Review AMD Ryzen 7 7800X3D Review: New Gaming Champ Beats Pricier CPUs

[TerryLaze]

Yeah, that's a fact. The making up your own fact is your explanation as to why it happens.

You are the one making up things, AM5 as a platform has the Volt levels it has, they are not too high since no CPU has ever blown up on the AM5 platform before.

This indictment needs to be weighted by how common it really is. So, please tell us roughly how many X3D CPUs have failed due to this problem.

A funny thing about human risk perception is that it's skewed by the severity of the outcome. If 0.01% of CPUs fail in a way that sounds violent or dramatic, people are much more fearful of it than if there were the same probability of failure by simply refusing to boot.

It's a higher percentage of other CPUs which is why people even bother talking about it, but more importantly why AMD bothered to do something about it, if they would be failing at the same rate as other CPUs then we would never have heared about it.

 

[SunMaster]

You are the one making up things, AM5 as a platform has the Volt levels it has, they are not too high since no CPU has ever blown up on the AM5 platform before.

That's your own, flawed, conclusion. Believing it doesn't make it true. How about backing up your claimed cause with some facts?

There is plenty information around from youtubers such as derbauer or gamers nexus which points to too high SoC voltage. Also from ASUS and AMD. But I guess they all lie, and have no idea what they're talking about, since you know differently.

 

[bit_user]

It's a higher percentage of other CPUs which is why people even bother talking about it, but more importantly why AMD bothered to do something about it, if they would be failing at the same rate as other CPUs then we would never have heared about it.

That's beside the point. My question was: roughly how often does it happen?

The reason that matters is if it's like a 1 in 10k risk or less, then you guys are really fear-mongering. And if you don't even know how big a risk it is, then that it's irresponsible to be talking like it's commonplace.

 

[Ogotai]

ahh, there is terry making up " facts " in any way possible to make amd look bad, at all costs.

terry, post a link to these " facts " , if you can't, then it is just your own personal opinion, nothing more, which is your agenda, make amd look bad.

this is no different then when intel " turns a blind eye " when a mobo maker does any thing it can, to allow intels cpus do and use what ever it needs to top the bench mark charts, i sure dont recall you saying anything about that, or complaining about it. its one of the reasons why some wont touch intel for the last 2 or 3 gens, as intel sucks power from the wall, (regardless which cherry picked graph or test you use.) and one reason why no one i know is looking at intel right now, some have picked up non x3d cpus and are running just fine .

 

[TerryLaze]

That's your own, flawed, conclusion. Believing it doesn't make it true. How about backing up your claimed cause with some facts?

There is plenty information around from youtubers such as derbauer or gamers nexus which points to too high SoC voltage. Also from ASUS and AMD. But I guess they all lie, and have no idea what they're talking about, since you know differently.

No it's YOUR own, flawed, conclusion.
Everybody showed that every mobo provides this level of Voltage TO ALL CPUS on that platform, but only the new CPUs have a problem with it.
The SoC voltage is only too high for the new CPUs.

 

[TerryLaze]

this is no different then when intel " turns a blind eye " when a mobo maker does any thing it can, to allow intels cpus do and use what ever it needs to top the bench mark charts, i sure dont recall you saying anything about that, or complaining about it.

The heck?!
I complain all the time about people calling this default...

its one of the reasons why some wont touch intel for the last 2 or 3 gens, as intel sucks power from the wall, (regardless which cherry picked graph or test you use.) and one reason why no one i know is looking at intel right now, some have picked up non x3d cpus and are running just fine .

So you don't buy intel because mobos aren't configured correctly but you still buy AMD because mobos aren't configured correctly.
The sense AMD fanboys are making...

 

[TerryLaze]

That's beside the point. My question was: roughly how often does it happen?

Often enough for AMD and board makers to panic and to release one bios/agesa after the other.

 

[Ogotai]

I complain all the time about people calling this default...

cause it IS the default, it is what intel " turns the blind eye " about and allows mobo makes to do what they want, at stock. AKA default settings as set by the mobo maker. its not our fault you dont know what stock settings mean, now is it ?

So you don't buy intel because mobos aren't configured correctly but you still buy AMD because mobos aren't configured correctly.
The sense AMD fanboys are making...

nice try word twister. that is no where near what i said. they ARE configured correctly, to use as much power and other setting as needed to make intel top the charts, as per what the mobo makers set as " stock settings " its just some wont use intel because of this, they dont like the amount of power, intel " turns a blind eye on " to run their cpus. the SAME can be said about the intel fanboys.

 

[bit_user]

Often enough for AMD and board makers to panic and to release one bios/agesa after the other.

