AMD Ryzen 7 5700X

[Dr. Dro]

5800X is B0 or new VRM-B2?

All newly manufactured processors are B2, but there's no difference between it and the original B0 stepping. There's no changes to the product design nor erratum fixes, the designation is added to processors built with AMD's new manufacturing process. Note I'm not talking about the lithography or changes in the node, just the changes in manufacturing that make building the processors easier and cheaper for the company.

Right now I am more concerned about, why reviewers don't tell people about coil whine in motherboards/gpus, it makes purchase 100 times harder now. But at least I heard about bad 1700 socket cover in time, almost bought it myself :C

Coil whine issues are very rare even in lower end motherboards nowadays. Whenever that occurs it's also generally due to a bad power supply with high ripple or otherwise dirty power as well.

By the way, even the 105W (142W PPT) spec processors like the 5800X and up have an Eco mode that sets them down to the 65W (84W PPT) spec. I believe Eco mode also brings the 65W processors down to the next tier below, which should be 45W (54W PPT), similar to the one used in mobile high performance (H-series) or desktop energy efficient (GE-series) processors.

 

[eidairaman1]

I got this piece (5800OEM) for 262 on Ebay last November (guy claimed it was 5800X but I called his bs out and he agreed to giving 100 back from 362.

 

[Pepamami]

there's no difference between it and the original B0 stepping

People in internets say it has better temperatures and oc clocks

Coil whine issues are very rare even in lower end motherboards nowadays.

my old 1070 Strix got awful coil sound, not a teapot, but still awful (compared to old 780), with bad PSU its maybe getting worse.
When I asked about it, people said: "Its not a harware defect! Its Design!".
MB/GPU coil sound must be in reviews, people asks a lot about it in stores Q/A section, they dont want surprises. Its like same goes for Display review nowdays, with PWM flickering checks.

And speaking about high end motherboards:
I found some annoying coil sounds on Asus Strix B550-i gaming, but did not find any of annoying sounds on gigabyte b550i aorus pro ax.

 

[Dr. Dro]

People in internets say it has better temperatures and oc clocks

People like to claim a lot of things, however, there is no evidence that to support these claims and nobody has been able to prove that thus far. AMD's official statement on B2 stepping is that these are the result of a production-chain change that would not affect or change anything in the silicon.

AMD Confirms B2 Ryzen 5000 Stepping Brings No Performance Improvements

AMD has recently confirmed that the upcoming B2 stepping for Ryzen 5000 processors will only improve production and availability with no impact on performance. The new stepping will be deployed within the next 6 months and will be fully compatible with existing motherboards without the need for...

www.techpowerup.com

Regarding better clocks and temperatures: I personally believe this to be anecdotal and highly unlikely, considering AMD's statements and how the Zen 3 architecture works.

My personal understanding/theory is that a a chip that goes far in the curve optimizer isn't a golden sample, it's quite the opposite, this being because Ryzen's v/f curve is managed in real time according to the processor's thermal and electrical conditions, as opposed to a traditional processor that would have fixed settings and need X voltage to operate at Y frequency. A higher-quality processor that generates less heat and requires less voltage would thus attain a nominally higher level of performance out of the box, because the reliability algorithm would allow it to sustain a slightly higher frequency before it decides "that's enough". Thus it stands to reason that AMD uses these better dies on higher-end SKUs with more aggressive clock speeds out of the box, something that may be corroborated by the narrower curve optimizer and PBO scalar range the further up in the product stack.

The 5800X3D seems to be an exception to this rule, they are locked and thus out of the box all behave the same, using PBO2 Tuner to access the curve optimizer (as Ryzen Master hides the clock control option and BIOSes disable access to the AMD Overclocking section if this specific processor is installed) shows that most of them will max out the negative adjustment or come close to it (-25 to -30), that plus the fixed v/f curve of that specific model leads me to believe AMD simply hasn't binned them and picked a conservative, safe curve for them that the engineers believe will be safest for the 3D V-Cache and shipped every unit that way.

Note that this is my opinion and it's based on my personal understanding of how these processors work, I may very well be talking gigantic nonsense here, but if you think about it, the reasoning is sound. I hope to see a review on this someday, though it would need someone with a lot of time and a generous amount of CPU samples throughout the entire product stack.

my old 1070 Strix got awful coil sound, not a teapot, but still awful (compared to old 780), with bad PSU its just getting worse.
When I asked about it, people said: "Its not a harware defect! Its Design!".
MB/GPU coil sound must be in reviews, people asks a lot about it in stores Q/A section, they dont want surprises. Its like same goes for Display review nowdays, with PWM flickering checks.

