Ryzen 5800X3D stock slowly getting depleted?

[FoulOnWhite]

Maybe.

I switch back and forth all the time.

I probably use mine 20% for gaming, i have other things to do, don't really have time to game all day. used to be more like 6hrs a day 20 years ago though. now its youtube, browse, games, and mostly household stuff

Imo get something thats good for all, even if it's not mega great for gaming.

 

[Sunny and 75]

I just put in an order for a 5700X3D

Congratz on the upgrade.

May it serve you well.

 

[RJARRRPCGP]

Makes me glad that I did major upgrades by January, 2024! IIRC, I got my 5800X3D after December. I got my 5900X on Christmas, 2023.

 

[Princess Garnet]

I get what he is saying I just don't think it applies to most me and him maybe though.

I hated the 7800X3D and most would love it.
I like the 7950X3D much better even with it's headaches.

But I'm definitely the minority.

Okay, but you're talking about something completely different then. You're basically saying "I would personally value more cores" which is neither here nor there because the statement I was responding to specifically isolated core count away from the comparison by stating "they just get slaughtered due to low clocks". So I was therefore only talking about that difference.

I dunno..

But there is a big difference between 4450MHz of the X3D, and 5050MHz for the X. You wouldn't notice more than half a GHz? To me there is more to computers than just playing games.

Where does a 4.5 GHz and 5 GHz comparison come from?

AMD lists the boost frequency of the 5800X as 4.7 GHz (which is 200 MHz, not 500 MHz, higher than the X3D), so it can't be the 5800X and 5800X3D.

All of the Ryzen 7000 series X3Ds boost to 5 GHz+, so it can't be them either.

What exact CPUs were you referencing? If you are comparing the 5900X or 5950X (neither of which boost to 5.05 GHz) compared to a 5800X3D, then suddenly you're not talking about a massive difference "due to low clocks" but due to "50% to 100% more cores" on top of the marginally faster cores.

And these are "up to" frequencies that only low core count loads will hold. What are the typical frequencies of all core loads? Does the gap (relatively, not absolutely) stay the same? Shrink? Grow?

It's easy to make something sound large when you phrase it "over half a GHz" but I fondly remember overclocking a 2500K from 3.3GHz to 4 GHz and beyond (same with an E8600), and the differences were not what I'd describe as so far apart that the slower was bad for the task, especially just going on perception. I really couldn't tell them apart at all for most comparisons. "Measurable, but mostly imperceptible" came to mind. And that was a bigger difference than the spread is now.

 

[oxrufiioxo]

Okay, but you're talking about something completely different then. You're basically saying "I would personally value more cores" which is neither here nor there because the statement I was responding to specifically isolated core count away from the comparison by stating "they just get slaughtered due to low clocks". So I was therefore only talking about that difference.

Def couldn't tell apart a 5800X3D/5800X or a 7800X3D/7700X unless gaming but the non X3D chips can be tweaked quite a bit if you don't care about power. I've installed 5700X3D ran benchmarks they did seem about 10% slower than my 5800X outside of gaming... Not sure if someone would notice that or not the 20% increase in gaming performance in some games is more noticeable.

Also the OP bought the 5700X3D it's the chip he's talking about boost clocks are 4.1ghz vs 4.4ghz on the 5800X3D..... and then another 300Mhz on the 5800X which is cheaper. so not a bad comparison.

 

[freeagent]

Where does a 4.5 GHz and 5 GHz comparison come from?

PBO.

5800X can do 5050, 5900X can do 5150, 5950X can do 5250.

Just like X3D, enable PBO.. though all it does it activates the curve.

 

[oxrufiioxo]

PBO.

5800X can do 5050, 5900X can do 5150, 5950X can do 5250.

Just like X3D, enable PBO.. though all it does it activates the curve.

The 5700X3D I worked with all hit 4ghz or lower all core though, the boost clocks on ryzen are pretty irrelevant the 5800X 4.5-4.6hz unless overclocked. So still about half a ghz holds true I guess.

It's been a minute but the 5800X3D isn't that far back maybe 2-300mhz all core though stock vs stock ofc.

Given identical core counts I would still want the X3D chip though.

 

[freeagent]

The 5700X3D I worked with all hit 4ghz or lower all core though, the boost clocks on ryzen are pretty irrelevant the 5800X 4.5-4.6hz unless overclocked. So still about half a ghz holds true I guess.

Given identical core counts I would still want the X3D chip though.

My X3D "feels" ok when just clicking around, and in game it does pretty good when running at under 4K. It uses hardly any power, and if I were running a heatsink, I would be able to run it with no fans installed on the cooler. Really good for games. Its an ok CPU in general. Its better than my tuned 5900X in a handful of instances.

 

[oxrufiioxo]

My X3D "feels" ok when just clicking around, and in game it does pretty good when running at under 4K. It uses hardly any power, and if I were running a heatsink, I would be able to run it with no fans installed on the cooler. Really good for games. Its an ok CPU in general. Its better than my tuned 5900X in a handful of instances.

