Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


AAE-1 submarine cable enters service
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

AAE-1 submarine cable enters service

Source : https://www.totaltele.com/497388/AAE-1-submarine-cable-enters-service

Good news for pakistan and other asian countries

Thanked by 1muratai
«1

Comments

  • Any decrease in pings? Or isps too slow to make the necessary changes?

  • muratai said: Any decrease in pings? Or isps too slow to make the necessary changes?

    This system was not really designed to reduce latency (it is actually a bit longer than existing and the route from start-end does not offer much more ways to cut larger distances), it however adds a LOT of capacity.

    CU seems live on it from HK to Italy (seen via Aliyun in SZ-GD)

    Real improvements should show up for:

    • Yemen (only one ISP which is consortium member) avoids likely the African routes and the low capacity FALCON system with it

    • Vietnam, as VNPT had issues the last months with cuts and never really enough capacity EU bound but also HK bound (VNPT is a major HKIX user, and for Asia surprisingly peering friendly)

    • Myanmar (complicated right now, not much connectivity and land not too stable border wise yet) as they rely on SeaMeWe3/5 too much

    Global Transit (Thailand, not to be confused with EU one), VNPT and PCCW should be able to offer transport and transit immediately.

    Thanked by 3muratai pike alexholl
  • Probly wont change a thing since the ISPs will push traffic through the cheapest pipe out there disregarding the latency.

  • I found out about tea-4 while reading aae-1 news. It seems russian land-fiber looks promising but I haven't noticed any major isp utilizing it other than rostelecom and ex-soviet countries like kazakhstan.

  • stefeman said: Probly wont change a thing since the ISPs will push traffic through the cheapest pipe out there disregarding the latency.

    Per Mbit this cable is likely far cheaper than existing by the way higher capacity.

  • noamannoaman Member

    @William said:

    muratai said: Any decrease in pings? Or isps too slow to make the necessary changes?

    This system was not really designed to reduce latency (it is actually a bit longer than existing and the route from start-end does not offer much more ways to cut larger distances), it however adds a LOT of capacity.

    CU seems live on it from HK to Italy (seen via Aliyun in SZ-GD)

    Real improvements should show up for:

    • Yemen (only one ISP which is consortium member) avoids likely the African routes and the low capacity FALCON system with it

    • Vietnam, as VNPT had issues the last months with cuts and never really enough capacity EU bound but also HK bound (VNPT is a major HKIX user, and for Asia surprisingly peering friendly)

    • Myanmar (complicated right now, not much connectivity and land not too stable border wise yet) as they rely on SeaMeWe3/5 too much

    Global Transit (Thailand, not to be confused with EU one), VNPT and PCCW should be able to offer transport and transit immediately.

    what about pakistan?

  • Mahfuz_SS_EHLMahfuz_SS_EHL Host Rep, Veteran
    edited June 2017

    Does it do any help for Bangladesh ?? I just heard we have only SEA-ME-WE4 !

  • hostdarehostdare Member, Patron Provider

    Jio in India bringing 100 gbps link with the fiber for now , it will be increased based on demand . Overall Maximum capacity for the cable now is 40 tbps which will be increased on based on need .

  • noamannoaman Member

    @hostdare said:
    Jio in India bringing 100 gbps link with the fiber for now , it will be increased based on demand . Overall Maximum capacity for the cable now is 40 tbps which will be increased on based on need .

    How can the capacity be increased from 40tbs?

    The number of pairs of fiber optic remains the same....

    Good news for Indian users

  • hostdarehostdare Member, Patron Provider

    noaman said: How can the capacity be increased from 40tbs?

    You are right , from my understanding total capacity of all fibers running that system is 40 tbps ,maybe that is the maximum possible capacity too .

  • ZerpyZerpy Member

    @noaman said:
    How can the capacity be increased from 40tbs?

    The number of pairs of fiber optic remains the same....

    It's not only about the optics, but also about the amount of colors you send over each optic.

    DWDM anyone?

    Eventually technology will become better, meaning that you can send more colors over the fiber, which increases the overall capacity in the end.

    Thanked by 1Clouvider
  • noamannoaman Member

    @Zerpy said:

    @noaman said:
    How can the capacity be increased from 40tbs?

    The number of pairs of fiber optic remains the same....

    It's not only about the optics, but also about the amount of colors you send over each optic.

    DWDM anyone?

    Eventually technology will become better, meaning that you can send more colors over the fiber, which increases the overall capacity in the end.

    Yes that makes much more sense thanx

  • ClouviderClouvider Member, Patron Provider

    @noaman said:

    @Zerpy said:

    @noaman said:
    How can the capacity be increased from 40tbs?

    The number of pairs of fiber optic remains the same....

    It's not only about the optics, but also about the amount of colors you send over each optic.

    DWDM anyone?

    Eventually technology will become better, meaning that you can send more colors over the fiber, which increases the overall capacity in the end.

