The performance of local drives is complicated enough, but networked storage is even worse. Just over three months ago I published a brief guide to what you should expect from several types of storage, including a little information from network-attached storage (NAS) systems. This article adds more data points to that, to help you decide which NAS to use.
Performance will vary across different NAS manufacturers and models, and figures here are based on those measured on a range of products, with a single system providing those for 10 GbE. However, I believe these should be realistic targets that all better models should be able to achieve if not exceed. All connections were made using recent versions of macOS with their respective SMB support, and no special features like iSCSI were used, to ensure these should be readily achievable during backup and file sharing.
This table compares performance of two types of local storage, hard disks and SSDs, connected via USB 3.2 Gen 2 at 10 Gb/s, with those achieved with five different NAS configurations, all using SMB. Hard disks in NAS systems were configured in RAID 1 (mirror) arrays, while those for SSDs are individual or as JBOD. Write speeds are given for:
the single 50 MB write test performed by Time Machine before each backup;
500 multiple concurrent writes of 4 KB each, performed in those same Time Machine tests;
calculated net write speed over a first full backup to APFS of at least 180 GB;
general write speed measurement using my app Stibium, which gives broadly similar results to other leading benchmarking apps.
General read speeds are also obtained using Stibium, and similar to other apps. All speeds are given as MB/s for consistency.
Storage medium
As you’d expect, hard disks are substantially slower than SSDs, although their difference isn’t as clear-cut when the effects of connection speed are taken into account. For example, overall backup speed to SSDs over 2.5 GbE was significantly slower than that to hard disks over 10 GbE.
Connection speed
The other dominant effect is that of connection speed, and the use of SMB over Ethernet. These are clearest when comparing NVMe SSD performance.
In general read/write testing, a local SSD achieves close to the maximum expected performance of 10 Gb/s at just under 1 GB/s. With 10GbE and SMB, write performance is about 60% of that, but 115% for reading. General performance of 2.5GbE is roughly a quarter of that of 10GbE and a local SSD.
Thus, overall performance is determined by a combination of medium, connection speed and type, each of which is rate-limiting. To get best performance improvement, you need to use both a fast storage medium (SSD) and fast networking (10GbE), in a NAS capable of delivering such good performance (processor, memory).
I/O Throttling
Behind all these performance figures lurks the mystery of how much they are affected by throttling by macOS. There would seem little point in spending good money for a higher-end NAS with 10GbE support, buying expensive network switches, and filling that NAS with costly SSDs, if macOS were then to throttle away those gains.
As I’ve explained, prior to changes in the documentation in 2019, I/O policy on throttling explicitly excluded remote volumes mounted through networks, but they’re now explicitly included in currently policy. However, throttling only comes into play when higher-priority I/O is competing for a share of the same disk bandwidth, and it isn’t a general restriction in transfer rate. There’s also evidence, presented there, that macOS runs first full backups at higher Quality of Service (QoS) to allow them to complete faster than scheduled automatic backups.
Unless a client Mac is simultaneously accessing a share on the same NAS or network, it therefore appears unlikely that I/O throttling should significantly reduce the speed of backups.
Recommendations
To improve performance of networked storage, particularly in Time Machine backups to a NAS:
Replace the NAS with a system designed for high performance, with processor, memory, 10GbE network connections and SSD slots capable of delivering that.
Upgrade your network to support 10GbE between clients and NAS.
Add SSDs initially to support caching for hard disk arrays, and eventually as primary backup storage.
Although suitable NAS systems will cost upwards of $/€/£ 1000, they should provide support for ample clients for years to come, and prove a worthwhile investment. Potential for improvement is greater than an order of magnitude: backups that previously had to be left running overnight, taking several hours, could take minutes, such as 180 GB in 8 minutes.
My sincere apologies – the table was omitted from the original version of this article. I have now restored that and hope it makes better sense as a result.
As the world exits the COVID-19 pandemic, more and more companies are pushing for workers to “Return to Office”. Many are also expecting full-time in-person work and dismisses remote work (or Work From Home, a problematic term that I will expand upon later) as “not real work”. There have even been instances where companies that once promised remote work will be implemented permanently turning its back on workers that have structured their life accordingly and forcing them to come back to the office instead. This is done even in companies that according to their own statistics that remote work is more productive.