As I explained, more outlandish and evocative outcomes skew risk perception. Even if it happened only a handful of times, the publicity it received due to the photos would be enough to prompt such a reaction.

For them, failing to act would be seen as aloof, uncaring, or irresponsible, even if the true failure rate were well within the estimates of their warranty underwriters. So, you really can't read into the mere fact that they addressed the problem. Had they not done so, I'd bet you'd be complaining about that, too.

 

[SunMaster]

No it's YOUR own, flawed, conclus

Unlike you, I can back up my claim.

Gamers Nexus

View: https://www.youtube.com/watch?v=kiTngvvD5dI&t=1s


Igors Lab

Pin analysis of the destroyed Ryzen 7800X3D - All burned pins supply the VDDCR (CPU Core Power Supply) | igor´sLAB

The process with the destroyed Ryzen 7000X3D CPUs has of course already interested me and I didn’t want to just cover a news and that was it. So I took the damage picture from Reddit and put the back…

www.igorslab.de

Tom's Hardware

AMD's 'Ryzen Burnout' Fixes Reportedly Plagued With Bugs

The bugs keep getting worse for AMD's AM5 platform.

www.tomshardware.com

Wccftech quoting AMD

AMD Identifies Root Cause of Ryzen 7000 CPU Burnout Issue, Related To Higher Chip Voltages

AMD has identified the root cause of the Ryzen 7000 CPU burnout issues which happen to be higher voltage on the chip.

wccftech.com

Where are your sources?

 

[LordVile]

No it's YOUR own, flawed, conclusion.
Everybody showed that every mobo provides this level of Voltage TO ALL CPUS on that platform, but only the new CPUs have a problem with it.
The SoC voltage is only too high for the new CPUs.

The issue lies with the board makers and AMD, they both cocked up with the 3D chips. Mainly ASUS on the board manufacturers side but AMDs affected other brands too

 

[TerryLaze]

cause it IS the default, it is what intel " turns the blind eye " about and allows mobo makes to do what they want, at stock. AKA default settings as set by the mobo maker. its not our fault you dont know what stock settings mean, now is it ?

nice try word twister. that is no where near what i said. they ARE configured correctly, to use as much power and other setting as needed to make intel top the charts, as per what the mobo makers set as " stock settings " its just some wont use intel because of this, they dont like the amount of power, intel " turns a blind eye on " to run their cpus. the SAME can be said about the intel fanboys.

Yes they are configured correctly just as the ryzen mobos are configured correctly, they use as much V as they need to make XMP work correctly so that the x3d CPUs can top the benchmark charts, and that makes the CPUs blow up. AMD also "turned the blind eye" toward it until the benchmark results came out, and only then they started to do something about it.

Unlike you, I can back up my claim.

Gamers Nexus

View: https://www.youtube.com/watch?v=kiTngvvD5dI&t=1s


Igors Lab

Pin analysis of the destroyed Ryzen 7800X3D - All burned pins supply the VDDCR (CPU Core Power Supply) | igor´sLAB

The process with the destroyed Ryzen 7000X3D CPUs has of course already interested me and I didn’t want to just cover a news and that was it. So I took the damage picture from Reddit and put the back…

www.igorslab.de

Tom's Hardware

AMD's 'Ryzen Burnout' Fixes Reportedly Plagued With Bugs

The bugs keep getting worse for AMD's AM5 platform.

www.tomshardware.com

Wccftech quoting AMD

AMD Identifies Root Cause of Ryzen 7000 CPU Burnout Issue, Related To Higher Chip Voltages

AMD has identified the root cause of the Ryzen 7000 CPU burnout issues which happen to be higher voltage on the chip.

wccftech.com

Where are your sources?

All of these videos prove my point and not yours, all of the mobos use the same V , that did not blow up any previous CPUs, on the x3d CPUs as on any other CPU.
AMD didn't provide a way of distinguishing the 3xd CPUs from the normal ones, that's only AMDs fault and nobody elses.
It's completely ridiculous to believe that AMD was somehow caught by surprise by the settings the mobo makers are using...that would mean that everybody at AMD is an complete idiot, and that's just not the case.

Had they not done so, I'd bet you'd be complaining about that, too.

They would only not have done anything about it if it wasn't a problem at all, so I would be complaining out of the blue for there not being any problem....
I'm known for making new topics every day just randomly complaining about AMD not having any problems...
AMD logic!

 

[msroadkill612]

So, if someone who already bought an AM5 CPU, which could have been purchased, AT MOST, 6 months ago, and has over-volted it, and has decided to swap in a 7800X3D....

