• Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
  • We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.

M.Bluth

Member
Oct 25, 2017
4,257
i was expecting the 5800X to be doing a lot better than that based on the reactions here lol.

it's good yeah but it's only just better than the 10900K. not really blowing it out the water. plus it's only at 1080p...not many people are buying these kind of CPUs for 1080p. at 1440/2160p there will be little difference.
Unlike the 10900K, the 5800X is far from being the top of the line in AMD's product stack. It's also probably $100 cheaper.
If the Zen 2 CPUs are any indication, 5900 and 5950 SKUs should boost higher while having more cores.

The 5xxx series is the last generation before the socket change, yes? So if I would like to maybe keep my motherboard across CPU upgrades I should wait for the next release?
Yes. Though I believe Zen 4 is gonna require DDR5, so it won't be cheap going in on the first gen of that.
 

Tora

The Enlightened Wise Ones
Member
Jun 17, 2018
8,640
i was expecting the 5800X to be doing a lot better than that based on the reactions here lol.

it's good yeah but it's only just better than the 10900K. not really blowing it out the water. plus it's only at 1080p...not many people are buying these kind of CPUs for 1080p. at 1440/2160p there will be little difference.
So that benchmark was at 4k

Look at the CPU framerate
-134 (5800x)
-115 (10900k)

That's quite a big difference
 

Darktalon

Member
Oct 27, 2017
3,266
Kansas
i was expecting the 5800X to be doing a lot better than that based on the reactions here lol.

it's good yeah but it's only just better than the 10900K. not really blowing it out the water. plus it's only at 1080p...not many people are buying these kind of CPUs for 1080p. at 1440/2160p there will be little difference.
Um what? First off, its a test at 4k. 2nd off, CPU's are typically single digit upgrades. Here we are seeing a 16.5% jump over literally the current fastest gaming CPU on the planet. You have some severely unrealistic expectations if that doesn't get the blood flowing. Not to mention the 10900k is a $500 product. The 5800X will be likely $350-$400.
 
Last edited:

Paganmoon

Member
Oct 26, 2017
5,586
AMD naming is starting to overlapp CPU and GPU, some grandma is going too buy a 5800X thinking it's a 5800XT
 

luminosity

Member
Oct 30, 2017
957
Watching the CPU race from the sidelines, but is the scuttlebutt that this is the gen when AMD finally surpasses Intel in all areas?
 

Darktalon

Member
Oct 27, 2017
3,266
Kansas
How come the cpu framerate is higher but the actual framerate is lower?
Because the RTX 2080 can't push more than 60fps at 4k in these tests... the difference between 59.3 and 60.3 is literally just a test variance.
Its literally a GPU bottleneck. But much like Horizon Zero Dawn's benchmark, it will display what the FPS your CPU would have let you had if you had a fast enough GPU.

To break this down farther, its literally counting how long it takes the GPU and the CPU to finish their workloads. 60 fps is 16.67ms. 134 fps is 7.46ms.
The cpu is finishing its work in 7.46ms and the gpu is taking 16.67ms. Whichever is the slower part is the FPS you will have displayed.

If you were using a 3090 OC'd, you would have the 10900k stuck displaying at 114 fps, and higher on the 5800X.
 

dgrdsv

Member
Oct 25, 2017
11,885
What benchmark is this and how much is it affected by RAM? Because the Ryzen bench was done with 32GB while the Intel only got 16GB.
Size of RAM has no impact here above 16GB. Speed of RAM can have an impact though. AotS is also one of those benchmarks which take advantage of many CPU cores somewhat well.
 

Darktalon

Member
Oct 27, 2017
3,266
Kansas
Looks great, but without knowing what clockspeeds and ram configurations both CPUs were using...
It's a bit pointless tbh.

My 3900x is for example reaching 58 FPS in the benchmark. The benchmark tops out at around 16 threads.
But it is a good deal slower than an 10900k in most games... Especially one with fast ram and running at 5GHz.

ashescrazy4knujk2.jpg
And how is the benchmark disagreeing at all? 115 fps (10900k) vs 58 fps (your 3900x) is 1.98x faster.

From here on, people, please stop confusing GPU FPS with CPU FPS in this benchmark.
 

Darktalon

Member
Oct 27, 2017
3,266
Kansas
For my defense:

I saw this one in my tweeter feed, and assumed it's the same.
And 63fps are not that much different to what I posted.