And speaking about high end motherboards:
I found some annoying coil sounds on Asus Strix B550-i gaming, but did not find any of annoying sounds on gigabyte b550i aorus pro ax.

I've owned two AM4 motherboards, the Crosshair VI Hero and the Strix B550-E (which I sidegraded to only because AMD was being stingy with X370 support for the longest time, mind you), and neither have had coil whine issues. It's been a long time since I had any component with coil whine... last GPU that I had with some was the original Titan, almost a decade ago. Maybe I've gotten lucky? Some coil whine is normal, but if it's audible beyond your cooling, then it can be quite annoying indeed. Unfortunately, an RMA is your only option if it gets too bad.

 

[PanicLake]

Oh right, let me run it today, will update the review in the evening

@W1zzard Thank you, really appreciate.

 

[Pepamami]

People like to claim a lot of things, however, there is no evidence that to support these claims and nobody has been able to prove that thus far

Production change can lead to quality improvements without change specification and features of product.
For example: i7 920 with steppings D0 (better) and C0, i5 10400F with steppings Q0 (better) and G1.

Unfortunately, an RMA is your only option if it gets too bad.

Thats why I am coplaining. I want to know about it before buying, not after.

 

[Dr. Dro]

Production change can lead to quality improvements without change specification and features of product.
For example: i7 920 with steppings D0 (better) and C0, i5 10400F with steppings Q0 (better) and G1.

Not the same, for example, the processors you've mentioned such as Bloomfield D0 contained erratum fixes, according to AMD, Vermeer B2 does not contain any changes to the hardware itself. It is true that over time, as the process matures, you may have an easier time finding higher quality silicon, but AMD's product stack is quite full and their server business is booming, all of which share the same dies. The only constant is how inconsistent processor binning will remain for the time being

Thats why I am coplaining. I want to know about it before buying, not after.

Unfortunately, coil whine is a per-piece issue, and without a large amount of samples tested, it's impossible for a reviewer to know beforehand if a board design is prone to coil whine or not. You may have it, but someone else with a completely identical board even coming from the same batch might not experience it. This would render coil whine section in a review useless.

 

[Wirko]

As a long time forum user - even before this account im typing on now - i see a "slip" in general attitude of this community from one of helpfulness to one of bitterness/discontent with each other..... even to go as far as arguing about nonsensical stuff.

I've been here for a year and a half and that's enough to notice the shift towards intolerance. Even some people here who contribute a lot to these forums using their knowledge and critical thinking, sometimes get drawn into this. Which makes me sad.

 

[kapone32]

How much better is the R 7 5700X compared to my R 7 3700X ?
I see no reason to change it.
Motheboard is a MSI MEG X570 Unify.

Honestly, about 10-20% in most applications.

 

[Pepamami]

Not the same, for example, the processors you've mentioned such as Bloomfield D0 contained erratum fixes, according to AMD, Vermeer B2 does not contain any changes to the hardware itself. It is true that over time, as the process matures, you may have an easier time finding higher quality silicon, but AMD's product stack is quite full and their server business is booming, all of which share the same dies. The only constant is how inconsistent processor binning will remain for the time being

I dont have proofs that B2 is better than B0, neither I dont have proofs that B0 is 100% the same as B2 in temperatures, power consumption and etc.
All I am doing here is asking: "which revision do u have?" and "what if B2 5800X is not same as B0 5800X, and u cant compare new 5700X to old B0?".

 

[Colddecked]

This is the CPU of choice for mini ITX builds and ecologically aware users as it combines performance with low temperatures and a low TDP.

The cpu of choice is the one you get the best deal for because you can set a 5800x to the same tdp.

 

[Dr. Dro]

I dont have proofs that B2 is better than B0, neither I dont have proofs that B0 is 100% the same as B2 in temperatures, power consumption and etc.
All I am doing here is asking: "which revision do u have?" and "what if B2 5800X is not same as B0 5800X, and u cant compare new 5700X to old B0?".

There is no need to burden yourself with proving that, as there is no B0 stepping 5700X, this is a newly released SKU. The closest thing to it would be a B0 5800X in eco mode, which is essentially what this processor is.

My 5950X is a B0, I bought it a few months before the B2 stepping was announced by AMD. It's about as tightly binned as a Ryzen can be

 

[bug]

There are a handful of review sites that will state outright if a reviewed product is no good.

