Intel B560 is a Disaster: Huge CPU Performance Differences and a Power Limit Mess

The best reason to buy an Intel CPU instead of AMD right now is called "the only one I could find today was Intel".

Conversely the best reason to buy an AMD CPU instead of Intel right now is "the only one I could find was AMD". 🙄🙄🙄

AMD may be objectively better this generation, but the supply situation is such a train wreck that day to day availability trumps everything else. 😭😭😭
 
The best reason to buy an Intel CPU instead of AMD right now is called "the only one I could find today was Intel".

Conversely the best reason to buy an AMD CPU instead of Intel right now is "the only one I could find was AMD". 🙄🙄🙄

AMD may be objectively better this generation, but the supply situation is such a train wreck that day to day availability trumps everything else. 😭😭😭

AMD cpu's appear to have better availability today. As for Intel, only buy Z590.
 
The article is clickbait because it doesn't really tell you what the real issue is until the very end, just like ND40oz says, it's the default settings of the boards.

I'm not sympathetic though. Intel is allowing manufacturers to play fast and loose with the turbo and power settings to make it seem like there is increased performance. They deserve articles like this.

The manufacturers are only interested in showing product differences. Why buy that cheap ASRock board when you can spend an extra $50 and get 30% increased performance out of the same chip on this fancy Gigabyte board?
 
For those too lazy to click on the article, the gaming performance isn't a huge disaster. If there's a silver lining in this I think it's that some board manufacturers will respect the power limits (though it looks like that can be tweaked).
 
The article is clickbait because it doesn't really tell you what the real issue is until the very end, just like ND40oz says, it's the default settings of the boards.

I'm not sympathetic though. Intel is allowing manufacturers to play fast and loose with the turbo and power settings to make it seem like there is increased performance. They deserve articles like this.

The manufacturers are only interested in showing product differences. Why buy that cheap ASRock board when you can spend an extra $50 and get 30% increased performance out of the same chip on this fancy Gigabyte board?
I disagree. It's not clickbait.

This is 100% intel's fault for not enforcing standards. Either the standard is set with the mobo manufacturer that the TDP limits are either set by default or not. It's not even that there are limits at all. There should never be this wild difference in performance. That would make your product look inconsistent. Intel has allowed too much leeway with the mobo manufacturers and this is the by product.

Intel could have fixed all of this by forcing limits by default but easily changed with settings. Problem solved.
 
  • Like
Reactions: ~CS~
like this
If you're building a 4K gaming system, the CPU barely matters. That's one reason you can basically buy whatever you can find.
 
I disagree. It's not clickbait.

This is 100% intel's fault for not enforcing standards. Either the standard is set with the mobo manufacturer that the TDP limits are either set by default or not. It's not even that there are limits at all. There should never be this wild difference in performance. That would make your product look inconsistent. Intel has allowed too much leeway with the mobo manufacturers and this is the by product.

Intel could have fixed all of this by forcing limits by default but easily changed with settings. Problem solved.
Shenanigans have been going on for years with performance impacting BIOS settings. Multi-core turbo has been the best known one on the desktop side. Motherboards often not enabling XMP ram settings by default is another one. (This is Ian Cutress @ Anandtech's hobbyhorse, and why the site only tests at JEDEC ram speeds.) Wide ranges in performance due to power different management settings has also been a thing in laptops for many years. Intel's never stopped any of them before; they're not going to do anything about this one either.
 
I disagree. It's not clickbait.

This is 100% intel's fault for not enforcing standards. Either the standard is set with the mobo manufacturer that the TDP limits are either set by default or not. It's not even that there are limits at all. There should never be this wild difference in performance. That would make your product look inconsistent. Intel has allowed too much leeway with the mobo manufacturers and this is the by product.

Intel could have fixed all of this by forcing limits by default but easily changed with settings. Problem solved.

Sure it is. Throw the terms, "Intel" and "disaster" in an article title to get clicks. Seems pretty straightforward to me.

I agree that for consistency sake they should have an "default" that all boards follow. Ironically, it seems like they are pulling an AMD Zen1 move. Sure the CPU is capable of more, but blaming the mobo manufacturers for not having good VRM, etc. to allow more consistent performance. Along with the wide variety of build quality, you lucky to get anything above the default.
 
Sure it is. Throw the terms, "Intel" and "disaster" in an article title to get clicks. Seems pretty straightforward to me.

