Discussion Ryzen 3000 series benchmark thread ** Open **

[piesquared]

Zen 2 and Navi both turned out to be excellent products. But where are the Ryzen and Navi combination reviews?? It seems pretty idiotic to have a simultaneous release and then reviewers ignore it??

 

[Fanatical Meat]

But you will not notice any difference between 150fps and 160fps so its not a selling point. And since 99% of people I assume do more than just play games hence amd wins. No reason to buy intel anymore.

Depends, if your goal is maximum frame rate and unlimited budget intel looks pretty good as for intel having clients with an unlimited budget is a good thing.
For the majority, I think most will chose 10 FPS lower and $100-250 cheaper that money can be put to a better video card or just pocket the cash. Which is a great volume proposition for AMD.
However it’s not that great to be the value leader.
Overall there is no answer that applies to everyone.

 

[Tup3x]

Extremely impressive in non-gaming workloads and good enough in gaming. Still behind Intel in gaming but they are getting close. X570 mobos are quite pricey though unfortunately.

I'll be keeping my 3770K another year. I'm interested to see what the next version can do and if the mobo situation would improve.

 

[coercitiv]

It's still faster in games.

Except some high FPS competitive games.

 

[IEC]

It's still faster in games.*

*At 1080p high refresh rate. Only some games. Loses in other applications. Uses more power. No stock cooler included. Costs more. Horse status: dead and beaten

Suddenly, the one major selling point left looks a lot less compelling. Especially when you have to sacrifice productivity performance and power efficiency for the sake of that last few %. Never mind that basically the entire HEDT lineup (AMD included) was upset by the 12c/24t 3900X. With a 16c/32t 3950X to follow...

Intel still has good products, but their pricing is going to have to come down to earth versus the perceived value of the Ryzen 3000 series. A $500 Ryzen 9 3900X 3950X that crushes productivity workloads and games within margin of error at 1440p+ is a big deal. A $200 Ryzen 5 3600 that comes within spitting distance of the $500 9900K is a big deal.

So really, everybody should be happy. Competition is back in a big way, and that means more bang for your buck.

 

[ubern00b]

It's still faster in games.

only 2 CPU's are "faster" in games from intel and that is 9700k and 9900k respectively at 3% and 5%, you must be talking about 1080p btw cause the difference between the 3 of them at 1440p is 3% and 0% at 4k

 

[gk1951]

IEC the 3900x is $500. The 3950x will be@$749.
Your point is well taken though. AMD has probably taken the lead in desktop cpu performance.

 

[moonbogg]

It's not just a 5-10% delta. It's as large as 20% + at 720p as shown by the Guru3D review. 720P matters because it frees up the CPU to do it's thing and you can compare them. It matters to high refresh gamers. Also, why would anyone buy a slower CPU anyway if all they want to do is game? People claim I'm trolling because my opinion reflects the data, lol. There are games where an i5 6600K (yes, a quad core skylake i5) is just as fast as these new CPU's in some games. It's in the Guru3D review.
I appreciate their review because they remove the GPU bottleneck sufficiently so you can actually see what the delta looks like. People argue that no one plays at 720p, therefore 720p results don't matter. That's a bad argument. It's obvious the reason people don't like 720p results is because it makes the AMD chips look disappointing, but those are the results. It's a good thing data is unbiased.
Yeah the chips are good at heavy workloads, but seriously that argument is less relevant to most people than the gaming discussion is.

 

[PotatoWithEarsOnSide]

...yet you're rocking a 4K monitor yourself.

 

[Topweasel]

But you will not notice any difference between 150fps and 160fps so its not a selling point. And since 99% of people I assume do more than just play games hence amd wins. No reason to buy intel anymore.

That's the thing. I mean it (9900k) doesn't even win every fight, this isn't even the "well in 1080p it's so much better for high refresh gaming that any serious gamer would get a Intel CPU". That is gone now. It's like the 2600k vs. 2700k if one of these two CPU's struggle in a game the other will too. 100MHz isn't going to save it. They are so close that that you have to look outside gaming to make a difference. Whether it's the extra compute on the 3900x or the lower cost of the 3700x/3800x. That's just the 9900k.

This gets worse as you down the line. All of the grasping at that ~5% game performance is completely wasted down the line. Anyone getting a 9700k down is making a huge mistake not looking at a Ryzen 3k alternative. Hell compared to the i5 lineup you can still get a Ryzen 7.

