ryzen 5000 real consumption?

Strange bird said:
I'm not gpu limited because I don't have a strong graphics card.
Click to expand...

You have it backward. You are GPU limited because you don't have a strong graphics card. Even in CPU intensive games I was starting to hit GPU limits with my previous 2080 at 1440p, and the 2080 is faster than the 3060 12GB. That's why I upgraded to a 4080. If I had been running 4K, I would have hit the limits of my old GPU much sooner. Running 4K on such a slow card (3060 12GB), you could upgrade to a 7800X3D or a 7950X3D and you would probably still be getting the same FPS as your 2700x.

If you want increased performance, upgrade the component that is holding you back. In your case, your GPU is most likely holding you back. The 3060 12GB is not going to do well at 4K in most newer games.
 
GotNoRice said:
You have it backward. You are GPU limited because you don't have a strong graphics card. Even in CPU intensive games I was starting to hit GPU limits with my previous 2080 at 1440p, and the 2080 is faster than the 3060 12GB. That's why I upgraded to a 4080. If I had been running 4K, I would have hit the limits of my old GPU much sooner. Running 4K on such a slow card (3060 12GB), you could upgrade to a 7800X3D or a 7950X3D and you would probably still be getting the same FPS as your 2700x.

If you want increased performance, upgrade the component that is holding you back. In your case, your GPU is most likely holding you back. The 3060 12GB is not going to do well at 4K in most newer games.
Click to expand...
Unfortunately nothing from that, I'm waiting for better times for gpu, here in my place the cheapest rtx 4070 (non ti) is 622 euros and up.
It would be nice to have more and more expensive components and gpu, but for now nothing of that kind.It's not going to happen.
 
The cheapest rtx 4080 is 1283 euros and up.
I don't know if they are normal or what? I haven't looked at gpu for a long time, who can buy it, only Kevin O'Leary.
 
Strange bird said:
The cheapest rtx 4080 is 1283 euros and up.
I don't know if they are normal or what? I haven't looked at gpu for a long time, who can buy it, only Kevin O'Leary.
Click to expand...

You might have a look at the used market. Buying an older used card can very cheap. Something like a 3080, or even a 2080 ti would still be a lot faster than the 3060.
 
GotNoRice said:
You might have a look at the used market. Buying an older used card can very cheap. Something like a 3080, or even a 2080 ti would still be a lot faster than the 3060.
Click to expand...
Nothing of that, and used ones are expensive, my upgrade would be rtx 4070ti or rtx 4080, nothing below that.
 
I guess it is, everything in my bios is mostly on auto settings.
I've already searched on the internet, but it's not really on the stock settings, I guess everything is oc 5600x.And when it is oc the results are higher, there is no one on the default settings.
I generally have both mc and sc higher than on the 2700x.

Temperatures go up to 65 and consumption up to 77w.There is a shift, but nothing terrible, and everyone said that the 2700x is very bad and the 5600x is much better.
 
Strange bird said:
I guess it is, everything in my bios is mostly on auto settings.
I've already searched on the internet, but it's not really on the stock settings, I guess everything is oc 5600x.And when it is oc the results are higher, there is no one on the default settings.
I generally have both mc and sc higher than on the 2700x.

Temperatures go up to 65 and consumption up to 77w.There is a shift, but nothing terrible, and everyone said that the 2700x is very bad and the 5600x is much better.
Click to expand...
I have had a 2700x, upgraded to a 3700x, then to a 5800x3D along with a secondary system that has been running a 5600x since around the time that they launched. Performance increases are very situational and gaming performance can be largely dependent on the graphics card that you are using, as others have mentioned earlier in the thread.

For me, there is almost no perceivable difference between the 5800x3D and 5600x while gaming on a 3060 at 2560 x 1440p resolution. There is a performance increase with a 3070 Ti, which moves the GPU bottleneck goalpost a little further.

In my experience, more RAM is generally better than faster RAM, unless you are utilizing integrated graphics. Synthetic benchmarks may indicate additional performance with faster RAM, all else being equal. Those synthetic benchmarks do not translate into a noticeable performance difference in most day to day tasks for me.

As an additional personal anecdote, my 5600x is a bit faster in Crystal Reports and Excel when working with large datasets. I'm pretty sure that the 3D vcache has no effect in Crystal Reports or Excel and that both applications are primarily single-threaded. The 5600x has an edge since it seems to maintain a higher boost frequency when using these applications.
 
