I have the impression that all Naim’s streamers are not stable, and maybe it’s not carrying a network or electricity…I’m not sure what I’m saying it’s only a thought in my mind? I experience frozen screen from the beginning and recently I am experiencing black screen and still Naim unable to fix using many firmware they already sent…maybe this is a hardware issue and not a software problem.It’s frustrating.
Interestingly, I’ve had no screen freezes over the last month since using my NDX2 with external DAC and without the 555 PSU connected.
Do the PSUs with NDX2/ND555 exacerbate the issue?
Okey…but what`s the benefit playing without 555 PSU
My Star locked most mornings until the last update and it was the 3rd unit as the first two were exchanged by Naim. Following on from your comment I now do wonder if you are right about it being a hardware issue and that the firmware is trying to find a workaround to the issue. Maybe those more knowledgable about these things could comment as I am only guessing about this.
The CPU overclocking that was mentioned by Naim support certainly reads like hardware is involved. It’s also a thing with firmware, it’s software but so closely entangled with the hardware that separation is not necessarily meaningful - if the software does something that makes the CPU overclock, surely unexpectedly, it can be just as much a hardware bug as a software bug, or the gray are of being both
When using the NDX2 as a transport only, an external PSU adds little if any benefit. There’s a good discussion here: Is a Naim External PSU Necessary
I am sorry but the overclocking theory makes no sense. CPU frequency scaling is fully controlled by the OS. There is nothing that can “make the CPU overclock” that is not under the control of the OS. And even on Raspberry Pi devices, the CPU is typically underclocked (not overclocked) for audio applications. If a Naim streamer had to overclock the CPU to cope with load requests, this would mean that the hardware of the streamer is not up to the task which, again, makes little sense to me.
First of all it’s not so much a theory as was said several times by Naim support. This and a memory leak. Second, you might be confusing an all-purpose OS with what is more like an embedded platform
It might have been said a number of times but this doesnn’t mean that it makes more sense. A memory leak is a software error that typically can be spotted quite easily by systematic usage of debugging tools. CPU overclocking is something that, if needed in audio applications, means that the hardware is not up to the task.
I worked on an embedded platform for years, we had custom drivers for everything (well even our own kernel), it was run close to spec for various reasons (you don’t want to waste money on hardware resources you don’t need, and the tasks are reasonably well defined compared to an all-purpose machine), and if a processor had had a hardware variability that meant that some chips were not quite up to the spec, something like this would have certainly been possible. There are variations in hardware, that’s why there are i3, i5, and i7 Intels.
I won’t rule out that the info from the devs was slightly misunderstood or dumbed down for customers. E.g. whether it is really the CPU as support mentioned, or a special-purpose processor
Memory leaks can be reasonably easy to find or terribly hard. I don’t even know which language and libraries they are using. Do they have source to everything? Might be wise, but not always possible. And yes, that would be a software problem, so? The info I got from Naim was " the display freeze is related to a memory leak and or CPU overclocking so I’ve been told".
Anyway, my original point was just that when you work close to the metal the border from hardware to software becomes blurry. Complex hardware has bugs just like software. You work around them in firmware/software, it’s an everyday thing. When your software workaround does not fix it completely, is it a software bug or a hardware bug?
Anyways, none of us have seen the code, but your constant assertions of this and that definitely not making sense are less based in facts than you make them out to be.
Words to me from Naim support in an email.
“Yeah, again from what you are describing - I’m fairly confident that you are having CPU overclocking issues, if after 3.5.2 has been installed you still have these issues, please get in touch with me again and I’ll see about what can be done for you. Again, please accept my apologies for the inconvenience.”
Which constant assertions are you referring to? All I have said is that I do not buy Naim’s explanation (CPU overclocking due to software error) for the screen freeze events observed by some users. You buy it and I do not, that’s all. Time will tell. Fact is, that CPU overclocking can easily be disabled for testing purposed. Another fact is that, so far, Naim has not managed to even improve the software via firmware releases, let apart fix it. Another fact is that, according to Naim, only a very small minority of users experience freezing problems. What the hell are these users doing to trigger CPU overclocking that most users do not do?
Well in the New Firmware Update thread, Clare wrote that the different hardware theory and a Naim dealer’s claim about this is untrue, and you went and said you still think it’s true.
Then you questioned Naim’s decision about canceling 3.5.2 and going with 3.6.0 right away. Then you doubted why they would beta test a release on a larger group of unaffected users.