 

[ubern00b]

It's not just a 5-10% delta. It's as large as 20% + at 720p as shown by the Guru3D review. 720P matters because it frees up the CPU to do it's thing and you can compare them. It matters to high refresh gamers. Also, why would anyone buy a slower CPU anyway if all they want to do is game? People claim I'm trolling because my opinion reflects the data, lol. There are games where an i5 6600K (yes, a quad core skylake i5) is just as fast as these new CPU's in some games. It's in the Guru3D review.
I appreciate their review because they remove the GPU bottleneck sufficiently so you can actually see what the delta looks like. People argue that no one plays at 720p, therefore 720p results don't matter. That's a bad argument. It's obvious the reason people don't like 720p results is because it makes the AMD chips look disappointing, but those are the results. It's a good thing data is unbiased.
Yeah the chips are good at heavy workloads, but seriously that argument is less relevant to most people than the gaming discussion is.

Who buys a $500 chip to game at 720p? c'mon man reeking of desperation now, Ryzen 2 spanks 9600k/9700k and mostly 9900k in everything outside of gaming, see my post above for the difference in 1080p gaming and above, stop clutching at straws man, oh and it does it whilst using about 60-75w less

I take it you game at 720p then? my bad, in your case, yea, pay $500 for a 9900k with your GT 1030 and you'll be a happy bunny

 

[Asterox]

https://www.techpowerup.com/review/amd-zen-2-memory-performance-scaling-benchmark/

 

[coercitiv]

...yet you're rocking a 4K monitor yourself.

And a HEDT Broadwell that never made it as gaming king, but was productivity master for a while

Oh the beautiful times when 6700K and 7700K beat HEDT in gaming but nobody was actually upset about it... those were the days.

 

[IntelUser2000]

And since 99% of people I assume do more than just play games hence amd wins. No reason to buy intel anymore.

Don't think in terms of numbers. If you are gaming and you want the best, what do you get?

This is why the high end and the low end happen to be the higher volume ones. Back when Nvidia had their Geforce 4 Ti series, the Ti 4200 and 4600 both outsold the Ti 4400. Because either you want value out of the money you spent, or get the *best* that there is to offer. Same reason why you'd get the RTX 2080 Ti.

There are niche categories for gamers that need high fps. Like those that have 144Hz or higher monitors. Or VR uses. In terms of overall, they are small, but for the people buying 9900K, it should be a high portion.

Plus applications that poorly scale with cores or are under 16 threads, the 9900K will do very well.

It's as large as 20% + at 720p as shown by the Guru3D review.

Also the IGP resolutions/settings at Anandtech.

 

[tamz_msc]

It's not just a 5-10% delta. It's as large as 20% + at 720p as shown by the Guru3D review. 720P matters because it frees up the CPU to do it's thing and you can compare them. It matters to high refresh gamers. Also, why would anyone buy a slower CPU anyway if all they want to do is game? People claim I'm trolling because my opinion reflects the data, lol. There are games where an i5 6600K (yes, a quad core skylake i5) is just as fast as these new CPU's in some games. It's in the Guru3D review.
I appreciate their review because they remove the GPU bottleneck sufficiently so you can actually see what the delta looks like. People argue that no one plays at 720p, therefore 720p results don't matter. That's a bad argument. It's obvious the reason people don't like 720p results is because it makes the AMD chips look disappointing, but those are the results. It's a good thing data is unbiased.
Yeah the chips are good at heavy workloads, but seriously that argument is less relevant to most people than the gaming discussion is.

Why assume that gaming is somehow more relevant to most people than productivity workloads? Especially when the main point of contention with regards to gaming is 720p performance with a 2080Ti, given that any real-world CPU+GPU pairing at resolutions at which people actually play would obliterate any 20%+ FPS difference.

 

[ubern00b]

Don't think in terms of numbers. If you are gaming and you want the best, what do you get?

This is why the high end and the low end happen to be the higher volume ones. Back when Nvidia had their Geforce 4 Ti series, the Ti 4200 and 4600 both outsold the Ti 4400. Because either you want value out of the money you spent, or get the *best* that there is to offer. Same reason why you'd get the RTX 2080 Ti.

There are niche categories for gamers that need high fps. Like those that have 144Hz or higher monitors. Or VR uses. In terms of overall, they are small, but for the people buying 9900K, it should be a high portion.

Plus applications that poorly scale with cores or are under 16 threads, the 9900K will do very well.

5% (at 1080p and dminishing as you increase resolution) is barely beyond margin of error but I guess there will be some CS:GO players out there playing at 720p that need 330fps as opposed to 315fps

 

[IntelUser2000]

Especially when the main point of contention with regards to gaming is 720p performance with a 2080Ti, given that any real-world CPU+GPU pairing at resolutions at which people actually play would obliterate any 20%+ FPS difference.