GPU Benchmark - Ashes of the Singularity

Video Card GPU Benchmarks, High quality testing of your video card and gaming system.

ejdffu2u4aaujjnlpkrn.png
Honestly I think there is something wrong with your 3900x benchmark, almost every single benchmark for 3900x & 2080ti has a higher score than you. Run all 3 batches.
 

Tora

The Enlightened Wise Ones
Member
Jun 17, 2018
8,640
What is the 5800x's score compared to a 3700x?

How much of a % difference, curious to see how much of a jump that is
 
Last edited:

Thraktor

Member
Oct 25, 2017
570
The b

The bottleneck would be the bandwidth of the ram, they would need something like a PCI E card with gddr ram on it for it to be really viable.

Then the bottleneck would be the PCIe interface. The real solution for this would be to use HBM, like the weird Intel CPUs with AMD integrated graphics released a few years ago. The issue is cost, though. APUs are aimed at the cheaper end of the market, and adding HBM to an APU would result in something far too expensive; you'd be better off just buying a cheap CPU and an entry-level discrete GPU. There's potentially a bit of a market for it in laptops, like those Intel CPUs, but from AMD's perspective it's probably not worth putting in a lot of R&D resources to create an APU with such a limited market.

If AMD do move to chiplets for their GPUs at any point then it may become an option; rather than developing an entire APU, they could take a CPU chiplet, a GPU chiplet, and they'd just have to create a (relatively cheap) I/O die which ties together the CPU, GPU and HBM. Still, I wouldn't hold out hope any time soon, and even if it did happen, it could be limited to the laptop market much like the Intel chips I mentioned.
 

Isee

Avenger
Oct 25, 2017
6,235
Honestly I think there is something wrong with your 3900x benchmark, almost every single benchmark for 3900x & 2080ti has a higher score than you. Run all 3 batches.

Now I understand what is going on here. There are two Ashes Benchmarks: One is GPU focused and one is CPU focused:
We obviously need to compare the correct ones with each other.


This is my 3900x vs the 5800x in the CPU focused benchmark.

3900x = 58 FPS
5800x = 63 FPS (+9%)

cpufocusedcomparision1ukb9.jpg



-----------------------------------------------------------

And this is the GPU focused Benchmark. If we concentrate on the CPU framerate here:

3900x = 109 FPS (base)
10900k = 115 FPS (+6%)
5800x = 134 FPS (+23%)

Now, my machine is obviously running a more powerful GPU, but that's reflected in the average Framerate graphs.
Has it an effect on the CPU graph. Maybe, though that's the best comparision I can provide.

gpufocusedbenchmarklpkcs.jpg


So, the difference plus is somewhere around 9 - 23%.
Still, the question is how fast is the 10900k running, what are the memory configurations for both CPUs etc.
Those things are important when comparing CPUs imo.

And let's not forget Ashes of the Benchmarks likes Ryzen CPUs.


edit:

This is a scene from the CPU focused benchmark, and this is a scene from the GPU focused benchmark.
The game doesn't seem to go above 16 Threads by a lot. Comparing the 5800x against the 10700k would make more sense. But it's not a big deal.
 
Last edited:

Sabin

Member
Oct 25, 2017
4,623
Fully expect Zen3 to also be in very short supply this year, yeah.


3080 stock situation will likely improve considerably in the second half of October once sea freight will arrive to key markets.

TSMC 7nm yields are fantastic tho.

Doubt it will be even remotely close to a 3080 situation and CPUs are always easier to get day 1 compared to GPU. It's actualy not even close.
 

GalvoAg

Member
Oct 30, 2017
2,385
Dallas
TSMC 7nm yields are fantastic tho.

Doubt it will be even remotely close to a 3080 situation and CPUs are always easier to get day 1 compared to GPU. It's actualy not even close.
Hopefully you're right, I can live without my GPU for a while and get by on my 1070ti. I just got a bunch of parts waiting for Zen 3 so hopefully I can get my hands on one of their 8cores so I can start building the new rig.
 

Vuze

Member
Oct 25, 2017
4,186
Looks very promising. Can't wait, hope I can snag one asap. It's the last missing part of my new build (well, the 980 Pro is still MIA, but it's scheduled for the 15th and I doubt Zen 3 will arrive any sooner anyway)
 

eddy

Member
Oct 25, 2017
4,741
If AMD trots out "Total War" benchmarks in their presentation, you know it's game over.
 

rashbeep

Member
Oct 27, 2017
9,464
sounds promising, if amd prices the ryzen 7 right it's gonna be an absolute pain in the ass to get. i have my doubts they will be able to meet the demand.

don't really mind waiting until they are readily available anyways
 

Ra

Rap Genius
Moderator
Oct 27, 2017
12,207
Dark Space
My understanding from this very forum is Turing is basically mid-tier and unlikely to keep up with next gen consoles.

