4 ms·

This is by far the most informative article coming from The Register, I'd had expected this article coming from AnandTech or similar news outlets. Back in last

This is by far the most informative article coming from The Register, I'd had expected this article coming from AnandTech or similar news outlets. Back in last year 2021, Linus was asked what is the most special about Linux in an interview, and he pointed to the Linux File System as the fastest in the business [1]. Now imagine an OS (Linux or not) having a first class in-memory file system, and it will blows away any conventional OS file-system, and Optane can provide this possibility at an affordable cost. I know that you can get several Terabytes of RAM with Xeon CPUs now, but at the moment the cost will be prohibitive. The above mentioned feature has a killer application namely non-linear time-frequency analysis [2]. This quadratic complexity algorithm need to have huge amount of working memory in order to work effectively. I have a paper on this topic applied to biomedical field and it is currently trending on top of the page in Google search for several years now due to the highest accuracy, specificity and sensitivity of more than 99%. Apparently it's a sleeper paper because other researchers would avoid citing the article due to its very high accuracy using commonly available datasets hence lower ranking (still high but not top or front page) on the subject in Google Scholar search. During a meeting with Micron engineers earlier this year I have conveyed my condolence to them. [1]https://www.tag1consulting.com/blog/interview-linus-torvalds-linux-and-git https://www.tag1consulting.com/blog/interview-linus-torvalds... [2]https://en.m.wikipedia.org/wiki/Bilinear_time%E2%80%93frequency_distribution https://en.m.wikipedia.org/wiki/Bilinear_time%E2%80%93freque...

lproven 4y ago
> This is by far the most informative article coming from The Register, I'd had expected this article coming from AnandTech or similar news outlets. Thank you. :-) I am trying to do some more in-depth analytical pieces for the site, and I am delighted that several of them have done very well -- not only on HN.
twoodfin 4y ago
The second part of your comment hints at the trouble with your first: Optane didn’t provide major benefits over existing storage architectures when used for a block-addressable file system. It had significant density benefits when used as RAM, but at an equally significant latency penalty. To really see wins, you needed an application data model that could benefit from the high density/single node capacity, byte-addressability, and persistence. Basically an in-memory database, either general purpose or embedded in a specialized data processing application. Those are premium, niche markets, and could never support the demand Intel needed to make commercial sense.
- lproven 4y ago
  > Those are premium, niche markets, and could never support the demand Intel needed to make commercial sense. That's exactly the sort of thing that people said about the Mac in the mid-1980s, and Windows in the early 1990s, and so on. :-)

lproven 4y ago

> This is by far the most informative article coming from The Register, I'd had expected this article coming from AnandTech or similar news outlets. Thank you. :-) I am trying to do some more in-depth analytical pieces for the site, and I am delighted that several of them have done very well -- not only on HN.

twoodfin 4y ago

The second part of your comment hints at the trouble with your first: Optane didn’t provide major benefits over existing storage architectures when used for a block-addressable file system. It had significant density benefits when used as RAM, but at an equally significant latency penalty. To really see wins, you needed an application data model that could benefit from the high density/single node capacity, byte-addressability, and persistence. Basically an in-memory database, either general purpose or embedded in a specialized data processing application. Those are premium, niche markets, and could never support the demand Intel needed to make commercial sense.

lproven 4y ago
> Those are premium, niche markets, and could never support the demand Intel needed to make commercial sense. That's exactly the sort of thing that people said about the Mac in the mid-1980s, and Windows in the early 1990s, and so on. :-)