Opponents of remote work often use the term “WFH”, or Work from Home to describe remote work, and it is often described as a perk. They often believe that working from the office is the only way to do real work.
This is a clear case of the Principal-Agent problem. The managers of the company are supposed to be working for the benefits of the shareholders and maximizing the profit potential. Instead, managers fall to their personal crave for the sense of control. I know someone that manages his team from Toronto that he forces to go into the New York office everyday. After all, how can they know you are doing real work unless they get to force you to commute 2 hours each way? Knowing that someone was forced to lose sleep, gain anxiety, be more stressed, is simply an irreplaceable joy that remote work can never offer. #slam_dunk_argument
Even if we ignore the Principal-Agent problem and pretend there is no personal motivation for the managers making such a decision, and it was purely made for the benefits of the business, it makes no sense.
Companies usually pay their workers something called a salary, along with possibility other perks. All of these compensation have a singular purpose, make the employee happy enough to keep doing the job. If a company can pay someone $5k a month to do the job, chances are they won’t find someone at $10k a month if they deliver the same quality of products. It is the same theme as the Murphy’s law of combat, “Remember, your weapons are made by the lowest bidder”. Considering this, allowing workers more freedom in deciding where they want to live would be an obvious way of improving their happiness. A happier worker = A more productive worker, so a manager who is forcing their team to go into a shoebox office is engaged in active sabotage against the company interest.
The auto plants of Detroit shutdown because of cheaper costs of producing in Japan. Outsourcing labor is just one of the many ways of remote work, but somehow with the advent of new technology that allows for a programmer to code from anywhere in the world, they are somehow not doing “real work” unless they go to a desk that has the same Wi-Fi connection as any other Starbucks?
As a woman, the traditional office environment can often be actively hostile. From the increased potential of physical sexual assault due literally being in the same physical location, to the air-con temperature that is often too cold for women’s comfort, it is simply a space that is not friendly, and therefore reduces the productivity. Many woman are also expected to bear household chores, and there are way more stay-at-home moms compared to dads. The inability to participate in the working world from your kitchen counter has been a huge career barrier for many women.
The gender pay gap exists for a reason, prejudice. However, I argue the solution is simple, let capitalism take over. If a woman’s work quality is the same as their male counterpart, fire the guy and hire another woman. Gender pay gap exists? Good! Exploit it!
Societal attitudes towards work changes depending on the era. When computer programming first started, newspapers pushed that women are more suited to do the job, then thought as mere clerical work, because women are more “careful”. It was only after men realized the job was important that the prejudice against female coders started and programming became a male dominated domain. This shows that societal attitudes towards work and its relationship with gender has nothing to do with objective reality.
Different societies also have different attitudes towards work. In this video, the Japanese salaried worker spends most of his days travelling across Tokyo to meet with clients face to face to resolve matters that can often be done on the phone, because Japanese culture believes face-to-face meetings to be more “polite”. He also arrived at the office 40 mins before the official start time and had work even after arriving at home after 8 pm. Japan is not known for creating the biggest startups, perhaps for a reason. After all, how much brain space do you really have for creativity after such a long day?
Japanese work culture is also known to be very prejudiced against women, who often have no real path towards career success and are often expected to marry, baby, and quit. How far can an economy go that ignores half of its highly educated population?
By not opening jobs that can done remotely to remote workers, a company ignores the entire global population, apart from wherever they happen to have an office at. Remote work is not “Work from Home”, which usually leads to the logical fallacy of “You are at home for the entire day, therefore you are not working, therefore WFH is not working, therefore remote work does not work”. Remote work is just work in a different environment, one that can be adjusted to fit the individual needs much better than a standardized office environment, one that boosts productivity, and eventually revenue.
Ignoring women means ignoring 50% of the potential talent pool, mandating in-person work means ignoring 99.99% of the potential talent pool. Remote work is simply, work. An employee of any gender is simply, an employee.