The mitigating factor is whether or not any given site receives reviewer samples from manufacturers. Most review sites simply do not have the budget to acquire all of their own review units. If they are forthright and honest about how much a product sucks, the flow of sample units will dry up real quick. In the early years of Tom's Hardware and Anandtech (back in the Nineties), they called out manufacturers periodically.

So most reviewers still need to find some sort of scenario to recommend the product being reviewed even if they start digging deep for a justification that was refuted earlier in the review.

No one is required by law to believe everything they read. Even if most of a review is plausible, a reader might disagree with certain parts. That's part of reading comprehension and being mature enough to decide how much of what you are reading is plausible.

No one can please everyone all the time. And we're not just talking about PC hardware reviews.

If you don't like a particular reviewer, don't read them. If you don't like any reviewers, skip all of them. Remember that writing isn't a precious skill anymore. Writing has been completely commodified thanks to the Internet.

For me reviews do still have some utility. I don't have the resources to test every single product on the market. The time factor is the biggest one. Some categories -- like power supplies -- require specialized test equipment and a much deeper knowledge of electrical engineering than I have.

Sometimes I end up at a conclusion that wasn't intended by the writer. Here's a Tom's Hardware legacy comparison chart of thermal pastes:

https://cdn.mos.cms.futurecdn.net/yqSq6qVDxvcsSQc3kFUvfG-970-80.png.webp

More important that learning which paste is "the best" is the fact that conventional thermal pastes are almost all within 3°C of each other. Only the liquid metal pastes have more conductive performance. Pads are slightly worse. So if a manufacturer includes some bundled thermal paste with a piece of hardware, I should comfortable using it because it's essentially no different than a premium paste like Noctua, Arctic, whatever.

Also a lot of reviews conclude with the reviewed product not being a standout value compared to the competition. Remember that most review judgments are based on the MSRP so when the pros and cons are added up and compared to the price asked, there's the fundamental question "Do I spend the money on this or something else?"

If you buy a bag of potato chips, most likely you left behind fifty other bags of chips and possibly hundreds of other snacks. We do this every day for multiple purchase decisions.

Long story short, I never meant to imply reviews are useless. Quite the opposite, in this day and age when brick and mortar stores are increasingly harder to find, reviews (for certain products) are more important than ever.
It's just that, like you said, most feel the urge to finish with a positive conclusion. And the way it was done this time was downright hilarious. For me, at least.

 

[Pepamami]

There is no need to burden yourself with proving that, as there is no B0 stepping 5700X, this is a newly released SKU

but there are B2 5800X and B0 5800X wot if they slightly different >.>

 

[Dr. Dro]

but there are B2 5800X and B0 5800X wot if they slightly different >.>

As clearly stated by AMD, there are no changes in the final product. Performance and hardware bugs are the same.

 

[Anymal]

Now we know why! W1zz, any comment why you labeled it Highly recommended if 5800x is only 10usd more and 12600k is cheaper and faster in games?

 

[cvaldes]

Long story short, I never meant to imply reviews are useless. Quite the opposite, in this day and age when brick and mortar stores are increasingly harder to find, reviews (for certain products) are more important than ever.
It's just that, like you said, most feel the urge to finish with a positive conclusion. And the way it was done this time was downright hilarious. For me, at least.

Giving a backhanded compliment is a learned skill. Recognizing a backhanded compliment is also a skill that a reader must learn.

Let's say someone gives you a homemade cookie that isn't very good. What are you going to say? It's awesome? It sucks? It's very distinctive? It's something your kids would enjoy?

You can say "this cookie sucks" but then you risk having that baker tell other bakers of your critical review and soon no one will give you cookies. And maybe those dinner party invites dry up.

A non-dcikhead will likely try to find some sort of positive comment. That's the natural response from a well-adjusted, socially mature adult.

Remember that some comments that can be interpreted as backhanded might actually be unintentional.

You also have to consider that any given reviewer is likely addressing a number of audiences. Not every PC hardware reviewer is going to pass judgment expecting all of their readership to be hardcore overclockers or hardware geeks. If they do that, they will limit the scope of their audience. And that reduces ad revenue opportunities for an ad-supported site like TPU.