That seems like an absurdly niche case. And, for the tiny number of people who are swapping from one AM5 chip to this one, and over-volted, one would think they'd be knowledgeable enough not to be that careless.

This seems like you're really grasping at straws to find a "flaw"/complain about AMD.

Its evident from the negative tone of his language.

 

[helper800]

Its evident from the negative tone of his language.

I am not defending him, however, I take issue with people attributing 'tone' to text. Tone is specifically an auditory verbal and body language que. You can talk all day about peoples writing 'style' but 'tone' is certainly an incorrect way to ascribe inferences, and assumptions we all make about reading language as text from someone else.

 

[bit_user]

I am not defending him, however, I take issue with people attributing 'tone' to text. Tone is specifically an auditory verbal and body language que.

I think you're splitting hairs. Sure, a better word exists, but we all know that writing can have a slant or subtext.

You can talk all day about peoples writing 'style' but 'tone' is certainly an incorrect way to ascribe inferences, and assumptions we all make about reading language as text from someone else.

According to Wiktionary, it's not wrong to use tone, in this context:

7. (literature) The manner in which speech or writing is expressed.

Example:

1850, William Cullen Bryant, Letters of a Traveller

Their tone was dissatisfied, almost menacing.​

https://en.wiktionary.org/wiki/tone#Noun​

​

 

[helper800]

I think you're splitting hairs. Sure, a better word exists, but we all know that writing can have a slant or subtext.

Maybe, however, the 'tone' of writing always seemed like an oxymoron to me.

 

[msroadkill612]

Thanks for the feedback.

Admittedly, DDR5 pricing has dropped pretty precipitously since I last checked, which couldn't have been more than two weeks ago. However, I think there are still some meaningful deltas.

Of note, your price comparisons above has what is effectively an overclocked DDR4-4000 kit (Raptor does DDR4-3200 at stock), and a DDR5-4800 kit that doesn't meet the minimum spec for AMD's chips — Zen 4 does DDR5-5200, so that memory kit is insufficient for basic operation.

As we all know, there is a sliding scale of price increases as you go further beyond the rated stock speeds — there isn't a linear correlation between speed and pricing, you really can spend stupid amounts on faster kits. So as a means of leveling the playing field, I prefer to compare memory kit pricing at the stock CPU speeds.

Here are a few PCPartPicker links to lists of kits available from any number of retailers. If you aren't reading this post today, be aware that memory pricing is very volatile and can change quickly.

DDR4-3200 (Raptor)

The ten cheapest 32 GB kits range from $56 to $59
The ten cheapest 64 GB kits range from $106 to $130

DDR5-5200 (Ryzen 7000)

The ten cheapest 32 GB kits range from $90 to $105
The ten cheapest 64 GB kits range from $199 to $224

I prefer to present data like this in table form, but the takeaway is that you will pay, at a minimum, 60% more for a 32GB kit and 87% more for a 64 GB kit.

Yes, the relative impact of paying any amount more for a component can be debatable when viewed through the prism of total system cost, but do realize that if you are paying more for one component you are draining funds from another. Whether it's the extra $35 or the extra $93, I would rather spend that cash on a better GPU or SSD, or just whatever (I have a thing for nice fans). Saving that cash means it is a cost that you don't have to pay elsewhere.

And yes, while the $449 Ryzen 7 7800X3D is a premium chip today, so was the $449 Ryzen 7 5800X3D at launch. Today, the $319 Ryzen 7 5800X3D is an absolute steal and I recommend it to anyone looking for a great gaming-centric chip (In fact, I left it in our Best CPU for Gaming list despite the addition of the 7800X3D today). This is all to say that the 7800X3D will cost less than $449 in the very near future, so it will not be limited to upper-tier builds forever. These reviews live a long time (hopefully forever), so we have to also account for the long term.

Even at $449, the 7800X3D isn't a super top-tier chip for folks that have unlimited budgets. Some enthusiasts will be trying to scrape up enough cash to build a 7800X3D system, and for them, these pricing deltas are something they should consider.

And yes, DDR5 will get cheaper in time, and then the deltas won't matter as much. Micron told me that bit crossover (the point when they sell more DDR5 than DDR4) will occur early next year. However, they still don't expect DDR5 to be cheaper than DDR4 because it simply requires more silicon to meet the same amount of capacity (new error correction overhead), and it also has onboard power circuitry that adds cost. According to them, DDR5 won't be cheaper than DDR4 until the latter is two generations behind. However, the price difference will eventually get smaller. Not to mention that newer chips will no longer support DDR4.

Finally, let's not forget the amount of cost that DDR5 support adds to motherboards. These boards require more robust materials and more expensive componentry to support the faster interface.