Soviet Union is dead, but capitalism has been defeated.
I shared my excitement about the PTE test result in the previous post. In this post, I will detail the English-speaking skills I learned during the three-month learning journey, which specifically meet the PTE test criteria.
PTE’s criteria
Let’s take a look at the two key criteria in the PTE’s speaking component: Pronunciation and Fluency.
Based on these criteria, I would like to highlight these key points:
Pronunciation:
Vowels and consonants;
Word stress;
Sentence-level stress;
Assimilation and deletions.
Fluency:
Rhythm and phrasing;
No hesitations.
As we can see, the PTE’s test criteria clearly show concepts we must fully understand and perfectly present if we want to achieve a higher score. In the following text, I will present my comprehension of these concepts, supported by related online resources.
Pronunciation
Vowels and consonants
In everyday conversations, sightly mispronounced words often do not significantly disrupt the flow of our discussion. But thanks to modern technology, PTE’s scoring is based on algorithms and is implemented by computers, which can easily detect each mispronunciation. Therefore, the ability to pronounce words clearly and accurately is crucial.
I have tried numerous methods to improve my pronunciation and reduce Chinese accent including speaking loudly, having more emotion, and directly imitating local accents. However, it didn’t work as expected, it did not meet my expectation, resulting in a low score in PTE practice.
Changes occurred the time I met Sun’s tutorials and BBC Learning English collection on YouTube. These pronunciation videos elaborate on vowel and consonant details, with vivid body language and emotion.
As non-native speakers who want to pronounce concisely, we must focus on these particular points:
Mouth Shape We can try to imitate the mouth shape that vowels and consonants request. For example:
When pronouncing /æ/ sounds like Agriculture and Activity, we should open our mouths as large as possible;
When pronouncing /i:/ sounds like These and Feed, the corners of our mouths should be as far apart as possible.
Tongue Position We should also pay attention to the tongue position. For example:
The /θ/ sound requires us to extend the tongue forward;
The /r/ sound requires us to pull the front part of the tongues back and up and keep the back stable, making our tongues fatter and thicker.
Breath We should also carefully control our breath, ensuring vowels and consonants are presented appropriately. For example:
When pronouncing /θ/ like Through and Thesis, we should ensure that the airflow passes through the gap between our teeth and lips;
When pronouncing /b/ and /p/ like Big and Picture, we should ensure that the airflow is completely blocked and then released suddenly.
Focusing on these points helps us ensure our study paths are on the right track, and consistently improve our pronunciation. In addition, there are two tips:
Don’t try to imitate accents. The PTE test does not expect you to have a perfect British or American accent. For beginners, attempting to achieve such an accent is time-consuming and pointless.
Never compare pronunciations between English and Chinese. For instance, the Cantonese pronunciation 士多啤梨 is often equated with Strawberry in English, which overlooks many nuances. Therefore, it is important to acknowledge that the pronunciation of English and Chinese words are completely different.
Each Chinese character has only one syllable, whereas English words typically consist of two or more syllables.
For example, the word water consists of two syllables: wa-ter, and phenomenon consists of five syllables: phe-nom-e-non. It depends on how many vowel sounds the word includes.
Furthermore, an English word consisting of two or more syllables includes both stressed syllables and unstressed syllables. These are indicated by the phonetic symbols found in dictionaries.
For example, the phonetic symbol of agri-cul-ture is /ˈæɡrɪ-kʌl-tʃər/, we can see the stress mark /'/ is placed in the first syllable /ˈæɡrɪ/, which known as a stressed syllable, while others are unstressed syllables.
An interesting rule to note is that the stressed syllable can be vary within the same word depending on its function in a sentence. For example:
When Project is acting as a noun, it is pronounced: /ˈprɑːdʒekt/
When Project is acting as a verb, it is pronounced: /prəˈdʒekt/
What should we do? Pronounce each syllable with different efforts:
Stressed syllables should be pronounced longer and louder;
Unstressed syllables should be pronounced shorter and thinner, or take place from the Schwa (we will discuss it later.)