You are actually missing Moonbogg's point.

The point of testing at low resolutions and settings such as 720p, is to isolate how the CPU performs in games.

Yes, for actual gaming it seems irrelevant, but hear me, it is. Because a single frame number like Average FPS does not show all the data. Benchmarks are limited. What you may feel or experience in the game can simply not be represented by one set of data.

What you are trying to do is eliminate the bottleneck in times when it might be one.

 

[GaiaHunter]

Don't think in terms of numbers. If you are gaming and you want the best, what do you get?

This is why the high end and the low end happen to be the higher volume ones. Back when Nvidia had their Geforce 4 Ti series, the Ti 4200 and 4600 both outsold the Ti 4400. Because either you want value out of the money you spent, or get the *best* that there is to offer. Same reason why you'd get the RTX 2080 Ti.

There are niche categories for gamers that need high fps. Like those that have 144Hz or higher monitors. Or VR uses. In terms of overall, they are small, but for the people buying 9900K, it should be a high portion.

Plus applications that poorly scale with cores or are under 16 threads, the 9900K will do very well.

True story - I wanted the Ti 4400 but it just went out of stock so I had to go with a Ti4200.

Why would you get a 9900K for gaming over the 9700K?

 

[guachi]

But you will not notice any difference between 150fps and 160fps so its not a selling point. And since 99% of people I assume do more than just play games hence amd wins. No reason to buy intel anymore.

I'd be better served spending the money I saved on an AMD CPU on a better video card.

And if I were wealthy enough I'd be gaming at 4k where the CPU didn't matter, anyway.

Or it would be like you describe. Even if the AMD and Intel chips were similar prices I'd never notice the difference in games (I mean, if I can afford a 9900K/3900X I'm probably not gaming at 1080) but I would in most other tasks that AMD wins in.

 

[PotatoWithEarsOnSide]

5% (at 1080p and dminishing as you increase resolution) is barely beyond margin of error but I guess there will be some CS:GO players out there playing at 720p that need 330fps as opposed to 315fps

...and they'd be better off with Ryzen anyway, since Ryzen now beats the 9900K at CS:GO.

 

[rbk123]

The point of testing at low resolutions and settings such as 720p, is to isolate how the CPU performs in games.

Nope. It isolates how the CPU performs at 720p. 1080p testing isolates the CPU at 1080p, etc. Different resolutions have different CPU demands; that's the point you and Moonbogg don't get.

 

[ubern00b]

Nope. It isolates how the CPU performs at 720p. 1080p testing isolates the CPU at 1080p, etc. Different resolutions have different CPU demands; that's the point you and Moonbogg don't get.

So basically if you're a RX 550/GT 1030 owner who games at 720p and buys a $500 processor, you'd be better off with Intel

 

[GaiaHunter]

Nope. It isolates how the CPU performs at 720p. 1080p testing isolates the CPU at 1080p, etc. Different resolutions have different CPU demands; that's the point you and Moonbogg don't get.

Yeah.

The problem is that reviewers never test the delta of the CPU at 720p when a new more powerful GPU comes out as one doesn't test GPUs at lower resolutions.

 

[chrisjames61]

It's not just a 5-10% delta. It's as large as 20% + at 720p as shown by the Guru3D review. 720P matters because it frees up the CPU to do it's thing and you can compare them. It matters to high refresh gamers. Also, why would anyone buy a slower CPU anyway if all they want to do is game? People claim I'm trolling because my opinion reflects the data, lol. There are games where an i5 6600K (yes, a quad core skylake i5) is just as fast as these new CPU's in some games. It's in the Guru3D review.
I appreciate their review because they remove the GPU bottleneck sufficiently so you can actually see what the delta looks like. People argue that no one plays at 720p, therefore 720p results don't matter. That's a bad argument. It's obvious the reason people don't like 720p results is because it makes the AMD chips look disappointing, but those are the results. It's a good thing data is unbiased.
Yeah the chips are good at heavy workloads, but seriously that argument is less relevant to most people than the gaming discussion is.

Your whole post is utter nonsense. I mean it is beyond ludicrous.

 

[Slaughterem]

Let moonbogg be happy with his comments. It is true that Intel is approximately 3% faster than AMD in 1080 games. Today he can't say that Adobe is the reason for having Intel, I use WinRAR and it is significantly better on Intel, the I9-9900K uses less power than Ryzen, Intel's multi threading is better, security is Intel's forte, Intel costs less, Sunny cove is announced for desktop. Let him hold onto his last bastion of grace, gaming performance.