I agree that for consistency sake they should have an "default" that all boards follow. Ironically, it seems like they are pulling an AMD Zen1 move. Sure the CPU is capable of more, but blaming the mobo manufacturers for not having good VRM, etc. to allow more consistent performance. Along with the wide variety of build quality, you lucky to get anything above the default.
Ok, from a language perspective I can see how inflammatory terms can drive readership. But, at the end, the sentiment is still true. Intel could do a lot to better this situation, IMO. I'm just a dude in the server arena, not a tech reviewer, but I don't think it's a crazy claim.
 
Ok, from a language perspective I can see how inflammatory terms can drive readership. But, at the end, the sentiment is still true. Intel could do a lot to better this situation, IMO. I'm just a dude in the server arena, not a tech reviewer, but I don't think it's a crazy claim.

I agree. There's definitely a problem, and it starts with Intel's lack of consistent direction with their motherboard partners.
 
Of all the "internet-isms" this one bugs me the most. You don't want to read but you still want to comment. RME

The fact that I could sum up that entire article in one sentence tells everyone everything they need to know about it.

I disagree. It's not clickbait.

This is 100% intel's fault for not enforcing standards. Either the standard is set with the mobo manufacturer that the TDP limits are either set by default or not. It's not even that there are limits at all. There should never be this wild difference in performance. That would make your product look inconsistent. Intel has allowed too much leeway with the mobo manufacturers and this is the by product.

Intel could have fixed all of this by forcing limits by default but easily changed with settings. Problem solved.

Intel has been doing this for 30 years, why should they change now? If mb manufacturers want to stick strictly to the spec, that's their choice. We'd have never had boards like the Abit BP6 if they did what you're suggesting.

Do your research before you buy a motherboard like we've been doing since the beginning of the internet. It's pretty simple.
 
The fact that I could sum up that entire article in one sentence tells everyone everything they need to know about it.



Intel has been doing this for 30 years, why should they change now? If mb manufacturers want to stick strictly to the spec, that's their choice. We'd have never had boards like the Abit BP6 if they did what you're suggesting.

Do your research before you buy a motherboard like we've been doing since the beginning of the internet. It's pretty simple.
Thats the problem, 99% of the population do not do their research, and Intel knows this and doesn't flinch. I agree the CPU's were running within specification. The problem is 99% of the population only see a CPU running out of spec with day 1 reviews on a Z590 or some B560 motherboards...,then they look at the graphs and decided to purchase the Intel CPU, and buy the cheapest motherboard they can.....thinking they will get the same kind of performance when they won't.

That is a disaster IMO. Sure most of us here on this website know this, and numerous other hardware tech people know better...but yea we are the minority.
 
Last edited:
  • Like
Reactions: kac77
like this
There isn't anything exactly wrong with buying a board, and it working out the gate to the advertised specifications with no need to manually tweak things. For many, that is kind of the goal not everybody wants to have to build a machine and spend days tuning it in and working to get it stable.
 
Of all the "internet-isms" this one bugs me the most. You don't want to read but you still want to comment. RME
I think they mean TLDR as in the short-sum of the article, as their description of it is pretty accurate, and not that they didn't actually read it.
It could have been worded better, but I do agree with their post.

You get what you pay for, and the board itself does operate within spec, so the fault is purely with Intel this round.

AKA "Welcome to the experience Dell buyers already have."
Ha!
To be fair, at least Dell does continue to support their products with drivers and BIOS/UEFI firmware years after the product has gone end of life, so credit to them for that.

Performance standards, especially with single-channel memory though, agreed.
 
I don't understand the problem. The way I see it, a motherboard maker has the option to sell a bargain basement board if they want by enforcing the power limit and cheaping out on the VRM. You get exactly the performance advertised and no more. On the other hand, if you want it to be almost as fast as a K-series, then you can spend a little more and get a better board. What's wrong with having the option? Is it too hard to look this stuff up before you buy a motherboard?
 
I don't understand the problem. The way I see it, a motherboard maker has the option to sell a bargain basement board if they want by enforcing the power limit and cheaping out on the VRM. You get exactly the performance advertised and no more. On the other hand, if you want it to be almost as fast as a K-series, then you can spend a little more and get a better board. What's wrong with having the option? Is it too hard to look this stuff up before you buy a motherboard?
They go over this in the article. Right now I can pick 3600 or whatever and put it into any MB bargain or otherwise and more or less aside from the typical manufacturer settings get performance within a few percentage points of what was interviewed.

The article goes over a specific motherboard chipset, if you were to pair it with any chip that you see reviewed it would instantly lead to 20 some percent less performance. That's the difference.
 