When training pronunciation, I strongly recommend exaggerating these nuances to ensure we are on the right track and fully comprehend this concept. Eventually, it should sound natural and require less effort.
Properly presenting word stress is key to making our speech more like English, and it can significantly help in shedding “Chinglish” tendencies.
To meet this criteria, there are two concepts we should understand: Content/Grammar Words and Stressed Words.
In the English world, there are two types of words within a sentence: Content words and Grammar words.
Content words like nouns, verbs, adjectives, and adverbs play an important role in sentence structure and convey main information.
Grammar words like prepositions, auxiliary verbs, and articles are used to link content words to make a complete sentence, we can’t understand a sentence that only includes grammar words.
In the sentence “I would like to read books,” the content words are I, like, read and books, they should be pronounced clearly and accurately, while would and to are grammar words, and they should be pronounced more softly than those content words.
The second point is to decide which words should be stressed. This is an easy-to-understand concept but hard to implement when we are facing a complex sentence. I will demonstrate it through Chinese examples:
We stress some special words in our mother tongue subconsciously, furthermore, and this often influence the meaning.
For instance, we can express “There are beautiful flowers in the park.” in these different ways:
“There are beautiful flowers in the park.” We are pointing out the location;
“There are beautiful flowers in the park.” We are emphasizing the flowers, not buildings or trees;
“There are beautiful flowers in the park.” We are emphasizing the place where the beautiful flowers are located.
I suggest following general rules to avoid the potential risk of making mistakes and mispronunciation because the given text is unpredictable when we are sitting at the PTE test. Here are two steps for consideration:
1. Understanding the text. Rather than speaking without consideration and comprehension, we should first grasp what ideas the writer attempting to convey before we open our mouths. Furthermore, analyzing the elements and structure of the sentence is crucial, including subjects, verbs, objects, content words, grammar words, and clauses.
2. Marking stressed words and unstressed words. Generally, we stress one word in a phrase, choosing from a range of words, including objects, gerunds, passive verbs, adjectives, and adverbs. Below are several examples from an actual test:
Globalisation refers to a set of changes rather than a single change.
Stress what authors attempt to emphasize. In this case, the author is declaring it is a set, not a single change.
You will be introduced briefly to the discipline of child psychology.
We always stress adjectives and adverbs that modify a noun. If they are connected, stress the first one. Similarly, when facing a compound noun, we stress the first noun generally.
Although choosing the stressed words is subjective, they should be chosen from an appropriate scope that I mentioned before.
These two concepts were the most interesting part of my learning journey. They make our English speaking vivid and dynamic.
Some voices can be transformed in specific circumstances. Below are several examples:
Using the Schwa. Grammar words and articles can be reduced to a “Schwa sound”, such as “to” /tu/ becoming /tə/, and "than" /ðæn/ becoming /ðən/;
Linking vowels. When a word ends with explosive sounds like /t/, /k/, and /p/ and the following word begins with a vowel sound, we link them together, such as that element becoming tha-telement;
Inserting new sounds. When a word ends with a vowel sound and the following word begins with a vowel sound, we often insert a connecting sound. For instance: many of can be many jof.
There are lots of variations in pronunciation that we need to learn and practice. While this might feel overwhelming for some beginners, it is a vital part of speaking like a native and sounding natural. Keep learning from online resources and practice consistently until you feel comfortable.
Some voices can be dropped in specific circumstances. Below are several examples:
Combining sounds. Link the same consonant sounds that are adjacent, such as in big garden, two /g/ sounds becoming one /g:/ sound but a bit longer;
Dropping sounds. In a rapid speech, we sometimes drop explosive sounds between two consonants, such as an important role becomes an importan role;
Holding sounds. Similarly, we hold back explosive sounds without fully releasing them when these sounds between a vowel and a consonant, such as that person becomes tha-person;
Reducing /h/ sounds. When a word ends with explosive sounds and the following word begins with the /h/ sound, we always reduce the /h/ sound, such as "might have" becomes "migh-t(h)ave" and "an hour" becomes "a-nour."