When everything is right and Ryzen Master is doing some auto overclocking it will lay some nice clocks https://valid.x86.fr/020edq

updated cpu- z and retuned it as to not auto overclock, just curve optimizer with 2658Mhz memory speed now at Cl 19 https://valid.x86.fr/yg0kv8

I think review sites just drop the chip in and call it a day, you should really let Ryzen Master take a look at what is under the hood and let it tune it to your board and memory, I turned XMP off on the memory and that is default settings, with just curve optimizer apply under Peerless Assassin 120 SE.
 
Last edited:
crazycrave said:
When everything is right and Ryzen Master is doing some auto overclocking it will lay some nice clocks https://valid.x86.fr/020edq

updated cpu- z and retuned it as to not auto overclock, just curve optimizer with 2658Mhz memory speed now at Cl 19 https://valid.x86.fr/yg0kv8

I think review sites just drop the chip in and call it a day, you should really let Ryzen Master take a look at what is under the hood and let it tune it to your board and memory, I turned XMP off on the memory and that is default settings, with just curve optimizer apply under Peerless Assassin 120 SE.
Click to expand...
none of that is good. your chip is under-boosting and your ram speeds are way out of wack. 2658 at c19 is nothing to brag about either. my 3600 ram is at c18, also nothing to brag about...
 
It will do the same thing if I raise to
pendragon1 said:
none of that is good. your chip is under-boosting and your ram speeds are way out of wack. 2658 at c19 is nothing to brag about either. my 3600 ram is at c18, also nothing to brag about...
Click to expand...
That memory does 3600Mhz Cl14, go eat some cheese with your wine.
 
crazycrave said:
It will do the same thing if I raise to

That memory does 3600Mhz Cl14, go eat some cheese with your wine.
Click to expand...
what "same thing"?
then why are you gimping it down to such a low random number?
youre giving bad advice, except for the pa120se part, those are awesome.
 
pendragon1 said:
it never seemed to do anything for me. my bios boost it to 4850 just fine too.
Click to expand...
I used it a couple of times when I was brand new to AM4. I quit using it when it would change my bios settings to its own settings lol. I lost my bios fear when I realized that all changes RM makes can be done without software :D

I mean I am sure it has its place, just not at my place.
 
MrCaffeineX said:
I have had a 2700x, upgraded to a 3700x, then to a 5800x3D along with a secondary system that has been running a 5600x since around the time that they launched. Performance increases are very situational and gaming performance can be largely dependent on the graphics card that you are using, as others have mentioned earlier in the thread.

For me, there is almost no perceivable difference between the 5800x3D and 5600x while gaming on a 3060 at 2560 x 1440p resolution. There is a performance increase with a 3070 Ti, which moves the GPU bottleneck goalpost a little further.

In my experience, more RAM is generally better than faster RAM, unless you are utilizing integrated graphics. Synthetic benchmarks may indicate additional performance with faster RAM, all else being equal. Those synthetic benchmarks do not translate into a noticeable performance difference in most day to day tasks for me.

As an additional personal anecdote, my 5600x is a bit faster in Crystal Reports and Excel when working with large datasets. I'm pretty sure that the 3D vcache has no effect in Crystal Reports or Excel and that both applications are primarily single-threaded. The 5600x has an edge since it seems to maintain a higher boost frequency when using these applications.
Click to expand...
Thanks for the reminder, I downloaded ryzen master and selected per core optimization in just two clicks. Now I will test with cinebench r23 to see what I got.
I don't have the time or knowledge to dig through bios, etc.
 
crazycrave said:
When everything is right and Ryzen Master is doing some auto overclocking it will lay some nice clocks https://valid.x86.fr/020edq

updated cpu- z and retuned it as to not auto overclock, just curve optimizer with 2658Mhz memory speed now at Cl 19 https://valid.x86.fr/yg0kv8

I think review sites just drop the chip in and call it a day, you should really let Ryzen Master take a look at what is under the hood and let it tune it to your board and memory, I turned XMP off on the memory and that is default settings, with just curve optimizer apply under Peerless Assassin 120 SE.
Click to expand...
Thanks for the reminder, I downloaded ryzen master and selected per core optimization in just two clicks. Now I will test with cinebench r23 to see what I got.
I don't have the time or knowledge to dig through bios, etc.
 
Strange bird said:
Suddenly 11919 Multi core, that's fantastic...
Consumption and heating remained similar to before

My patriot memory now with ryzen 5600x works at 3200MHz with latencies 24-23-23-53 tRC 76 tRFC 560
When I bought it, it was quite expensive, now I see that the price seems to have dropped.
It is about this memory:
https://www.bhphotovideo.com/c/prod...t_pvs416g320c6k_viper_steel_16gb_2.html/specs
Click to expand...
Why are the timings so loose? That RAM is rated for CAS 16 and you’re running it at 24?
 
gtrguy said:
Why are the timings so loose? That RAM is rated for CAS 16 and you’re running it at 24?
Click to expand...
because I set the memory to 3200mhz and auto latency and that's how bios set it... I don't know what I should put, that memory has an xmp profile but that's for intel processors
 
Strange bird said:
because I set the memory to 3200mhz and auto latency and that's how bios set it... I don't know what I should put, that memory has an xmp profile but that's for intel processors
Click to expand...
IF just setting xmp doesnt work on its own, open the bios, set it to xmp but dont save/reboot, copy the values, then turn off xmp and manually set it to those values.
 
