Linus Torvalds On The Importance Of ECC RAM, Calls Out Intel's "Bad Policies" Over ECC

[schmidtbag]

I don't disagree with Linus, but I feel he is overreacting here. Intel offers ECC even on relatively low-end chips. Take this for example:

https://www.newegg.com/intel-xeon-e-2136-lga-1151/p/N82E16819118007?&quicklink=true

That's a socket 1151 6c/12t CPU with ECC support. It's a bit overpriced for what it is, but if ECC is that important to you then speak with your wallet: pay the extra price or go to AMD.
Sure, it sucks that they don't offer ECC on consumer-grade CPUs, but the vast majority of people don't care.

[vladpetric]

Can Intel just die please?

Competition is still a good thing ...

[CommunityMember]

Sure, it sucks that they don't offer ECC on consumer-grade CPUs, but the vast majority of people don't care.

I believe the issue is not that most people don't care, but they are not aware they should care or need to make a choice.

When it comes to such purchases, some people (a lot of people, actually) want to go to Wally World (or BestBuy) and get a cheap, but serviceable tool, and some people want the best, no matter what the price.

One tends to choose ECC only when one is spending OPM, and not so much when you have to choose to spend your own money on your families devices (unless your income is like high 6/7 figures like Linus's).

Of course, Linus is speaking only to the choir about the value of things like ECC. One should certainly ask him whether his laptop, and his phone, and all those he has bought for his family has ECC, or whether he has chosen "good enough" in that case, which is, for better or worse, where most people tend to end up (I am only aware of a handful of laptop vendors that offer ECC, and while I suspect there is some specific phone with ECC, that is not the norm unless your have a .gov at the end of your email address).

[schmidtbag]

I believe the issue is not that most people don't care, but they are not aware they should care or need to make a choice.

This is one of those things where if you don't know whether you need it or even know what it is, you very likely (but not assuredly) don't need it. You don't need ECC on a family PC. You don't need ECC for a gaming PC. You don't need ECC for a home media center. You don't need ECC for an office PC that just runs a web browser and MS Office all day. Bit flipping is a very real and dangerous problem but it's not enough of a threat to the average user. If it were, either all RAM would be ECC or all CPUs would support ECC.
Of course, you'd be an imbecile if you built a server or high-end workstation without ECC.

Of course, Linus is speaking only to the choir about the value of things like ECC. One should certainly ask him whether his laptop, and his phone, and all those he has bought for his family has ECC, or whether he has chosen "good enough" in that case, which is, for better or worse, where most people tend to end up (I am only aware of a handful of laptop vendors that offer ECC, and while I suspect there is some specific phone with ECC, that is not the norm unless your have a .gov at the end of your email address).

Is he, though? Because the choir are using Xeons and Threadrippers/Epycs. Like I said, you can get affordable hardware with ECC support.
Maybe there's a little more truth when it comes to laptops, though I figure most people who are doing something critical enough to warrant ECC on a laptop are running most of their tasks on a server.

[bug77]

Here's another way to look at it: ECC may be good, but AMD needs it more. Because Zen needs faster memory to really shine and faster memory is more error-prone. Plus, Zen just seems more picky about RAM choice in general.

On a more practical note, ECC means higher latency. Whoever moves their consumer chips to ECC first, will be slaughtered in benchmarks.
Also, as noted above, it's pretty pointless to improve the reliability of your RAM, when your software is 1,000x more likely to blue screen/core dump anyway.

I was with Linus when we "talked" about AVX, but I think he's not seeing/considering the whole picture here.

[Zan Lynx]

This is one of those things where if you don't know whether you need it or even know what it is, you very likely (but not assuredly) don't need it. You don't need ECC on a family PC. You don't need ECC for a gaming PC. You don't need ECC for a home media center. You don't need ECC for an office PC that just runs a web browser and MS Office all day. Bit flipping is a very real and dangerous problem but it's not enough of a threat to the average user. If it were, either all RAM would be ECC or all CPUs would support ECC.

People need it even if they don't know it.

I've helped out friends and family who have corrupted documents. Was it their drive? I guess they should have been running btrfs or ZFS. They didn't know it, but they needed it. Or it could have been a RAM error while copying or saving the file. However, since their computer didn't have any error correction there is no way to know what error correction would have helped.

Nice Catch-22 there isn't it. Without ECC you don't know if you needed ECC.

[Zan Lynx]

On a more practical note, ECC means higher latency. Whoever moves their consumer chips to ECC first, will be slaughtered in benchmarks.

This has to be old information. I am running with ECC RAM on this Ryzen 5950 and its latency stats are right in line with any other 2,666 MHz CL19 RAM. In other words, it isn't great, but that has nothing to do with the presence of ECC. It's because the RAM manufacturers don't bother to produce high speed ECC. If they did, it would be exactly the same.

AMD (and Intel I assume) build the RAM controller with integrated things like ECC, AES encryption, PCIe and cross-CPU (for EPYC) routing. All of this runs inline at full RAM speed. At the transistor level this is not hard.

[zxy_thf]

I know where I can buy Xeon CPUs, but please show me where I can buy Ryzen Pros. There is absolutely a difference, here! And you could argue that Intel is actually the better of the two, because they will sell you a Xeon if you want it, but you can only get a (used) Ryzen Pro APU on ebay or in a prebuilt corporate-oriented PC with specs that might be a poor fit for your needs.

Threadripper CPUs have ECC hardware, although they're more expensive than the combo of Xeon E3 (Now it's called Xeon E I think?) and a motherboard.

[schmidtbag]

People need it even if they don't know it.

Again, if that were true, it would be the standard. You wouldn't have an option.

I've helped out friends and family who have corrupted documents. Was it their drive? I guess they should have been running btrfs or ZFS. They didn't know it, but they needed it. Or it could have been a RAM error while copying or saving the file. However, since their computer didn't have any error correction there is no way to know what error correction would have helped.

I'm far more confident that was due to drive failure and not a flipped bit in RAM. That you can actually prove by doing drive integrity tests.

[qarium]

i buy AMD for ECC alone for years.
i had 32GB ECC ram for my FX8320
and i will soon buy 128gb ddr4 ECC ram for my Threadripper 1920X...

yes AMD also build cpu in the APU market without ECC but the situation with AMD is much better than intel.

today AMD is faster than intel means just avoid intel at any cost and buy ECC ram for it.

companies like intel only chance their mind if no one buy intel anymore.