Basically you are making lots of strong assertions on this matter without having any idea about the facts, like we all don’t, and you are constantly “not buying” things with a tone that makes it seem as if you believe someone was hiding something.
In addition you are making claims about development in general that I simply disagree with, rightly or wrongly, but based on my experience that things often are not as simple as they seem or like they look from the outside for people missing all of the relevant information.
Your choice, sorry if it rubs me the wrong way because I am dealing with outsiders who think they know better than my company on a daily basis. I would recommend to keep a more open mind.
Like about your question, “What the hell are these users doing to trigger CPU overclocking that most users do not do?”, don’t you think that there are possible scenarios that you simply didn’t think of and that might not be obvious to someone with essentially zero real knowledge about the hardware platform and the code?
Sure, it is in conceivable that a minority of users are consistently bringing their devices through states that the majority of users never enter. Given the very long observation period, this theory seems a bit unlikely to me. But it is logically possible. If the theory is correct, “faulty” devices should then behave flawlessly when operated in one of the environments of the “majority”.
I would imagine the hardware is not suitable for the task if they have to over clock the processor?
As I said, I do not buy the overclocking theory and it is not even cear which CPU it refers to. The Naim streamer rely on a streaming board manufactured by StreamUnlimited and designed in close collaboration with Naim. It seems very unlikely that Naim have co-designed a board whose CPU is not up to the envisaged task.
If the screen freeze events are not caused by hardware defects then, as pointed out by @Suedkiez , many, many scenarios are conceivable. There doesn’t even need to be an issue with the firmware of the streamers, the events could be triggered by the control app or by certain iOS setups, who knows?
I do not think that you are fair in this judgment. If you carefully read my posts, you will realize that I am not making any “strong assertions” but merely drawing the logical consequences from what we know. Also, what I am not buying are explanations that, to me, seem unlikely or logically or technically implausible. It goes without saying that what seems unlikely to me might seem likely to others. To each his own.
I don’t think they mean that they are overclocking like a PC tinkerer on purpose running a processor at a higher clock than the spec says.
I read it not as “we are overclocking it” but “it is overclocking”, i.e., either they are running it at a clock speed where it should work stably but turns out it does not, or the clock governor is telling it to run at speed x but it runs at x+1 and they don’t know why (which both could be a bug deep in the silicone). In either case, the problem is most likely not occurring on all machines or is difficult to reproduce.
I may have and apologize if so. It may well just be a language thing or how you phrase things that seems to me to come on too strongly, given that we know very little and close to nothing of relevance. (And maybe I am not the only one, Richard also wasn’t too happy in the other thread, so maybe you do phrase a little strongly).
I reread the thing in the other thread about the hardware differences and realized that while I strongly disagreed with not believing Clare, I actually agree with your subsequent post #1615, “As a matter of fact, no matter how careful quality control is, small hardware differences (a defect batch, a change in the production flow, in component tolerences, etc.) can never be excluded with certainty.” With this clarification it doesn’t contradict Clare in fact - I believe Clare debunked the notion that Naim built “different hardware” into the NDX2 at different times, but I don’t read her posts as necessarily saying that such tolerance differences in e.g. different batches of the same hardware are impossible.
It was me that mentioned different hardwares in ndx2’s.
This came from my dealer and direct to them from Naim reps. Who ( so my dealer tells me) had been summoned because of " issues" with 2020 produced high end products.
The conversation was around supply and the quality of that supply since March.
The story was something like Naim were having quality issues from their supply chain. In particular screens and Naim illuminated logos. To the point that there were a lot of semi finished products sitting in an external ( to Salisbury) warehouse. There was also mention of " hardware issues " with other bits too.
The point being that because of supplier quality issues Naim were being compromised. In quality and production output. Semi finished products didn’t help.
This could be bunk. My dealer when challenged recently reiterated what he was told.
Make if it what you will…
I think that we should take all what comes from non-technical staff with a grain of salt. Claire and Richard are doing a great job of mediating between frustrated customers and the company but they might not know the whole story or they might be beholden not to tell it. Dealers are in a similar situations. The most reliable source of information for understanding the issue are probably those users who are plagued by screen freeze events. They are in the best situation to run systematic tests and eventually provide Naim and other users with useful observations. Time will tell whether this is a software issue, a hardware quality issue or something else and whether Naim’s software support and communication policy on this specific issue has been appropriate or not.