I don't understand the problem. The way I see it, a motherboard maker has the option to sell a bargain basement board if they want by enforcing the power limit and cheaping out on the VRM. You get exactly the performance advertised and no more. On the other hand, if you want it to be almost as fast as a K-series, then you can spend a little more and get a better board. What's wrong with having the option? Is it too hard to look this stuff up before you buy a motherboard?
The problem is 99% of the people look at day 1 reviews, and expect that performance they see. Not knowing if they go and purchase the cheapest motherboard they will not get that same performance, and thats even with a locked CPU.

Even in their article they state that some B560 motherboards act like Z590 motherboards and run the CPU's out of spec. But, not all B560 motherboards act the same way. So, 99% of the un-informed users won't this and get boned in the long run. Then you see posts popup everywhere on forums asking why their brand new Intel CPU isn't running properly, when they are actually running within spec....

This is why it is a disaster.
 
The problem is 99% of the people look at day 1 reviews, and expect that performance they see. Not knowing if they go and purchase the cheapest motherboard they will not get that same performance, and thats even with a locked CPU.

Even in their article they state that some B560 motherboards act like Z590 motherboards and run the CPU's out of spec. But, not all B560 motherboards act the same way. So, 99% of the un-informed users won't this and get boned in the long run. Then you see posts popup everywhere on forums asking why their brand new Intel CPU isn't running properly, when they are actually running within spec....

This is why it is a disaster.

In fairness, an uninformed user probably isn't going to notice the missing "30% performance" either.
 
They go over this in the article. Right now I can pick 3600 or whatever and put it into any MB bargain or otherwise and more or less aside from the typical manufacturer settings get performance within a few percentage points of what was interviewed.

The article goes over a specific motherboard chipset, if you were to pair it with any chip that you see reviewed it would instantly lead to 20 some percent less performance. That's the difference.

The problem is 99% of the people look at day 1 reviews, and expect that performance they see. Not knowing if they go and purchase the cheapest motherboard they will not get that same performance, and thats even with a locked CPU.

Even in their article they state that some B560 motherboards act like Z590 motherboards and run the CPU's out of spec. But, not all B560 motherboards act the same way. So, 99% of the un-informed users won't this and get boned in the long run. Then you see posts popup everywhere on forums asking why their brand new Intel CPU isn't running properly, when they are actually running within spec....

This is why it is a disaster.

So the problem is that Intel released processors that are not as locked down as Xeons, but not as not-locked-down as K-series, and people are confused because they don't bother to fully understand the hardware before they buy it.

I don't think I'm alone on this one, but before I choose a motherboard, I research the shit out of it. Besides reading specs and reviews, I spend a lot of time reading forum threads from actual owners who are using it in the same ways that I actually will. I find out if there are compatibility issues, stability issues, how it overclocks, what RAM it likes, what BIOS settings it likes, etc. I do that BEFORE I buy it, and I do it for several different competing boards as well. It's always better to learn from other people's mistakes than to make them yourself. I have no sympathy for people who put in less effort and then cry about not getting a level of performance they mistakenly believe they are entitled to.
 
So the problem is that Intel released processors that are not as locked down as Xeons, but not as not-locked-down as K-series, and people are confused because they don't bother to fully understand the hardware before they buy it.

I don't think I'm alone on this one, but before I choose a motherboard, I research the shit out of it. Besides reading specs and reviews, I spend a lot of time reading forum threads from actual owners who are using it in the same ways that I actually will. I find out if there are compatibility issues, stability issues, how it overclocks, what RAM it likes, what BIOS settings it likes, etc. I do that BEFORE I buy it, and I do it for several different competing boards as well. It's always better to learn from other people's mistakes than to make them yourself. I have no sympathy for people who put in less effort and then cry about not getting a level of performance they mistakenly believe they are entitled to.

Your experience isn't that of the average person either. The fact that you know what to research at all puts you in the top 2-3% of knowledgeable computer purchasers. A lot of people are just buying something off the rack so their kids can do school from home.
 
You want to laugh, some of the things I looked for in motherboards are things like VGA, DVI output and optical output. I didn't see a board with VGA/DVI outputs and optical output at the same time so I bought the AsRock B560M-HDV. (Oddly enough, I want to build shelves for a while before I build my computer. Imagine just having an i5-11500 etc. just sitting there for a while I'm doing something else...)
 
Your experience isn't that of the average person either. The fact that you know what to research at all puts you in the top 2-3% of knowledgeable computer purchasers. A lot of people are just buying something off the rack so their kids can do school from home.

