other hand need optimizing for Ryzen and DX12.

 

[JETFOURLITRE]

Who's gonna shill out for a high end CPU and then put up with 1080p anyway? To me, 4K+ is the future so a Ryzen system with a high end graphics card that can compete with an intel system in gaming and then flog it in just about any other task, for less money, is the way to go. Who gives a crap about 4FPS at 1080P? Only those who can't see the whole picture. Ryzen will mature and get better. Intel will respond. Who benefits? All of us.

 

[InvalidError]

Who's gonna shill out for a high end CPU and then put up with 1080p anyway? To me, 4K+ is the future

But for a CPU review, performance at 4K where every CPU performs practically the same due to hitting a hard GPU bottleneck tells you absolutely nothing about the CPUs relative performance. CPU benchmarks are run at 1080p (some reviewers go even lower) to eliminate as much of the potential GPU bottleneck as possible - at least enough of it to believe with reasonable certainty that the GPU shouldn't be a major limiting factor in most cases.

 

[elbert]

Who's gonna shill out for a high end CPU and then put up with 1080p anyway? To me, 4K+ is the future so a Ryzen system with a high end graphics card that can compete with an intel system in gaming and then flog it in just about any other task, for less money, is the way to go. Who gives a crap about 4FPS at 1080P? Only those who can't see the whole picture. Ryzen will mature and get better. Intel will respond. Who benefits? All of us.

Me for one. I'm a modorate gamer thats heavy into development. I build game maps mostly for CSGO and it takes a while to compile maps, models, and work on textures at the same time. I stick to a mid level GPU so my maps are playable at reasonable fps.

 

[PaulAlcorn]

"The 1800X enjoys a 7% performance boost from the update and jumps up several positions."

How did you come up with 7%?
My English is not to good but I know the difference between "al least 7%" and "7%".

1800X CPU Test 1920x1080
New: 45.5
Old: 35.1

(45.5-35.1)/(35.1/100)=15.43%

This 15.43 is about the same percentage other sites have come up with (15-16%), only Toms calculated half the acceleration.
I usually don't comment this kind of "mistakes" and I'm really not a conspiracy guy, but reading Rayzen articles on Toms I can't help impression that you guys are not objective in AMD vs Intel "race".

Of course this is just my opinion.

We listed a 16.36% increase for the CPU test. I'm not sure where you found the numbers you used for your calculations.

As stated in the article, the 7% increase applies to the GPU test.

 

[rdgoodri]

Um "Ryzen 1800X will still be a bit behind the 7700K "

 

[PaulAlcorn]

Bonscott, and Tom's first review barely mentioned much about RAM speed and Rank and i had to go back and hunt for the DDR speeds used. I know they were rushed but Tom's is usually good about explaining that stuff.

We included a table on the opening page with memory requirements.

http://www.tomshardware.com/reviews/amd-ryzen-7-1800x-cpu,4951.html

 

[mason-the-deathbat]

It would be more exciting if they didn't focus on a game that looks like it should be as easy to run as League of Legends.

 

[jimmysmitty]

"The 1800X enjoys a 7% performance boost from the update and jumps up several positions."

How did you come up with 7%?
My English is not to good but I know the difference between "al least 7%" and "7%".

1800X CPU Test 1920x1080
New: 45.5
Old: 35.1

(45.5-35.1)/(35.1/100)=15.43%

This 15.43 is about the same percentage other sites have come up with (15-16%), only Toms calculated half the acceleration.
I usually don't comment this kind of "mistakes" and I'm really not a conspiracy guy, but reading Rayzen articles on Toms I can't help impression that you guys are not objective in AMD vs Intel "race".

Of course this is just my opinion.

Reading is important. You cut half of that off. Let me add it back there so we can read it properly:

We shifted gears to the GPU test, which reveals little post-update variation for the Intel line-up. The 1800X enjoys a 7% performance boost from the update and jumps up several positions.

The most important part is the first part which sets the context of the next part:

We shifted gears to the GPU test

So this is talking about the second set of slides right above it. In this the pre-patch tests have the 1800X at 55.46 FPS average and the post patch 1800X hit 59.41 FPS. That's is roughly 7.12% of a performance increase in the GPU test vs the pre-patched version.

