Raspberry Pi 4 2GB 版本「暫時」漲價,另外 1GB 版本回歸

Raspberry Pi 宣佈 Raspberry Pi 4 的 2GB 版本將從 US$35 漲到 US$45,另外本來淘汰掉的 1GB 版本回歸,賣 US$35:「Supply chain, shortages, and our first-ever price increase」。

In February last year, we announced that we were discontinuing the 1GB variant of Raspberry Pi 4, and moving the 2GB product to our signature price of $35. We’re still glad we did this, as countless young people made use of this device as they studied from home during the pandemic.

Unfortunately, cost increases caused by the current shortage mean that this product is not currently economically viable at this reduced price point. We are therefore moving it back to $45 on a temporary basis.

To support the many industrial customers who have designed the 2GB variant of Raspberry Pi 4 into their products, we are reintroducing the 1GB variant at the $35 price point. This provides a degree of choice: less memory at the same price; or the same memory at a higher price.

然後共用 40nm 製程的 3B+ 會暫時先被降低優先權,產能會優先提供給 CM3CM3+3B

In allocating our limited stocks of 40nm silicon, we will prioritise Compute Module 3, Compute Module 3+, and Raspberry Pi 3B, and deprioritise Raspberry Pi 3B+.

整個供應鏈的情況被炸的亂七八糟...

Amazon EC2 推出 m6i 的機器

AWS 給了公告,在 Amazon EC2 上面推出了 m6iIntel-based 新機種:「New – Amazon EC2 M6i Instances Powered by the Latest-Generation Intel Xeon Scalable Processors」。

這好像是第一次看到 Intel-based 機種加上了 i 的 suffix...

這次比較大的兩個差異,與 m5 相比,多出了 m6i.32xlarge

A larger instance size (m6i.32xlarge) with 128 vCPUs and 512 GiB of memory that makes it easier and more cost-efficient to consolidate workloads and scale up applications.

另外看了一下 us-east-1 上的單價,看起來與 m5 系列的機器價錢一樣,但是效能提昇了 15% (然後很假掰的寫了 price/performance?):

Up to 15% improvement in compute price/performance.

單以數字看起來的話還是 m6g 系列會比較香?當然如果只有 x86-64 binary 的話看起來還是可以考慮換到 m6i 上跑...

AWS 同一區的 VPC Peering 流量不收費了

AWS 在同一個 AZ 裡面的流量是不收費的,但如果是跨帳號的話,還是要當作 inter-AZ 流量 (收 USD$0.01/GB 的費用),現在則是宣佈不用了:「Amazon VPC Announces Pricing Change for VPC Peering」。

要注意的是不同帳號的 a 不一定相同 (像是 us-east-1a 在不同帳號對應到的實際 AZ 不同),得透過 AWS 提供的資料確認底層實際的 AZ 是哪個。

回朔到這個月月初生效:

Starting May 1st 2021, all data transfer over a VPC Peering connection that stays within an Availability Zone (AZ) is now free. All data transfer over a VPC Peering connection that crosses Availability Zones will continue to be charged at the standard in-region data transfer rates. You can use the Availability Zone-ID to uniquely and consistently identify an Availability Zone across different AWS accounts.

RDS 推出 ARM 版本

Amazon RDS 推出了 ARM 的版本:「New – Amazon RDS on Graviton2 Processors」,包含了 MySQLMariaDBPostgreSQL 的版本都有支援,不過看起來需要比較新版的才能用:

You can choose between M6g and R6g instance families and three database engines (MySQL 8.0.17 and higher, MariaDB 10.4.13 and higher, and PostgreSQL 12.3 and higher).

官方宣稱可以提供 35% 的效能提昇,考慮費用的部份會有 52% 的 c/p 值提昇:

Graviton2 instances provide up to 35% performance improvement and up to 52% price-performance improvement for RDS open source databases, based on internal testing of workloads with varying characteristics of compute and memory requirements.

對於 RDS 這種純粹就是個服務的應用來說,感覺應該不會有什麼轉移成本,只要測過沒問題,換過去等於就是現賺的。看起來等 RI 約滿了就可以切...

Google Cloud Platform 在台灣的機房可以開 Standard Network 的機器了

Google Cloud Platform 一開始是提供 Premium Network,會透過 Google 自家的網路骨幹連到最近的點,然後再透過當地的機房交換出去,這樣可以確保頻寬的穩定性,但成本當然也就比較高...

後來提供了 Standard Network 則是從機房出去後就直接交換,成本會比較低 (參考「Network Service Tiers - Custom Cloud Network」這篇),但在台灣的機房一直都沒有提供 Standard Network (好像是需要另外申請?),所以我每個月月底的時候都會測一下看看開放了沒... 然後剛剛發現可以開起來了,不確定是已經全開了還是分批開。

測了一下發現網路相當... 爛?是還在調整嗎...

