9 ms·
"Some workstation users will need ECC memory, and up to 512GB of it. When memory has an error rate of 1 error per GB per year, using 512GB ensures almost two bi
by jonbaer 9y ago
"Some workstation users will need ECC memory, and up to 512GB of it. When memory has an error rate of 1 error per GB per year, using 512GB ensures almost two bit errors per day: something that a 60-day simulation would find catastrophic."
- theyregreat 9y agoBitsquatting, rowhammer and silent corruption. Basically, ECC should be the universal standard in all network gear, servers and endpoint (user) devices. Not creates an attack surface and opportunity for data-loss. https://cr.yp.to/hardware/ecc.html https://cr.yp.to/hardware/ecc.html
- cbetz 9y agoI had the same thought and am wondering about the economic reality of doing this. Does anyone have an idea of the cost increase for using ECC RAM in a mobile device for instance? Are there also power usage concerns to worry about? AFAIK there are basically zero consumer devices using ECC RAM.
- barrkel 9y agoECC is a market segmentation tool. Keeping ECC away from consumer hardware means you can charge more to make it available for industrial, professional etc. use.
- ericfrederich 9y agoInteresting thought, but kind of sounds like a conspiracy theory. Anything else to back it up?
- skummetmaelk 9y agoObviously ECC requires more resources and is therefore intrinsically more expensive. That does not preclude it from being used for market segmentation to get higher margins. Same reason ECC is not supported on consumer processors from Intel. Here it would be almost trivial to add, but they don't. Simple business strategy.
- nirvdrum 9y agoYou're probably aware, but the desktop i3 does support ECC. Well, at least through the 7th generation. It looks like now that it's added 2 cores in the 8th generation, ECC has been dropped (NB: I didn't look at all models). So, they're okay with it on consumer parts, as long as you're not looking to do anything resembling professional work I guess.
- sitkack 9y ago> ECC requires more resources and is therefore intrinsically more expensive manufacturing cost is non-linear. Obviously intrinsically word gunk.
- skummetmaelk 9y agoCan you expand on this? Memory already includes controller logic and adding ECC logic does not need a seperate process. All you need is more silicon real estate because of the extra bits needed for the ECC. Silicon is cheap.
- dom0 9y agoIn most previous generations all Core iX dice have had support for ECC, selectively disabled through fusing. Obviously, all internal busses and caches use error correction/detection independent of what the memory controller does.
- Dylan16807 9y ago> Obviously ECC requires more resources and is therefore intrinsically more expensive. Technically. A fraction of a percent increase in die size for the memory transceivers, basically nothing for making and verifying checksums. It's all business strategy.
- Bromskloss 9y agoDo you see every market segmentation as a conspiracy? If so, a conspiracy between whom?
- tormeh 9y agoNot really. Most people don't care about ECC and gamers actually prefer the faster non-ECC ram. Since ECC is more expensive to make, the industry is obviously not going to sell ECC to people not willing to pay more for it. And in a lucky coincidence it turns out that those who want ECC are willing to pay the moon for it, which of course gatekeepers like Intel exploit.
- ticviking 9y agoI'd take a slight slowdown in ram access if I knew for sure that my game won't bluescreen.
- oldsklgdfth 9y agoYour point seems to make sense in GPU lands as well. Nvidia offers ECC on their workstation class GPUs - quadro - but not on their gaming GPU - gtx.
- dom0 9y agoGaming GPUs have different qualification targets anyway. They will error fairly frequently compared to e.g. your CPU; something that doesn't matter when most errors become invisible after about 16 ms, but not really something you want to see e.g. in a simulation.
- DCKing 9y agoWell I don't think there's anything about ECC, or at least unbuffered ECC, that should make memory slower than non-ECC memory. "Fast" memory it seems to me is also just a marketing thing. If ECC was considered mainstream as it should be then I think we'd have overpriced '4000MHz Gaming X Raptor' ECC modules all the same. I'm also not convinced that unbuffered ECC makes memory modules significantly more expensive. Basically there's no evidence that ECC memory is signficantly slower or more expensive than non-ECC memory. Obviously it's slightly more complex than non-ECC memory, but it's not a particularly high-tech addition. For all its supposed complexity/slowness, Samsung is using it in its caches on their Exynos SoCs. Everything can be perfectly explained in terms of market segmentation.
- KGIII 9y agoMany and CPUs support ECC. It will take a bit longer to find a motherboard that supports them, but they are out there. Consumer ECC is an option but few seem to know about or opt for it. I want to say I had ECC in my old AMD K6-350, but that was a lot of years ago. Either way, I'm pretty sure they have supported it for a long time and in many of their CPUs, even the CPUs aimed at consumers. Here's a comment I made last night, if you want to dig in and find some current offerings: https://news.ycombinator.com/item?id=15406565 https://news.ycombinator.com/item?id=15406565
- ju-st 9y agohttps://www.eetimes.com/author.asp?section_id=36&doc_id=1325816 https://www.eetimes.com/author.asp?section_id=36&doc_id=1325...
- jandrese 9y agoTheoretically going with ECC Memory should add about 20% to your memory budget, but due to artificial chipset limitations and "enterprise pricing" it is often more like 500%. Personally, I've always found this to be very annoying. The only people using non-ECC memory should be Xtreme Gamer type people who don't care about their data and only care about squeezing that last 2% out of their system to land at the top of the penis size chart. Non-ECC should be a feature like liquid N2 cooling.