You are referencing the CPU test, not the GPU test, where they stated a 16.36%, higher than what you calculated.

So how are they not being "objective"? It seems to me a issue on your end, not theirs.

If it is true that Ryzens are going into the next generation consoles this will ensure that it gets thoroughly optimized for.

How long will it take before people realize that console side optimizations mean nothing for PC games? Yes they run the same hardware base, x86-64. However, as I have said many times, they use two parts that are very different.

The XB1 has a custom Windows 10 kernel, based around the hardware and ERAM that PCs do not use, and a custom version of DirectX. To add to that the actual hardware used is a custom version of the chips that may or may not come to PC.

The PS4 uses a custom Unix kernel based on FreeBDS. It also uses a custom OpenGL API. Again built around the specific hardware they have.

Any optimizations done on consoles do not translate to PC.

Ryzen is just going to be like the Rx 400 series. Slower then the competition at launch but overtime through driver updates it'll get faster and catch up. AMD is back baby!

I am more interested in how long they can keep it up. Will they keep a decent release schedule to compete with Intel? I am sure Intel is going to put it into high gear, they don't want to lose any of the pie they have been enjoying. Intel is planning to move to 10nm while AMD is said to be waiting fro 7nm.

Intel also has other platforms coming up for updates. Will it compete with that as well? How long will it compete with that?

Where are the DX11 numbers for comparison? We all avoid DX12 like the plague because of low FPS.

Depends on the game. Hitman runs better in DX12.

Also the very beginning of the article stated they plan to release DX11 results later.

 

[mason-the-deathbat]

Hitman runs better in DX12.

I mean, you may be right but

I can't even get it to even open in DX12, yet 11 provides no problems with that. My GPU supports 12 and I am running Windows 10 (not to mention I play Deus Ex: Mankind Divided in 12), so that shouldn't be why.

 

[harly2]

Whats odd is my stock Ryzen 1800x system scores better than every single Toms test..

Ryzen is fast even in an unflattering setup but it really starts gaining that 5 to 10% when in a proper "un-crippled setup"

 

[harly2]

Toms gets 7% everyone else gets 20%. lol

 

[jimmysmitty]

Whats odd is my stock Ryzen 1800x system scores better than every single Toms test..

Ryzen is fast even in an unflattering setup but it really starts gaining that 5 to 10% when in a proper "un-crippled setup"

Toms gets 7% everyone else gets 20%. lol

Read my first reply to that. The 7% is specifically with the GPU test, not the CPU tests. And that makes sense considering that the GPU is the main stressor in that test so the CPU will not make that much of a difference unless it is just a really bad uArch design (i.e. Netburst or Bulldozer).

 

[mediasearchprogram]

Why are people claiming there are bottlenecks in the GPU when testing out the new patch? could someone explain to me how the GTX 1080 is getting bottlenecked by this benchmark?

 

[CRO5513Y]

It would appear the stakes have Ryzen!

 

[spdragoo]

Why are people claiming there are bottlenecks in the GPU when testing out the new patch? could someone explain to me how the GTX 1080 is getting bottlenecked by this benchmark?

It's not, & they're just yammermouth idiots?

 

[harly2]

Whats odd is my stock Ryzen 1800x system scores better than every single Toms test..

Ryzen is fast even in an unflattering setup but it really starts gaining that 5 to 10% when in a proper "un-crippled setup"

Toms gets 7% everyone else gets 20%. lol

Read my first reply to that. The 7% is specifically with the GPU test, not the CPU tests. And that makes sense considering that the GPU is the main stressor in that test so the CPU will not make that much of a difference unless it is just a really bad uArch design (i.e. Netburst or Bulldozer).

Instead of read my response how about write a decent clear review. I also completely out benchmark everything you do on a no overclock 1800x in all CPU dependent tasks. Why AMD is doing a Q&A with this website is beyond me. Same story.

A number of good youtube reviewers have been doing some reasonable research on Toms benches over time.Considering their view count your click count is likely dropping.

 

[mediasearchprogram]

Whats odd is my stock Ryzen 1800x system scores better than every single Toms test..