Good point. Like you said in your other post, the very casual consumer will never even know whether or not they have an unlocked TDP. I think the problem is the half-informed consumer. Half of them will get the slow motherboard and complain that it's too slow in Cinebench or whatever and will ask if they should try to RMA the chip, and the other half will get the fast motherboard and complain that their CPU with its stock fan is hitting 100C in some program they don't actually use for anything and they will also ask if they should RMA the chip.
 
TLDR, we bought a 65 watt locked CPU and paired it with a bottom of the barrel budget boards and it only performs to the specs listed without having to go in and manually tweak things.

Hardware unboxed is like a bitter ex. that won't go away towards Intel and Nvidia. :ROFLMAO:
 
More on this topic. Their point is that these boards ARE advertising 125w support but don't officially support it. Now it does appear that some really do while others really don't.
 
Last edited:
More on this topic. Their point is that these boards ARE advertising 125w support but don't officially support it. Now it does appear that some really do while others really don't.


In fairness that looks like a problem with the board manufacturer not necessarily Intel at this point. ASRock can be kind of shady at times compared to top tier manufacturers. I've had good luck with them though for most things. Their budget boards I would stay away from though.
 
In fairness that looks like a problem with the board manufacturer not necessarily Intel at this point. ASRock can be kind of shady at times compared to top tier manufacturers. I've had good luck with them though for most things. Their budget boards I would stay away from though.
This... their top end boards are usually innovative and quite good, but their cheap stuff is... cheap
 
In fairness that looks like a problem with the board manufacturer not necessarily Intel at this point. ASRock can be kind of shady at times compared to top tier manufacturers. I've had good luck with them though for most things. Their budget boards I would stay away from though.
Intel can set the language in their supplier agreements to force the manufacturer's hand, 100%. If Intel wanted to fix this, it could. Question is, does it want to?
 
Intel can set the language in their supplier agreements to force the manufacturer's hand, 100%. If Intel wanted to fix this, it could. Question is, does it want to?
More to the point, do we want them to? If Intel "fixes" it they're most likely to insist on 100% adherence to the power management spec, meaning all mobos will operate in the slow mode.
 
Of all the "internet-isms" this one bugs me the most. You don't want to read but you still want to comment. RME
You failed to understand what TLDR means.
He provided a summary of the article for those that found it Too Long and Didnt Read.
It doesnt mean he didnt read the article, your dig at him is unwarranted.
 
In fairness that looks like a problem with the board manufacturer not necessarily Intel at this point. ASRock can be kind of shady at times compared to top tier manufacturers. I've had good luck with them though for most things. Their budget boards I would stay away from though.
Yes and no. There's looseness in the spec that allows for these variations. We aren't talking about buying the Abit board for 10-15% more performance here. We are looking at 50% differences.

So something that I often wanted to see I'm seeing now and that is great variation between motherboards. I only wanted to see it for excitement. But that doesn't necessarily mean that it's good.

Basically if you're buying one of these boards you need to pay extra special attention to which one you're buying.
 
My question is why did Intel remove the raid ROM from the b560 when it was on the B460? Every Intel board going back 10+ years I own had RAID available.
 
My question is why did Intel remove the raid ROM from the b560 when it was on the B460? Every Intel board going back 10+ years I own had RAID available.
If not for maximally cynical reasons, then because it pushed the ROM size above a chip threshold and would've added an extra $0.50 to $1/board to the BOM. Running out of space in the bios flash when adding support for next years CPUs has been a major factor in BIOS updates dropping features a year after release. Recently this has been more of an issue on the AMD side because their sockets are longer lived; but I could see Intel deciding to head the problem off by dropping features in advance. Just like they tick the socket every other year to avoid having to try and cram 3 generations of config into a min-spec bios chip.
 
If not for maximally cynical reasons, then because it pushed the ROM size above a chip threshold and would've added an extra $0.50 to $1/board to the BOM. Running out of space in the bios flash when adding support for next years CPUs has been a major factor in BIOS updates dropping features a year after release. Recently this has been more of an issue on the AMD side because their sockets are longer lived; but I could see Intel deciding to head the problem off by dropping features in advance. Just like they tick the socket every other year to avoid having to try and cram 3 generations of config into a min-spec bios chip.

But with Intel, there's hardly any supported CPUs (relatively speaking). There should be PLENTY of room (compared to AMD). Plus, the B460 board supports the same processors that the B560 would support.
 
Back
Top