- ensignavenger 9y agoCurrent market prices for ECC RAM are approximately twice as much. I just built anew Ryzen system and considered using ECC, but I couldn't find any DDR4 3000 RAM, and the DDR 4 2666 ECC Ram was just under twice as much as the DDR4 3000 non ECC I ended up with. Opportunity for a future upgrade , though. (My motherboard does support ECC :) The real question though for a large scale consumer product manufacturer is how much margin they could convince the RAM makers to give up on ECC. Once it became standard, it probably would not cost much more than non ECC. Maybe 15% more? (1/8th extra memory, plus some extra for other expenses.)
- flyinghamster 9y agoI found that, at least for DDR4-2400, the ECC price premium is about $50 per 16 GB stick. I sprang for it anyway, but still winced at the cash register racking up the numbers for my own Ryzen upgrade. It is nice to see "EDAC amd64: DRAM ECC enabled." in my dmesg output. :) That, and the NVMe SSD is screaming fast. I'm debating whether to try my hand at overclocking my R7-1700, but even leaving it at the stock clocking it's pretty speedy, and it runs cool with the stock cooler.
- dom0 9y agoECC is often used without being advertised. For example, hard drives aren't advertised as using these kinds of codes, yet all of them do. Similarly, switches will often use ECC memory without making it an explicit point in the advertising. Industrial computers tend to have ECC, yet you will have to look closely at the data sheet to see this. And of course all internal busses and caches in pretty much any desktop/server/workstation CPU use error correction/detection. The moral of this story is that ECC is only ever advertised explicitly, when it is special to have, i.e. because there are segments in the market where it is not used. (A somewhat curious case are small microcontrollers. Often these control relatively important aspects of machinery, for example, drives [say a bit-flip in the drive controller turns STOP into RUN, which can get rather ugly on e.g. a lathe]; these usually don't have ECC. However, they are manufactured in very coarse processes and generally don't use DRAM.)
- mafsedan 9y agoExactly. Any CPU for real work that can't support ECC memory is useless.
- robin_reala 9y agoDoes anyone have any info as to which study lead to the 1 error per Gb per year statistic?
- keldaris 9y agoJust to offer a partial counterpoint to the popular "ECC is a must for real work" sentiment, I've ran many 60-day simulations where a few bit errors would be completely irrelevant. Many physics simulations include internal checks (that must be there regardless of ECC, for algorithmic reasons) that would catch virtually all such issues. While a bit error could certainly crash the simulation, long running simulations can be resumed from checkpoints with minimal loss of time and the probability of a crash is still very low for most real workloads. In truth, there are plenty of good reasons, both for data integrity and security, to have ECC. However, there are real costs to ECC too - 10-30% immediate performance overhead (depending on your application) and much more than that for memory-bound workloads (due to lower frequencies and subtiming issues), not to mention hardware cost efficiency. Instead of oscillating between "ECC for everything" and "ECC is useless" people should rationally compare the costs and benefits for their specific workload and choose accordingly.
- dna_polymerase 9y agoThank you! That sentiment on HN about how "Everything should have ECC" really grinds my gears. There is no need for ECC in your iPhone or on your MacBook. Your average emoji-laden WhatsApp Conversation with your friends does not require Bit Error Correction. People who really need ECC do exactly what you just mentioned: Evaluate, and often enough they will find little to gain from ECC.
- sitkack 9y ago> There is no need for ECC in your iPhone or on your MacBook. Scientists, Engineers and Economists shouldn't use a Phone or MacBook? Your handle is ironic. The lack of ECC is a method to artificially segment markets and nothing more. Not ECC, but "single bit errors" can be fatal in other domains [0] [0] https://gizmodo.com/382026/a-cellphones-missing-dot-kills-two-people-puts-three-more-in-jail https://gizmodo.com/382026/a-cellphones-missing-dot-kills-tw...
- sixbrx 9y agoI imagine bit errors are highly unwanted on any domain when they occur in pointers, especially on those that are written through. Some languages make that more likely than others of course with heavy use of references.
- semi-extrinsic 9y agoDepends on your simulation. Are you using your workstation to do molecular dynamics? Or turbulent fluid simulations? A little extra randomness in two bits every day won't hurt you. See e.g. this paper by Walker and Betz from the San Diego Supercomputing Centre (paper from XSEDE '13) on the effect of ECC on GPUs for MD simulations: https://dl.acm.org/citation.cfm?id=2484774 https://dl.acm.org/citation.cfm?id=2484774
- sitkack 9y agoSounds like we need a kernel module to introduce bit flips for bespoke errors.
- dom0 9y agoThe Linux EDAC (for ECC memory controllers) module allows injecting address-based errors (simulating the MC reporting a fault). You can also use ptrace to corrupt the memory of a process more subtly.
- Asooka 9y agoExcept if the bit error happens in a pointer and you end up reading from the wrong part of memory. Or in your file system driver and you corrupt the file system. Yes, it will most likely be in the 500GB of float arrays, but why would you risk it?
- iainmerrick 9y agoExactly. I'm amazed at how blithely people dismiss the impact of bit errors. A random bitflip could break anything! That's terrifying. Even if it's just in a float, it looks like a single bitflip could change FLOAT_MAX to NaN. That could hurt.
- cjbillington 9y agoYeah but you'll notice, and resume the simulation from a previous checkpoint or something. The risk is with errors that you don't notice, but that nonetheless change the outcome in a meaningful way. That seems to me to actually be a pretty small space of situations.