Here is how I derive the comparisons for motherboard pricing:

Intel DDR4 B760 starts at $109 to $155
AMD B650 starts at $119 to $195
Intel DDR5 B760 is $140 to $195

Intel DDR4 Z790 is $179 to $275
Intel DDR5 Z790 is $195 to $250
AMD X670 is $259 to $335

Raptor can drop into previous-gen boards, too (if they have the right BIOS):

Intel DDR4 Z690 $129 to $195
Intel DDR4 B660 $85 to $127

Oscar Wild ? "I spent a tiring day editing, In the morning I added a comma, and in the afternoon I removed it."

 

[phxrider]

Did you guys ever consider running the benchmarks with something other than a 4090, like GPUs in the $500-$1100 range? You know, so those of us who aren't going to spend over $1600 on a GPU can get an idea of how much these CPUs actually would benefit us?

I promise I do understand the concept of comparing CPUs without introducing GPU bottlenecks, but how they perform with the most outrageously priced gaming GPU in the world is not much help to the rest of us trying to figure out how much CPU we need to avoid a CPU bottleneck with a more reasonably priced GPU like my 6800xt, or even a 4080 or 7900xtx....

 

[King_V]

Because the goal of the benchmarks is to compare the CPUs, exclusive of other considerations.

What you're asking for is out of scope, and would not constitute a CPU performance comparison. It would have to involve numerous factors, not just CPU and GPU, but RAM, monitor, which specific games you're talking about, etc. And because of those other factors, the idea of an overarching "bottleneck" on the system is completely meaningless.

Literally the same system can have the CPU as the "bottleneck" in one game, and the GPU as the "bottleneck" in a different game.

 

[bit_user]

Because the goal of the benchmarks is to compare the CPUs, exclusive of other considerations.

What you're asking for is out of scope, and would not constitute a CPU performance comparison. It would have to involve numerous factors, not just CPU and GPU, but RAM, monitor, which specific games you're talking about, etc. And because of those other factors, the idea of an overarching "bottleneck" on the system is completely meaningless.

Literally the same system can have the CPU as the "bottleneck" in one game, and the GPU as the "bottleneck" in a different game.

You can argue from a purist's perspective, but if Tom's wants to provide useful information to PC builders, then there's definitely a role for measuring ideal component pairings, in order to help people find combinations that yield a balanced system.

I'm not saying that has to be in the CPU review, itself. However, one bit of advice I offered to the Editor In Chief was to publish more benchmarks of the sort that would help people pair the right CPU cooler with a given CPU. The scenario raised by @phxrider is another variation on that theme.

What I have seen, on this site and others, are articles around launch time of new AA and AAA titles that measure its performance on different graphics cards. I've never read them, since I don't play games, but perhaps they have useful information about good CPU pairings.

 

[TJ Hooker]

Most graphics settings in game have little or no direct impact on CPU demand. So by testing in a nearly pure CPU-bound scenario you know what FPS a given CPU is capable of in a variety of games. You can then looks at graphics card reviews to see what FPS cards are capable of at various settings. To get a rough idea of how a particular CPU + GPU combo would perform, just take the lower of the two FPS values from the CPU and GPU reviews. Just pick a CPU and GPU that both comfortably exceed whatever FPS threshold you desire in the sorts of games you play.

 

[King_V]

What I have seen, on this site and others, are articles around launch time of new AA and AAA titles that measure its performance on different graphics cards. I've never read them, since I don't play games, but perhaps they have useful information about good CPU pairings.

Now, that, I can see being a thing. But, honestly, I think that any sort of review like this really would have to be game-specific.

I'm not sure how practical it would be to try to have such an article that would cover multiple games/most games/etc. I was about to suggest maybe a clustering of games based on the 3D engine they use, but even then, I think there's too many variables to make this practical much beyond a per-game level.

 

[bit_user]

Now, that, I can see being a thing. But, honestly, I think that any sort of review like this really would have to be game-specific.

Yes.

Starfield PC Performance: How Much GPU Do You Need?

FSR 2 upscaling and optimized for AMD.

www.tomshardware.com

A few years ago, Toms would regularly spec out gaming PCs at different price points. This was an interesting exercise, since it meant matching CPU, GPU, cooler, etc. I haven't seen one of those articles, in a while.

 

[King_V]

Could be related to the decrees from the current overlords, perhaps?

 

[bit_user]

Could be related to the decrees from the current overlords, perhaps?

Maybe they just lost the budget for actually building & benchmarking the PCs they'd spec out.

They used to do a lot of giveaways... I haven't seen those in a while, either (although I think most were manufacturer sponsored).