Apple buys the premium (middle) of all wafers which is more expensive. Apple memory has always been more reliable.
This guy appears to have schooled a whole bunch of you. His point exists in isolation from the notion of needless overcharging and gives real world context to all the "it's fine" or "just to be safe" hand-waving. What he said should most certainly trump any discussions of cost until he is adequately (in your opinion) refuted.
If you're convinced that, for example, buying OWC RAM compromises nothing, or there's no reason you can't otherwise just circumvent Apple and buy RAM that has the same specs, that's fine; simply be prepared to hold your own in discussions about wafers and the wafer supply chain.
That being said, there appears to have been very little meaningful conversation on this topic since he raised the stakes. Either defer to the knowledge of such people, ask them to elaborate, or question how they came to know what they claim to know.
I'll do it right now, I swear! How do you know? In disseminating this fact are you violating a current NDA? Or, and I believe this to be of equal importance, are you "violating" the terms of an expired NDA?
I can fling that back at you. How do you know?
i can say with a lot of experience behind it that OWC memory, among others, compromises nothing.
Apple buys the premium (middle) of all wafers which is more expensive. Apple memory has always been more reliable.
This guy appears to have schooled a whole bunch of you. His point exists in isolation from the notion of needless overcharging and gives real world context to all the "it's fine" or "just to be safe" hand-waving. What he said should most certainly trump any discussions of cost until he is adequately (in your opinion) refuted.
If you're convinced that, for example, buying OWC RAM compromises nothing, or there's no reason you can't otherwise just circumvent Apple and buy RAM that has the same specs, that's fine; simply be prepared to hold your own in discussions about wafers and the wafer supply chain.
That being said, there appears to have been very little meaningful conversation on this topic since he raised the stakes. Either defer to the knowledge of such people, ask them to elaborate, or question how they came to know what they claim to know.
I'll do it right now, I swear! How do you know? In disseminating this fact are you violating a current NDA? Or, and I believe this to be of equal importance, are you "violating" the terms of an expired NDA?
I can fling that back at you. How do you know?
i can say with a lot of experience behind it that OWC memory, among others, compromises nothing.
I know what you mean. I made a concerted effort to not appear to be taking a stance one way or another.
Yes, I have myself witnessed and investigated persistent issues that were later confirmed to be caused by subpar wafers. Without (i.e. before) such confirmation, they can be issues that really throw a wrench into a hardware engineer's grasp on reality. And yes, I would be violating an NDA to discuss it much further.
I suppose my overall point was that I believe such wafer-related knowledge SHOULD be common knowledge; we (including you) are more than capable of comprehending the details of how and why deviations in wafer quality can lead to system issues. There are apparently very real reasons and explanations for it, but the prevalence of NDAs and proprietary knowledge can prevent such details from ever leaving company control. This essentially FORCES otherwise eager-minded people to just kind of 'black box' the process, and not truly get it. Myself included. It also forces company after company to slowly and redundantly acquire the same knowledge. This keeps innovation at a near-standstill, relatively speaking.
And I haven't discounted the fact that OWC may very well have the same 'premium' purchasing standards as Apple, but no one has said that yet.
Granted, I can always do further research to determine exactly how much of the aforementioned knowledge is indeed freely and legally available. But it's frustrating enough that I can't just rattle off my experience in order to contribute to the conversation.
I kind of like to think that that guy WAS breaking some rule in telling us that. But then at the same time that would make me feel like a bit of a pussy.
Apple buys the premium (middle) of all wafers which is more expensive. Apple memory has always been more reliable.
This guy appears to have schooled a whole bunch of you. His point exists in isolation from the notion of needless overcharging and gives real world context to all the "it's fine" or "just to be safe" hand-waving. What he said should most certainly trump any discussions of cost until he is adequately (in your opinion) refuted.
If you're convinced that, for example, buying OWC RAM compromises nothing, or there's no reason you can't otherwise just circumvent Apple and buy RAM that has the same specs, that's fine; simply be prepared to hold your own in discussions about wafers and the wafer supply chain.