像是 1.1.1.1 的 latency 很高 (自家的 8.8.8.8 當然就沒這個問題):

PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
64 bytes from 1.1.1.1: icmp_seq=1 ttl=49 time=48.9 ms
64 bytes from 1.1.1.1: icmp_seq=2 ttl=49 time=48.5 ms
64 bytes from 1.1.1.1: icmp_seq=3 ttl=49 time=48.5 ms

--- 1.1.1.1 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2003ms
rtt min/avg/max/mdev = 48.554/48.691/48.964/0.193 ms

然後 168.95.1.1139.175.1.1 也都很差 (61.31.1.1 不給 ping):

PING 168.95.1.1 (168.95.1.1) 56(84) bytes of data.
64 bytes from 168.95.1.1: icmp_seq=1 ttl=239 time=21.2 ms
64 bytes from 168.95.1.1: icmp_seq=2 ttl=239 time=21.3 ms
64 bytes from 168.95.1.1: icmp_seq=3 ttl=239 time=21.4 ms

--- 168.95.1.1 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2003ms
rtt min/avg/max/mdev = 21.276/21.367/21.454/0.072 ms
PING 139.175.1.1 (139.175.1.1) 56(84) bytes of data.
64 bytes from 139.175.1.1: icmp_seq=1 ttl=53 time=63.4 ms
64 bytes from 139.175.1.1: icmp_seq=2 ttl=53 time=62.9 ms
64 bytes from 139.175.1.1: icmp_seq=3 ttl=53 time=62.9 ms

--- 139.175.1.1 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2003ms
rtt min/avg/max/mdev = 62.967/63.139/63.455/0.303 ms

不過學術網路倒是還不錯:

PING 140.112.2.2 (140.112.2.2) 56(84) bytes of data.
64 bytes from 140.112.2.2: icmp_seq=1 ttl=51 time=5.13 ms
64 bytes from 140.112.2.2: icmp_seq=2 ttl=51 time=4.40 ms
64 bytes from 140.112.2.2: icmp_seq=3 ttl=51 time=4.52 ms

--- 140.112.2.2 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2003ms
rtt min/avg/max/mdev = 4.405/4.690/5.138/0.325 ms
PING 140.113.250.135 (140.113.250.135) 56(84) bytes of data.
64 bytes from 140.113.250.135: icmp_seq=1 ttl=55 time=5.87 ms
64 bytes from 140.113.250.135: icmp_seq=2 ttl=55 time=5.97 ms
64 bytes from 140.113.250.135: icmp_seq=3 ttl=55 time=6.11 ms

--- 140.113.250.135 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2002ms
rtt min/avg/max/mdev = 5.872/5.987/6.119/0.135 ms
PING 140.117.11.1 (140.117.11.1) 56(84) bytes of data.
64 bytes from 140.117.11.1: icmp_seq=1 ttl=242 time=9.52 ms
64 bytes from 140.117.11.1: icmp_seq=2 ttl=242 time=9.17 ms
64 bytes from 140.117.11.1: icmp_seq=3 ttl=242 time=9.20 ms

--- 140.117.11.1 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2003ms
rtt min/avg/max/mdev = 9.172/9.298/9.521/0.176 ms

有需要的人可以測試看看了...

AWS 日本區 EC2 與 S3 的傳輸費用降價...

Twitter 看到 Jeff Barr 提到日本區的 EC2S3 傳輸費用降價:

網站的說明文章則是在「AWS Data Transfer Price Reductions – Up to 34% (Japan) and 28% (Australia)」這邊。分成幾個部份降價:

  • 10TB 以下的費用從 USD$0.14/GB 變成 $0.114/GB (約 19%)。
  • 10TB 到 50TB 從 USD$0.135/GB 變成 USD$0.089/GB (約 34%)。
  • 50TB 到 150TB 則是 USD$0.13/GB 變成 USD$0.086/GB (約 34%)。
  • 超過 150TB 的部份從 USD$0.12/GB 變成 USD$0.084/GB (約 30%)。

然後自動回朔到 2018/09/01 開始算:

Effective September 1, 2018 we are reducing prices for data transfer from Amazon Elastic Compute Cloud (EC2), Amazon Simple Storage Service (S3), and Amazon CloudFront by up to 34% in Japan and 28% in Australia.

另外有提到 CloudFront 也有降價,以及澳洲的部份,應該不太會碰到所以就跳過去...

但就文章這樣寫明 EC2 與 S3 的流量有降價,應該表示從 ELB 出去的流量就不算在這次的降價?除非你是直接 S3 裸奔,不然對大多數的人應該沒差?

