Dear Prof. Kurtsiefer, Your manuscript, referenced below, has been reviewed for Applied Physics Letters. "Random numbers from vacuum fluctuations" L16-01425R1 The reviewers' comments are included below and/or attached. In view of their recommendations, we cannot accept your paper for publication in its present form. If you revise the paper to meet the reviewers' objections, we will be happy to give it further consideration. (Please see Editor's Comments below.) Please indicate how the manuscript has been revised in a separate Response Letter file so that the editors can see whether you have complied with the reviewers' comments. Please use add file to upload the Response Letter file, and use replace for any files that have been revised or changed. Revised manuscripts must be submitted through the online submission system. They are not accepted by email. The revised manuscript should be returned to the Editor promptly. Your revision is due by 10-Jul-2016. A manuscript returned more than 60 days from today should be submitted as a new manuscript and will be given a new receipt date. Please go to the URL below to submit the revised version. To meet AIP Production requirements, please provide a separate figure file for each cited figure number (all parts in one file), in addition to your article-text file. http://apl.peerx-press.org/cgi-bin/main.plex (If clicking on the above URL address directly from your mail program is unsuccessful, please copy and paste the complete address into your browser.) Thank you for the opportunity to examine this work. Sincerely yours, Kenjiro Miyano Associate Editor, Applied Physics Letters AIP Publishing Suite 300 1305 Walt Whitman Road Melville, NY 11747-4300 USA Phone: 516-576-2344 E-mail: apl-edoffice@aip.org --------------------------------------------------------------------- Manuscript #L16-01425R1: Editor's Comments: Some comments by Reviewer #2 below are related to the form letter that we use; therefore he/she is complaining to the Editor. However, this point set aside, the reviewer's criticism is clear: since the random number generation has been around for some time, the technical advantage proposed here is not novel enough for a Letter journal. From the comments below, we have to agree with the reviewer. If the intention of your previous rebuttals were not meant to be as interpreted by the reviewer and you want to resubmit a revision, we will seek an independent opinion. Reviewer Comments: Reviewer #2 Evaluations: Does this paper meet APL's standards: No Is the paper scientifically sound with the assertions made and conclusions drawn well supported: No Is the discussion of related work and associated references adequate?: Yes Is the English satisfactory?: Yes Is the title short, interesting, and descriptive of the contents?: No Is the paper well organized and understandable?: Yes Reviewer #2 (Comments to the Author): Dear Editors of Applied Physics Letters I am responding to your letter of 7 June, in which you write: Recently you reviewed the manuscript referenced below for possible publication in Applied Physics Letters. You rejected the manuscript but the author has submitted a revised version with a rebuttal. Will you look over the revised material and return it with your recommendation as to the manuscript's suitability for publication? If you find it unsuitable for publication, please indicate what further revision is required. In my previous review, I indicated two main deficiencies of the work, and corresponding revisions that would be required for it to be suitable for publication. I quote directly from my previous report and the authors' present reply. Deficiency: The work contributes very little that has not already been demonstrated, and nothing in the area of applied physics. The only novelty in the hardware implementation is the use of a shift register + CPLD for randomness extraction. These off-the-shelf components are wired up using standard techniques, so there is no physics in the use of these devices. Proposed remedy: The title "Random numbers from vacuum fluctuations" suggests a novelty that is not present here. The title "Inexpensive randomness extraction with a shift register" would more accurately describe the contribution of the work. Response from authors: For an application, it is important to have an effective practical and fast randomness extraction scheme, and this is what we demonstrate in this work. While LFSR have been around for ages, we believe their use for this purpose makes fast randomness extraction very simple, even though much higher rates have been reported with what we believe are more complex methods. My response: I am a bit puzzled by the authors' response. The authors claim they show an effective method for randomness extraction, which is not true because they don't know if their scheme actually works (see next point). Also, the argument that using a shift register + CPLD makes the system more practical is not altogether obvious. While it simplifies the randomness source per se, I suspect that in most cases it actually makes the global system more complex. Consider that most systems that would consume high-speed random numbers, e.g. cryptography systems, will already have modern devices such as FPGAs and CPUs in them. This is certainly the trend in modern electronics, to reduce the number of special-purpose components in favor of general, re-configurable components like FPGAs. Deficiency: The authors implement an unproven method of randomness extraction and claim that it succeeds. Proposed remedy: If the authors can find a proof showing that their extraction method succeeds *for some model* of the input, and are willing to state this model as an assumption of the work, I would reverse my position on this point. Response from authors: As stated in our earlier reply, we do not have a proof yet for the LFSR method to be a "good" extractor, hence we also can not state under what assumptions it works. My response: The authors agree the effectiveness of their method is unproven. And yet in the article they still claim that it works: "with an efficient randomness extractor, we obtain an unbiased, uncorrelated stream of random bits" and "we are able to generate uniformly distributed random numbers." These claims are unsubstantiated and misleading. Since the authors have not been able to respond to the above points, my stated opinion is unchanged: the work does not contain enough novel applied physics for publication in Applied Physics Letters, and its main claim (to have generated "unbiased" and "uniformly distributed" random numbers using an "efficient" randomness extraction) is unsubstantiated. I would like to note, for the benefit of the authors, that the letter they received, beginning "The reviewers are of the opinion that it should be revised and the English needs improvement," did not accurately represent my previous referee report. I recommended for rejection, not revision, for the reasons given above. I suspect this may have led the authors to underestimate the seriousness of my objections, and the amount of revision required to make the work acceptable. ---------------------------------------------------------------------