That being said, there appears to have been very little meaningful conversation on this topic since he raised the stakes. Either defer to the knowledge of such people, ask them to elaborate, or question how they came to know what they claim to know.
I'll do it right now, I swear! How do you know? In disseminating this fact are you violating a current NDA? Or, and I believe this to be of equal importance, are you "violating" the terms of an expired NDA?
I can fling that back at you. How do you know?
i can say with a lot of experience behind it that OWC memory, among others, compromises nothing.
I know what you mean. I made a concerted effort to not appear to be taking a stance one way or another.
Yes, I have myself witnessed and investigated persistent issues that were later confirmed to be caused by subpar wafers. Without (i.e. before) such confirmation, they can be issues that really throw a wrench into a hardware engineer's grasp on reality. And yes, I would be violating an NDA to discuss it much further.
I suppose my overall point was that I believe such wafer-related knowledge SHOULD be common knowledge; we (including you) are more than capable of comprehending the details of how and why deviations in wafer quality can lead to system issues. There are apparently very real reasons and explanations for it, but the prevalence of NDAs and proprietary knowledge can prevent such details from ever leaving company control. This essentially FORCES otherwise eager-minded people to just kind of 'black box' the process, and not truly get it. Myself included. It also forces company after company to slowly and redundantly acquire the same knowledge. This keeps innovation at a near-standstill, relatively speaking.
And I haven't discounted the fact that OWC may very well have the same 'premium' purchasing standards as Apple, but no one has said that yet.
Granted, I can always do further research to determine exactly how much of the aforementioned knowledge is indeed freely and legally available. But it's frustrating enough that I can't just rattle off my experience in order to contribute to the conversation.
I kind of like to think that that guy WAS breaking some rule in telling us that. But then at the same time that would make me feel like a bit of a pussy.
Yes, there are different bins in memory chips as well as with CPUs. But the bins are speed, and sometimes, latency related rather than reliability related. Nobody wants to release unreliable chips.
i have no doubt that if a company has thousands of computers with hundreds of terabytes of RAM, they will get the occasional bad chip. Bit memory sticks are tested before they leave the factory. The testing is pretty sophisticated, and while that testing can’t test every cell, because that would take too long, it weeds out almost every bad stick. As I said, I’ve installed a vast amount of memory over the decades, including soldering memory in, and the only bad stick I got was back in 1992 from Apple, for my Quadra 950.
Comments
i can say with a lot of experience behind it that OWC memory, among others, compromises nothing.
Yes, I have myself witnessed and investigated persistent issues that were later confirmed to be caused by subpar wafers. Without (i.e. before) such confirmation, they can be issues that really throw a wrench into a hardware engineer's grasp on reality. And yes, I would be violating an NDA to discuss it much further.
I suppose my overall point was that I believe such wafer-related knowledge SHOULD be common knowledge; we (including you) are more than capable of comprehending the details of how and why deviations in wafer quality can lead to system issues. There are apparently very real reasons and explanations for it, but the prevalence of NDAs and proprietary knowledge can prevent such details from ever leaving company control. This essentially FORCES otherwise eager-minded people to just kind of 'black box' the process, and not truly get it. Myself included. It also forces company after company to slowly and redundantly acquire the same knowledge. This keeps innovation at a near-standstill, relatively speaking.
And I haven't discounted the fact that OWC may very well have the same 'premium' purchasing standards as Apple, but no one has said that yet.
Granted, I can always do further research to determine exactly how much of the aforementioned knowledge is indeed freely and legally available. But it's frustrating enough that I can't just rattle off my experience in order to contribute to the conversation.
I kind of like to think that that guy WAS breaking some rule in telling us that. But then at the same time that would make me feel like a bit of a pussy.
i have no doubt that if a company has thousands of computers with hundreds of terabytes of RAM, they will get the occasional bad chip. Bit memory sticks are tested before they leave the factory. The testing is pretty sophisticated, and while that testing can’t test every cell, because that would take too long, it weeds out almost every bad stick. As I said, I’ve installed a vast amount of memory over the decades, including soldering memory in, and the only bad stick I got was back in 1992 from Apple, for my Quadra 950.
i suppose I’m somewhat lucky too.