These two concepts are related to the term “Thought groups.” When speaking English, we always separate the sentences into several groups by their meanings, emotions, structures or lengths. Here is an example from the real test:
Many papers you write in college / will require you to include quotes / from one or more sources.
The speaker reminisces about his views / of the English Revolution / when he was a student.
In particular, we break sentences down before prepositions such as “of”, “in” and “that.” Importantly, we should NEVER separate compound words like “the English Revolution.”
Furthermore, I suggest breaking the sentence into smaller fragments for practice, like this:
Many papers / you write in college / will require you / to include quotes / from one or more sources.
However, the PTE test would perfer a longer phrase, so I suggest that each group should have 4 to 7 words.
Now we know what is the term “Thought groups” and how to divide a sentence, the next step is to learn how to present it well. This related to the term ‘intonation and it means words pronounced in a high or low pitch accordingly and intermittently.
Intonation can bring rhythm to speaking, however, it is hard to handle and can cause trouble easily for beginners.
What should we do?
Always present a low pitch to the last word of a thought group and the sentence;
Carefully present a slightly high pitch to the adjectives or adverbs that modify a noun or a noun phrase.
No hesitations
PTE test can detect any hesitation or mispronunciation which can negatively influence our final score, especially in the Read Aloud and Repeat Sentence module.
Despite numerous challenges on test day, such as being disrupted by other test-takers or encountering unfamiliar words, I strongly recommend speaking slowly and confidently to avoid potential risks and maintain fluency.
This strategy is crucial: when we face a word or phrase that is difficult to express and may cause hesitations unavoidably, this may affect our scores in both Pronunciation and Fluency. However, if we express these challenging words slowly and confidently, maintaining a natural flow, it might primarily affect our Pronunciation score.
This is why I strongly advocate for speaking confidently, even when making mistakes.
Summary
This article discusses the knowledge I gained on my English learning journey, including methods to improve pronunciation and an understanding of the PTE speaking module criteria.