 

[W1zzard]

Are some of you seriously claiming the data in my review suggests "the 5700X sucks" and that I'm sugarcoating that just to keep AMD happy? It's a really really good CPU for gaming, applications, low power scenarios, weak cooler scenarios, works on a gazillion existing motherboards, uses dirt cheap DDR4 memory .. sure it could be cheaper like everything else on the planet, but the pricing is definitely not totally insane

 

[bug]

Are some of you seriously claiming the data in my review suggests "the 5700X sucks" and that I'm sugarcoating that just to keep AMD happy? It's a really really good CPU for gaming, applications, low power scenarios, weak cooler scenarios, works on a gazillion existing motherboards, uses dirt cheap DDR4 memory .. sure it could be cheaper like everything else on the planet, but the pricing is definitely not totally insane

I am seriously not suggesting anything beyond what I have explicitly said: it's funny that you point out there are better picks both in AMD and Intel camp and then you go "highly recommended". If I can get significantly more performance for $10 extra from 5800X or about the same performance for less money from 12600k, it's pretty hard to see what's "highly recommended" about this.

When you feel like you need top open with "it's a really good CPU for gaming", that's a red flag right there: every CPU is really good for gaming, because gaming is rarely CPU bottlenecked. "Really good for applications"? It loses almost every applications test to both 5800X and 12600k. Sure, it's not a slow CPU by any metric. But it doesn't live in a void, all things considered this is "highly recommended" if you can't get any of the other two, for some reason.

 

[Dr. Dro]

I am seriously not suggesting anything beyond what I have explicitly said: it's funny that you point out there are better picks both in AMD and Intel camp and then you go "highly recommended". If I can get significantly more performance for $10 extra from 5800X or about the same performance for less money from 12600k, it's pretty hard to see what's "highly recommended" about this.

When you feel like you need top open with "it's a really good CPU for gaming", that's a red flag right there: every CPU is really good for gaming, because gaming is rarely CPU bottlenecked. "Really good for applications"? It loses almost every applications test to both 5800X and 12600k. Sure, it's not a slow CPU by any metric. But it doesn't live in a void, all things considered this is "highly recommended" if you can't get any of the other two, for some reason.

However, the 5800X has almost double the TDP, and buying it to use in Eco mode doesn't make sense if you have to spend more money for it, even if it's $10, those ten bucks will pay for the TIM or a better heatsink

That it's a drop in upgrade to virtually any AM4 board warrants the recommended status IMHO. And only because AMD finally conceded on updating all motherboards.

 

[Shtb]

If I can get significantly more performance for $10 extra from 5800X or about the same performance for less money from 12600k, it's pretty hard to see what's "highly recommended" about this.

Why is it hard for so many people to consider, that for 12th generation Intel you will always need a new, more expensive motherboard,
when even for Ryzen 5000 there has been an AM4-platform for several years, which obviously saves you money, if you want more performance.

Is it so hard to consider this?
Is it so hard to keep in mind not only CPU price, but of the band - CPU plus motherboard?
Or is it all about fanboyism for Intel?

 

[tussinman]

However, the 5800X has almost double the TDP, and buying it to use in Eco mode doesn't make sense if you have to spend more money for it, even if it's $10, those ten bucks will pay for the TIM or a better heatsink

To add to this the main reason why the 5800x got to it's current price is because the 5700x originally came out at a much lower price.

Not sure why that user has such a snarky attidude about the 5800x being not much more expensive, the reason why it's not much more expensive is literally because the 5700x came out........ (that's not a negative for the 5700x that's a plus, it was not only cost effective at release but it brought down the price of the chip above it)

 

[Udyr]

Apparently, this needs to be clarified

 

[mahoney]

Real shitty from AMD not releasing these cpu's in 2020. Now you basically have the 5700x and 5800x with less than 10€ difference and the 5600x and non x going for the same price.

 

[Dr. Dro]

Apparently, this needs to be clarified
View attachment 254357

*Widely available and quite affordable AM4 motherboards for new buyers, or a platform that's been around since 2017 with intense market penetration and a huge install base
*Not necessarily low-budget, you could use this with an NH-L9a-AM4 chromax.black for a distinct build look, for example
*Bingo, which is what these standard power desktop processors target, simpler builds with leaner power requirements
*So, the largest majority of Ryzen users?

Real shitty from AMD not releasing these cpu's in 2020. Now you basically have the 5700x and 5800x with less than 10€ difference and the 5600x and non x going for the same price.

Precisely. And for refusing to support older motherboards well until the pressure from Alder Lake came in. But this is where you're sorely reminded that AMD is a business, not our friends, much to their fans' dismay who still see them as some sort of underdog that needs our unwavering support in every move. TRX40 owners would love a word, they're still out in the cold, too.