測試 TPUv2 的 C/P 值

有人用相同演算法實際測試 Google 的 TPUv2 與 NVIDIATesla P100 的 C/P 值了:「Benchmarking Google’s new TPUv2」。

如果以 ResNet-50 當作計算的演算法,可以看到其實 C/P 值的差距沒有想像中大。主要原因是 GPU 可以使用較低的精度計算以加快速度,而非 Google 之前新聞稿故意使用較高精度比較 (TPU 使用 8-bit matrix engine,所以 GPU 使用較低的 fp16 版本比較會比較有參考價值):

真正的差異是在 LSTM

It turns out that the TPU is even faster on the LSTM model (21402 examples/s): ~12.9 times faster than a P100 (1658 examples/s) and ~7.7 times faster than a V100 (2778 examples/s)!

不過這邊就沒特別提到精度了...

AWS Price List API

AWS 把價錢資訊也 API 化了:「New – AWS Price List API」。

除了可以透過 API 取得資訊外,還可以透過 Amazon SNS,在價錢有變動時得到通知:

You can also elect to receive notification via Amazon Simple Notification Service (SNS) each time we make a price change.

讓 billing 的各種計算變方便。

Oracle 也要推動自己的 EC2 了...

在「Oracle finally launches Elastic Compute Cloud, 9 years after Amazon debuted EC2」這邊看到 Oracle 也要推動自己的 Amazon EC2 了。Oracle 官方的新聞稿在「Oracle Updates Oracle Cloud Infrastructure Services」這邊可以看到。

官方網站在「Oracle Cloud Infrastructure as a Service (IaaS)」這邊,價錢理論上可以在「Compute Cloud Pricing」這邊可以看到,不過沒看到 Price 欄位啊...

不知道什麼人會去用... 綁專案?

CloudFront Price Class 100 奇怪的現象

剛剛發現奇怪的現象,晚點來問問看好了... 手上的 test-priceclass100.gslin.org 設定為 Only Use US and Europe:

但在 SmokePing 上卻看到極低的 latency 而覺得納悶:

動手查了一些資料發現是被導到台灣的點... 可以看到先問 168.95.192.1 (HiNet 光世代預設的 DNS) 後得到了這些 IP:

gslin@GSLIN-HOME1404 [~] [02:43/W4] dig test-priceclass100.gslin.org @168.95.192.1

; <<>> DiG 9.9.5-3ubuntu0.5-Ubuntu <<>> test-priceclass100.gslin.org @168.95.192.1
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 48453
;; flags: qr rd ra; QUERY: 1, ANSWER: 9, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 3072
;; QUESTION SECTION:
;test-priceclass100.gslin.org.  IN      A

;; ANSWER SECTION:
test-priceclass100.gslin.org. 300 IN    CNAME   d265g70k07is7o.cloudfront.net.
d265g70k07is7o.cloudfront.net. 60 IN    A       54.192.215.203
d265g70k07is7o.cloudfront.net. 60 IN    A       54.192.215.98
d265g70k07is7o.cloudfront.net. 60 IN    A       54.239.176.96
d265g70k07is7o.cloudfront.net. 60 IN    A       54.192.214.40
d265g70k07is7o.cloudfront.net. 60 IN    A       54.239.176.216
d265g70k07is7o.cloudfront.net. 60 IN    A       54.192.212.118
d265g70k07is7o.cloudfront.net. 60 IN    A       54.239.176.223
d265g70k07is7o.cloudfront.net. 60 IN    A       54.239.176.56

;; Query time: 178 msec
;; SERVER: 168.95.192.1#53(168.95.192.1)
;; WHEN: Thu Oct 22 02:43:24 CST 2015
;; MSG SIZE  rcvd: 228

接下來跑 mtr 看,發現是台灣的節點:

gslin@GSLIN-HOME1404 [~] [02:43/W4] mtr --report -n 54.239.176.56
Start: Thu Oct 22 02:43:37 2015
HOST: GSLIN-HOME1404              Loss%   Snt   Last   Avg  Best  Wrst StDev
  1.|-- 168.95.98.254              0.0%    10    8.6   9.1   8.4   9.9   0.3
  2.|-- 168.95.25.50               0.0%    10    8.1   9.0   8.0   9.9   0.3
  3.|-- 220.128.11.226             0.0%    10    9.3  11.8   8.1  31.7   7.1
  4.|-- 220.128.5.94               0.0%    10   12.9  13.9   9.6  17.9   2.9
  5.|-- 220.128.4.177              0.0%    10    8.7  13.1   8.7  35.3   8.0
  6.|-- 203.75.228.29              0.0%    10    9.9  10.1   9.3  10.9   0.3
  7.|-- 223.26.65.19               0.0%    10    9.4  10.3   9.4  11.4   0.3
  8.|-- 202.133.255.122            0.0%    10   11.0  10.8   9.5  13.5   1.0
  9.|-- 54.239.176.56              0.0%    10   10.7   9.4   8.4  10.7   0.7

這是哪招呢...