Additionally, I’ve decided to update posts in English from now on. It may contain numerous grammatical errors, awkward phrasing and word-choice issues, it’s still a necessary step forward. ‘Practice makes perfect’ is the key lesson from this journey.
當身體中的 ATP 量已經超出需求時,ATP 可以把身上的能量貢獻給合成反應,製造脂肪、肝醣、或蛋白質,儲存在脂肪細胞、肝臟、和肌肉中。
熱量盈餘能幫助增肌,是因為當攝取的營養素能提供夠多的能量時,身體就能有多餘的 ATP 去提供氨基酸合成成蛋白質的能量。如果想增肌減脂,就須由庫存的脂肪提供合成蛋白質所需的能量,但這個過程只有肥胖者或新手較容易達成。
ATP:人體的能量電池
與其說 ATP 是能量貨幣,不如說是能量「電池」。充飽電時這個電池就是 ATP,並在釋放能量後變成沒電的 ADP。
「ATP 是人體的能量貨幣」是課本最常用來描述 ATP 的一句話,代表身體需要能量時,就會使用 ATP。或許就是因為如此,讓許多人自動把 ATP 與「能量」劃上等號。但 ATP 其實是一種「高能分子」,而不是能量。ATP(adenosine triphosphate)的中文為三磷酸腺苷,顧名思義就是「有三個磷酸根的腺苷」。
你可能還是會想說:磷酸根和腺苷又是什麼東西?別擔心,你不需要搞懂詳細的化學,只要知道 ATP 的親戚們是誰就好。跟 ATP 擁有類似結構的分子,稱為「核苷酸」,也就是人體遺傳物質的構成原料。RNA 是由核糖核苷酸組成,而 DNA 則是由去氧核糖核苷酸組成。可以參考下表:
當身體需要能量時,會把 ATP 和水找來,讓它們化學反應變成 ADP 和 P。這個反應會產生能量,而身體就能將這個能量用在需要的活動上,如收縮肌肉或合成身體組織。
因此,與其說 ATP 是能量貨幣,不如說是能量「電池」。充飽電時這個電池就是 ATP,並在釋放能量後變成沒電的 ADP。講到這邊,大家應該已經可以理解,物質跟能量是不能混為一談的。
*註:我還是認為課本把 ATP 比喻為能量貨幣是合理的,因為當我們在分析生理反應時,有 ATP 確實就等同於有能量。本文是為了讓大家能分辨「能量」和「物質」的差異,才用電池做比喻。
人體也是類似的情況。身體以醣類和脂肪酸(及少量的氨基酸)作為生產 ATP 的燃料。當循環系統中的燃料不夠用時,我們需要吃東西以提供更多營養素;如果吃的食物又不夠時,就會分解肌肉及肝臟中的肝醣和脂肪細胞中的脂肪來提供醣類和脂肪酸。這些來自食物或來自身體組織的分子,就能被粒線體分解,產生 ATP,供身體使用。
舉例來說,運動會消耗大量能量,所以肌肉中的很多 ATP 會被使用並轉化成沒電的 ADP。此時,身體需要將這些 ADP 「充電」變回 ATP,才能繼續給肌肉取用。而要把 ADP 變回 ATP,就需要靠分解營養素所提供的能量,所以就得從食物或身體組織中取得營養素。若從身體的庫存中取用大量營養素,體重就會降低,這就是減重的原理,也是為何熱量消耗需要大於熱量攝取,才有可能取用大量的庫存脂肪。
體內能量很多時,會消耗 ATP 並用釋放的能量來合成身體組織
如果活動量減少或食物攝取增加,就會使體內循環的營養素增多,此時製造出來的 ATP 會高於身體所需。既然沒有什麼需要使用 ATP 的事,那就把 ATP 的能量儲存起來吧!
熱量盈餘能幫助增肌,是因為當循環中的營養素夠多時,身體就有許多的 ATP 可以運用。等這些維持生理機能、提供運動能量的需求都被滿足後,剩下的 ATP 就能幫助循環中的氨基酸合成成蛋白質,讓肌肉增大。(這也是為何除了熱量要夠外,蛋白質也要充足,不然就算有了 ATP 也沒有材料可以建造肌肉)
但熱量在平衡時、或甚至些微赤字時,也有機會能夠增肌同時減脂。請看回圖六,如果來自小腸的營養素不夠的話,是可以動用庫存的脂肪來生產 ATP,然後再拿這些 ATP 去合成肌肉的。不過,這對肥胖者(脂肪庫存多)和新手(容易合成肌肉)才比較有效,對於已經偏瘦或重訓經驗豐富的人,還是要在熱量盈餘較能有效增肌。
總結
希望透過這篇文章,能讓大家對於「食物、能量(熱量)、和身體組織」的關係更加了解。如果下次有人問你:「熱量是怎麼變成肌肉的?」相信你已經可以自信地回答:「食物中的能量在經消化與分解後,會暫時儲存在 ATP 中,此時只要有足夠的 ATP 及氨基酸,ATP 就能提供氨基酸合成成蛋白質所需的能量,最終讓肌肉增大。」
① POL,也就是最常见的大号「煤气罐」,准确地说,叫「液化石油气罐」。我这边日常可以买到的,有 3.7kg 和 8.5kg 两种容积。大的更划算,但我的床板下面只能放进小号的,换一瓶气大约 $20,Bunnings 和很多加油站都有换。
还有一种 LCC 27 接口,是 POL 的升级版。近年来政府渐渐把 POL 气罐,升级成更安全的 LCC 27 接口。这个是向下兼容的:原先用在 POL 上的管线,仍然可以拧进 LCC 27 的气罐;反之则不行,LCC 27 专用的管线,不能用在 POL 气罐上。所以,使用 POL 的管线,就不必在乎每次换到的气罐,是旧接口还是新接口。
② 3/8″ BSP-LH,另一种大号石油气罐的接口,通常只有专门的户外型房车才会使用。加油站很少见,更换气瓶也远不如 POL 方便。可以很方便地改成 POL,户外店有转接头卖($15)。
③ UNEF 1″ / BOM,北美常见的一磅重的绿气罐,北美的加油站和便利店到处都是,但澳洲和中国很少,只有专门户外店才有。