The sole reason I'm looking to upgrade.
No one has called the 2080 Super "mid-tier", come on you're being unrealistic.

You very much have a card that could allow you to skip Ampere to wait for RTX 40/RDNA3.
 

piratepwnsninja

Lead Game Designer
Verified
Oct 25, 2017
3,811
Also a lot of people recommend Asus TUF X570 and it is a great board with exceptional VRM for the price - unless you have a modern size GPU. The chipset fan is completely blocked by GPU and in my case it means the chipset idle temperatures are at around 75C with the fan running at around 3000RPM and during load it's at around 80C with 3500RPM. Can't even adjust the chipset fan speed with Asus software or via bios. In case someone wants to adjust the fan on this board, try Libre Hardware Monitor. Temperatures don't really seem to change whether the fan is at 2000RPM or 3500RPM, in other words, the fan is shit (but unfortunately necessary, fan at 0RPM made temperatures rise quickly).

So the X570 TUF is a great fairly priced board with one critical flaw that in practice makes it impossible to recommend to almost anyone.

Is the problem noise, then? I've had a TUF sitting here for a few months for my build. Noise really won't be a problem.
 

Phinor

Member
Oct 27, 2017
1,241
Is the problem noise, then? I've had a TUF sitting here for a few months for my build. Noise really won't be a problem.

Well noise seems to always be pretty subjective, some people think their quiet setup is quiet but someone else might consider the same setup noticeably loud. That being said, the chipset fan noise isn't really an issue for me and I like to think my setup is pretty quiet, especially when idle. I'm more concerned about longevity of the chipset and the fan. The fan is pretty much always running at 3000+ RPM and I'm not sure it will last long running like that (bad experiences with chipset fans, nForce 2 etc.) Also chipset being at 70C+ at all times, again, not sure how this particular chipset enjoys heat but I'm not used to chipsets running that hot when idle. Also I wonder if thermal throttling is a thing with these chipsets and what the thresholds are.

Also the chipset fan probably has a slight effect on my GPU cooling (particularly the third fan that is directly on top of the chipset fan). If I drop the chipset fan down to 2000RPM, my GPU seems to run ever so slightly colder compared to chipset at 3500RPM under load.
 

piratepwnsninja

Lead Game Designer
Verified
Oct 25, 2017
3,811
Well noise seems to always be pretty subjective, some people think their quiet setup is quiet but someone else might consider the same setup noticeably loud. That being said, the chipset fan noise isn't really an issue for me and I like to think my setup is pretty quiet, especially when idle. I'm more concerned about longevity of the chipset and the fan. The fan is pretty much always running at 3000+ RPM and I'm not sure it will last long running like that (bad experiences with chipset fans, nForce 2 etc.) Also chipset being at 70C+ at all times, again, not sure how this particular chipset enjoys heat but I'm not used to chipsets running that hot when idle. Also I wonder if thermal throttling is a thing with these chipsets and what the thresholds are.

Also the chipset fan probably has a slight effect on my GPU cooling (particularly the third fan that is directly on top of the chipset fan). If I drop the chipset fan down to 2000RPM, my GPU seems to run ever so slightly colder compared to chipset at 3500RPM under load.

Ah. Well, guess I'll see if Amazon will let me return this since it's unopened and maybe get the MSI Meg or something.
 

bigJayy

Member
Dec 30, 2019
87
The chipset fan is completely blocked by GPU and in my case it means the chipset idle temperatures are at around 75C with the fan running at around 3000RPM and during load it's at around 80C with 3500RPM.

What case do you have and how many fans are installed?
I ask because every review for the ASUS TUF Gaming X570-Plus i looked up showed great Temperatures.
 

dgrdsv

Member
Oct 25, 2017
11,885
TSMC 7nm yields are fantastic tho.

Doubt it will be even remotely close to a 3080 situation and CPUs are always easier to get day 1 compared to GPU. It's actualy not even close.
Yields have nothing to do with it really. Bad chip yields reflect in prices and die sizes, not market availability.
The problem is that they will prioritize DC products (Epycs) over consumer ones and there won't be a lot of them at start anyway (because TSMC is fully loaded at the moment and not with AMD chips only).
3080's issues isn't yields, its demand at its price. And Ryzen 5K will be hit with the same insane demand I'm afraid which will make its stock very elusive. Likely less so than that of 3080 at the moment but still.