Something I found odd is 5000X3D with low latency 3800 CL14 memory felt snappier in normal desktop usage than 7800X3D with 6000 CL30 memory..... The first time I noticed it I thought maybe it was a one off but 3 for 3 all fresh windows installs at least stock vs stock....

That isn't the case with my 5950X vs 7950X3D so not sure what's up with that I wish I had a dual CCD 5000X3D to compare lol but na AMD want's to milk them no vcache chips smh.

 

[freeagent]

Something I found odd is 5000X3D with low latency 3800 CL14 memory felt snappier in normal desktop usage than 7800X3D with 6000 CL30 memory..... The first time I noticed it I thought maybe it was a one off but 3 for 3 all fresh windows installs at least stock vs stock....

That isn't the case with my 5950X vs 7950X3D so not sure what's up with that I wish I had a dual CCD 5000X3D to compare lol but na AMD want's to milk them no vcache chips smh.

I saw a lot of people saying X3D did not respond to memory. Mine does. I would say even more so than my 5900X, as the ccd latency dumbs down most latency improvements that are gained. Most.. but not all of course.

I was going to get me some Zen 4, but its a little late in the game now, especially with new chipsets and boards coming out. It was really late in the game when you guys helped me build my first AM4 rig, but compared to what was out at the time, it has aged well..

 

[oxrufiioxo]

I saw a lot of people saying X3D did not respond to memory. Mine does. I would say even more so than my 5900X, as the ccd latency dumbs down most latency improvements that are gained. Most.. but not all of course.

I was going to get me some Zen 4, but its a little late in the game now, especially with new chipsets and boards coming out. It was really late in the game when you guys helped me build my first AM4 rig, but compared to what was out at the time, it has aged well..

The fact that technically someone who grabbed a 1600X/2600X to start can now get a 5700/5800X3D chip is wild. Going from something that generally lost to a 7600K in gaming to something that trades blows with a stock 12900K is crazy..... I get part of that is the 1000/2000 chips weren't super impressive but still you'd need to change boards multiple times in that span with intel. It's had an awesome run no lets see how AM5 ages.

I got A die on my DDR5 kit haven't had time to really tweak timings but they matter more on 7000X3D than I thought just a slight tweak made a huge difference in consistency.

Kinda a bummer that all the DDR4 left is pretty crap and the good kits are an arm and a leg now glad I got 8 good sticks of Bdie lol....

 

[freeagent]

The fact that technically someone who grabbed a 1600X/2600X to start can now get a 5700/5800X3D chip is wild. Going from something that generally lost to a 7600K in gaming to something that trades blows with a stock 12900K is crazy..... I get part of that is the 1000/2000 chips weren't super impressive but still you'd need to change boards multiple times in that span with intel. It's had an awesome run no lets see how AM5 ages.

Even Zen 2 to Zen 3 is a huge kick in the pants when you tune PBO. Even my mind was blown lol.. thats why I went hard on Zen 3 lol.. I still have my 5600X too chugging away at 4800MHz in my sons rig. With a single fan PA120SE ARGB too lol..

I got A die on my DDR5 kit

Can you still buy it? If so can you link me to them or something similar? I thought I heard something awhile back about the good stuff being harder to find, maybe I am mistaken..

Kinda a bummer that all the DDR4 left is pretty crap and the good kits are an arm and a leg now glad I got 8 good sticks of Bdie lol....

I have 6 good sticks too, very thankful to have them as they are super flexible.

 

[oxrufiioxo]

Can you still buy it? If so can you link me to them or something similar? I thought I heard something awhile back about the good stuff being harder to find, maybe I am mistaken..

It's a lottery you either Get A or M die.. Both are good at different things.

At least with 6000 CL30

If you go up to like an 8000 CL34 kit pretty sure that's A die

You might want to PM @ir_cow he might know ways to get a specific type based on timings/Frequency.

 

[kapone32]

The fact that technically someone who grabbed a 1600X/2600X to start can now get a 5700/5800X3D chip is wild. Going from something that generally lost to a 7600K in gaming to something that trades blows with a stock 12900K is crazy..... I get part of that is the 1000/2000 chips weren't super impressive but still you'd need to change boards multiple times in that span with intel. It's had an awesome run no lets see how AM5 ages.

I got A die on my DDR5 kit haven't had time to really tweak timings but they matter more on 7000X3D than I thought just a slight tweak made a huge difference in consistency.

Kinda a bummer that all the DDR4 left is pretty crap and the good kits are an arm and a leg now glad I got 8 good sticks of Bdie lol....

To be honest for me RAM quality does not matter for X3D chips. I was running some 18 timed 3600 32GB kits before I got those Team Extreme 3600 15-15-15-16 32GB kit. It literally made no difference with the 5800X3D. It did make the 5950X snappier though.

 

[oxrufiioxo]

To be honest for me RAM quality does not matter for X3D chips. I was running some 18 timed 3600 32GB kits before I got those Team Extreme 3600 15-15-15-16 32GB kit. It literally made no difference with the 5800X3D. It did make the 5950X snappier though.

Cool! 3600 CL15 isn't very fast maybe 4000 CL15 I'd consider an ok bin, in my experiences ryzen doesn't like odd primary timings but I've been running 3800 14-14-14-24 so long honestly can't remember. 3600 CL14 kits were pretty cheap for a while ryzen owners should have jumped on them when they were 115-120 bucks.