Ryzen is fast even in an unflattering setup but it really starts gaining that 5 to 10% when in a proper "un-crippled setup"

Toms gets 7% everyone else gets 20%. lol

Read my first reply to that. The 7% is specifically with the GPU test, not the CPU tests. And that makes sense considering that the GPU is the main stressor in that test so the CPU will not make that much of a difference unless it is just a really bad uArch design (i.e. Netburst or Bulldozer).

Instead of read my response how about write a decent clear review. I also completely out benchmark everything you do on a no overclock 1800x in all CPU dependent tasks. Why AMD is doing a Q&A with this website is beyond me. Same story.

A number of good youtube reviewers have been doing some reasonable research on Toms benches over time.Considering their view count your click count is likely dropping.

http://www.hardware.fr/news/15076/maj-total-war-warhammer-ryzen.html
Im a tech youtuber, currently researching for my part 2 in ryzen has an achilles heel. I'm trying to correct any misinformation. I know in this benchmark so far there is no bottleneck on the GPU, although others say otherwise. <self promotion removed by moderator>

 

[jimmysmitty]

Why are people claiming there are bottlenecks in the GPU when testing out the new patch? could someone explain to me how the GTX 1080 is getting bottlenecked by this benchmark?

At higher resolutions the GPU becomes the bottleneck and the load is lifted from the CPU. At lower resolutions the CPU becomes the bottleneck and the GPUs load is not as heavy due to this.

That is what they are saying, not that the GPU is a bottleneck just because but if you game at higher resolutions and settings the GPU will become the main factor.

The reason sites will use a GTX 1080 and low resolutions and settings is to make the CPU the main bottleneck to see what kind of performance it has and will possibly have when GPUs advance.

Whats odd is my stock Ryzen 1800x system scores better than every single Toms test..

Ryzen is fast even in an unflattering setup but it really starts gaining that 5 to 10% when in a proper "un-crippled setup"

Toms gets 7% everyone else gets 20%. lol

Read my first reply to that. The 7% is specifically with the GPU test, not the CPU tests. And that makes sense considering that the GPU is the main stressor in that test so the CPU will not make that much of a difference unless it is just a really bad uArch design (i.e. Netburst or Bulldozer).

Instead of read my response how about write a decent clear review. I also completely out benchmark everything you do on a no overclock 1800x in all CPU dependent tasks. Why AMD is doing a Q&A with this website is beyond me. Same story.

A number of good youtube reviewers have been doing some reasonable research on Toms benches over time.Considering their view count your click count is likely dropping.

I think it is a case of poor reading comprehension. The second set of slides show new numbers and it literally says

We shifted gears to the GPU test, which reveals little post-update variation for the Intel line-up. The 1800X enjoys a 7% performance boost from the update and jumps up several positions.

The post I initially responded to left off the first part of the sentence, which is important as it sets the context, and jumped to the conclusion that Toms was being biased even though in the paragraph above the second set of slides and just below the first set of slides had the "16%" the poster had calculated.

The article was written just fine. I am almost wondering if the person purposely dropped the first line in order to cause some sort of controversy or just only saw that line.

Oh and AMD is doing a Q&A with Toms because besides you jumping to a random conclusion about nothing, Toms is one of the oldest and most respected tech sites out there.

On a side note, I highly doubt your doing anything vastly above Toms benchmark wise if you use the same hardware and settings. Had to bold it so that you don't misread what I said. It is near impossible to get vastly different results using the same exact setup beyond margin of error, +/- a few %.

Also, I am not part of the staff. I just moderate for free in my spare time. I have no bearing on the reviews or how they do them.

 

[spdragoo]

And we greatly appreciate the work that you & the other moderators do on the forum, as it helps maintain a certain level of civility & quality to match the overall reputation of Tom's Hardware.

And no, before anyone thinks to ask, I don't even moderate here, let alone work here. I've just been enjoying the site for some years now as a consumer & participant.

 

[mediasearchprogram]

@jimmysmitty thanks for the reply, as I've heard from AMD themselves running at higher resolution 1440P and 4K you would get close to similar FPS (AMD or Intel) possibly due to bottlenecks on the GPU, but 1080P, you won't. I asked this because a certain youtuber has made a statement saying the reason they only (Toms hardware) got 7% is due to possibly a GPU bottleneck.. I was thinking, what is this individual saying... so now I'm making a video to correct any misinformation out there.