Strange bird said:
I don't know what I should put, that memory has an xmp profile but that's for intel processors
Click to expand...

Why do you think your motherboard (B450M Steel series correct?) BIOS offers the option to select an XMP profile if it isn't compatible? While it is not guaranteed to work, it is possible to select XMP and test if it works.


steel series bios.jpg
 
Thank you, I put xmp in the bios and now it works on 16-20-20-40, it was pushed to windows.
I never used that xmp because it always caused me problems on the 2700x processor, it never worked.That's why I always used the latencies that the bios automatically set for me.
 
Strange bird said:
Thank you, I put xmp in the bios and now it works on 16-20-20-40, it was pushed to windows.
I never used that xmp because it always caused me problems on the 2700x processor, it never worked.That's why I always used the latencies that the bios automatically set for me.
Click to expand...
You had problems with the XMP profile with a 2700x because as many people told you in your previous threads that Zen+ was a crapshoot when trying to run anything above 2933 in most cases and it was even worse if you were trying to run more than 16GB. The memory controllers on them were never the best.

My son and I have 2600x's which do run 3200 (16GB) with xmp profiles but the CPUs were manufactured well after Zen 2 had already been out and so they tended to be much better silicon.

If there's a 5600x out there which won't run the RAM at 3200 with decent timings, it needs to be RMA'd. Unless you're using a large amount of RAM such as 64GB+ any 5600x should run just fine with 3600 RAM kits. This is NOT a recommendation that you try to run your RAM at 3600. It's a 3200 kit and you should leave it at that.
 
SmokeRngs said:
You had problems with the XMP profile with a 2700x because as many people told you in your previous threads that Zen+ was a crapshoot when trying to run anything above 2933 in most cases and it was even worse if you were trying to run more than 16GB. The memory controllers on them were never the best.

My son and I have 2600x's which do run 3200 (16GB) with xmp profiles but the CPUs were manufactured well after Zen 2 had already been out and so they tended to be much better silicon.

If there's a 5600x out there which won't run the RAM at 3200 with decent timings, it needs to be RMA'd. Unless you're using a large amount of RAM such as 64GB+ any 5600x should run just fine with 3600 RAM kits. This is NOT a recommendation that you try to run your RAM at 3600. It's a 3200 kit and you should leave it at that.
Click to expand...
Maybe i could lower the latencies or overclock, in what way and by how much? Maybe that ryzen ram calculator? I have 32gb of ram 4x8gb but they are not exactly the same, 2x8gb is hynix and the other 2x8gb is micron, there are some differences and I could not find the same patriot memory with hynix chips anywhere, they are no longer produced.
 
I changed a machine from 9700k to 5950x. Idle power went up by 22 watt (with all other hardware ported over in identical form).

Do I miss some power management software somewhere?
 
uOpt said:
I changed a machine from 9700k to 5950x. Idle power went up by 22 watt (with all other hardware ported over in identical form).

Do I miss some power management software somewhere?
Click to expand...
It is due to the constantly active I/O die of the Zen 3 (and Zen 2) CPUs. And that separate I/O die within the CPU package, which is manufactured on an older 12 nm process node, is drawing a lot of power (relatively speaking) from the PSU.

So, all Ryzen CPUs from Zen2 onwards that are not APUs use a multi-chiplet design with differing process nodes.

As for the Ryzen 5 5500 that the OP mentioned earlier in this thread, it is an APU with a graphics portion so defective that it is disabled at manufacturing level. And all Ryzen APUs all the way up to and including Cezanne (on the desktop side) are monolithic (as opposed to chiplet) designs.
 
  • Like
Reactions: uOpt
like this
pendragon1 said:
set it to what its supposed to be and leave it alone.
Click to expand...
yes daddy:D
There is a video where they used that 2x8GB memory for lower latency with 3200MHz and also achieved 3600MHz.There is no evidence, only the latencies and speeds were shown in the video, but not from a program,written on the video screen.
 
Is there a possibility that amd ryzen master automatically turns on per clock optimization? Or do I have to start the program every time I turn on Widnows and click on apply?
 
You must log in or register to reply here.
Back
Top