Most my testing was with a 2080ti/3080ti haven't really done any with the 4090 but it's noticeable down to a 4070.

 

[Princess Garnet]

PBO.

5800X can do 5050, 5900X can do 5150, 5950X can do 5250.

Just like X3D, enable PBO.. though all it does it activates the curve.

Looking into it, I am finding some anecdotes of the ~5 GHz value with PBO for the 5800X, but I'm also finding more anecdotes of less. Things can start getting vague real quick with stuff like that. I was presuming more of a typical scenario and was using the listed, and thus more expected to be typical, boost frequencies.

While using PBO on my prior 3700X, I didn't observe it boosting higher than without, but it was using twice the peak power draw. That was on an ROG Strix B550-F and Dark Rock Pro 4 so I find it hard to imagine either the board or cooling were limiting it, but who knows. The 3700X isn't the 5x00X too, so there's that, but perhaps it explains why I wasn't putting much consideration into what something "might do" or "up to".

Peak, low core count boosts are one thing. Typical spreads are another. If something is recording a given peak frequency in a fringe and brief moment but not actually being separated by that same spread on average, then that's making exaggerated comparisons in my mind. And even if they are 500 MHz apart on average, I suppose we're digressing because the ultimate point is that the performance data exists as it does and shows the typical, expected performance spread. And these differences for tasks that don't benefit from cache tend to fall within "certainly measurable, but typically imperceptible" (while being more than made up for by larger gains the other direction in the tasks that do benefit from cache). So my overall point is that calling "X minus 5% to 10%" as "getting slaughtered" seemed being a bit dishonest in my eyes.

 

[freeagent]

but I'm also finding more anecdotes of less.

I think it comes down to tuning, and the tuners ability personally. Also, Zen 2 behaves much differently than Zen 3 does in just about every aspect. Thinking back to my 3600XT is all..

I ran this for you yesterday before I dropped my X3D back in. Just wanted to show what the CPU (5900X) can actually do.

 

[azrael]

Well, I got my (boxed) Ryzen 5700X3D from Amazon yesterday. Was dreading the price would be further reduced on Prime Day, but apparently it's gotten a few Euros more expensive.

I recently found out that you can buy processors, including the 5700X3D, on AliExpress. Could've saved about 40€, but tbh I don't really trust AliExpress for stuff like that. Coincidentally, 5800X3Ds are also very hard to find on the site and those that are there are even more overpriced than what I'd have to pay in Denmark.

All in all, I'm a happy camper. Now I just need to install the new hardware. Is Arctic MX4 still good? I have some around from my last (aka. current) build about 4 years ago.

Finally,, I would like to once again thank you all for your input! Very much appreciated.

 

[68Olds]

All in all, I'm a happy camper. Now I just need to install the new hardware. Is Arctic MX4 still good? I have some around from my last (aka. current) build about 4 years ago.

If I had some MX4 on hand, I would be more than happy to use it. Congrats on the new hardware!

 

[Chomiq]

Meanwhile:

Looks like 5800X3D at €300 was just an anomaly.

 

[azrael]

Meanwhile:
View attachment 355348
Looks like 5800X3D at €300 was just an anomaly.

It's certainly gotten even more expensive since I created the thread. Not sure about availability. Might be the same.

 

[oxrufiioxo]

Well, I got my (boxed) Ryzen 5700X3D from Amazon yesterday. Was dreading the price would be further reduced on Prime Day, but apparently it's gotten a few Euros more expensive.

I recently found out that you can buy processors, including the 5700X3D, on AliExpress. Could've saved about 40€, but tbh I don't really trust AliExpress for stuff like that. Coincidentally, 5800X3Ds are also very hard to find on the site and those that are there are even more overpriced than what I'd have to pay in Denmark.

All in all, I'm a happy camper. Now I just need to install the new hardware. Is Arctic MX4 still good? I have some around from my last (aka. current) build about 4 years ago.

Finally,, I would like to once again thank you all for your input! Very much appreciated.

You're probably fine but I typically throw out paste that is over 2 years old, not going through unmounting the cooler and cleaning off thermalpaste to save 5-10 usd lol if there is an issue with it.

Congrats on the new cpu.

 

[freeagent]

I just went on Newegg to see if the new XT CPU's might be out, nope.. but I did not see 58X3D listed.. crazy.

 

[oxrufiioxo]

I just went on Newegg to see if the new XT CPU's might be out, nope.. but I did not see 58X3D listed.. crazy.

I did notice that the 7950X3D is back down to what I paid for it $476 actually slightly cheaper with a promo code 464.... But yeah 5800X3D sold out it seems.

 

[68Olds]

I just went on Newegg to see if the new XT CPU's might be out, nope.. but I did not see 58X3D listed.. crazy.

I did notice that the 7950X3D is back down to what I paid for it $476 actually slightly cheaper with a promo code 464.... But yeah 5800X3D sold out it seems.

The 5800X3D is still available at both my local Microcenter locations, 10 at one store & 21 at the other. Definitely dwindling in numbers.