    Yes that makes much more sense thanx

    ^ this and bed table reading is here: https://en.wikipedia.org/wiki/Wavelength-division_multiplexing for those interested. Current MUX/DEMUX can happily inject up to ~80+ lengths on 10G channels so you can get more than 800G out of a single pair. Fun!

    Thanked by 2Aidan netomx
  • noamannoaman Member

    @Clouvider said:

    @noaman said:

    @Zerpy said:

    @noaman said:
    How can the capacity be increased from 40tbs?

    The number of pairs of fiber optic remains the same....

    It's not only about the optics, but also about the amount of colors you send over each optic.

    DWDM anyone?

    Eventually technology will become better, meaning that you can send more colors over the fiber, which increases the overall capacity in the end.

    Yes that makes much more sense thanx

    ^ this and bed table reading is here: https://en.wikipedia.org/wiki/Wavelength-division_multiplexing for those interested. Current MUX/DEMUX can happily inject up to ~80+ lengths on 10G channels so you can get more than 800G out of a single pair. Fun!

    Well than why arent try upgrading previous SEA-ME-WE4 and 5

    Over here in Asia we are duying for better speeds

    Thanked by 1muratai
  • Is there a connection to China?

  • AidanAidan Member

    Is there a connection to China?

    No, closest is Hong Kong.

    Thanked by 1server1
  • noaman said: Well than why arent try upgrading previous SEA-ME-WE4 and 5

    Because these are absolutely full wave wise and upgrading to more speed per wave requires new repeaters - the optical amplification is limited by the, basically, refresh frequency the diode can output.

    New cables have other amplifiers that can do more now and MIGHT be able to be upgraded in-use (but this is not likely).

    Thanked by 1Clouvider
  • Aidan said: No, closest is Hong Kong.

    HK is China in network sense - all big 3 carriers run Tbit/s of capacity through Shenzhen to HK.

    Thanked by 3Aidan Zerpy netomx
  • karanchookaranchoo Member
    edited July 2017

    ---edited

  • licsonlicson Member

    @William said:

    muratai said: Any decrease in pings? Or isps too slow to make the necessary changes?

    This system was not really designed to reduce latency (it is actually a bit longer than existing and the route from start-end does not offer much more ways to cut larger distances), it however adds a LOT of capacity.

    CU seems live on it from HK to Italy (seen via Aliyun in SZ-GD)

    Real improvements should show up for:

    • Yemen (only one ISP which is consortium member) avoids likely the African routes and the low capacity FALCON system with it

    • Vietnam, as VNPT had issues the last months with cuts and never really enough capacity EU bound but also HK bound (VNPT is a major HKIX user, and for Asia surprisingly peering friendly)

    • Myanmar (complicated right now, not much connectivity and land not too stable border wise yet) as they rely on SeaMeWe3/5 too much

    Global Transit (Thailand, not to be confused with EU one), VNPT and PCCW should be able to offer transport and transit immediately.

    Actually there are real ping improvements in HK side. It's known that Cogent Communications uses the new AAE-1 system to connect Hong Kong to Marseille. The latency of that particular link is 172ms which I think is the express routing that AAE-1 claims to avoid traditional areas to decrease latency.

  • licson said: Actually there are real ping improvements in HK side

    Ping to asia is only secondary to most, capacity is far more issue - you can take care of ping once there is a stable all around network as in Europe and the US.

    HE should also have capacity on this, or turn up shortly, to SG/HK - this makes sense as it is cheaper than existing options (or, for large players - so not HE Asia lol - offers easier usage by less waves).

  • licsonlicson Member

    @William said:

    Aidan said: No, closest is Hong Kong.

    HK is China in network sense - all big 3 carriers run Tbit/s of capacity through Shenzhen to HK.

    In political sense, yes but in a network sense, not really. Although Chinese operaters all have terabits of capacity to Hong Kong, they are all dark fibres and hardly utilized to serve the country. Also their routing policy prefers sending traffic over to the US directly. To compare the resillency of the networks, HK is better in terms of available network peers and number of submarine cables.

    In mainland China, most providers need to participate in remote peering via submarine cables as most peers are not available inside mainland China. This renders the peering unstable as the link overloads or cable outages. In contrast, there are more peers in Hong Kong and an IX that has importance on local peering. Therefore providers can have their peering more easily and do not need to participate in remote peering.

    In the numbers of submarine cables, Hong Kong has landed more submarine cables than the whole mainland China combined, with 11 activated submarine cables and 2 in progress. This also makes network resillency higher as more fallback links are available. In mainland China, a cable break can cause Internet disruption in a greater sense, especially for the TPE trans-Pacific cable that is known to be as fragile as the AAG cable system that caused chaos in Vietnam. With the older CUCN cable deactivated, Chinese operaters have fewer fallback links to the US, making the situation even worse.

    You'll start to understand more if you're a Hong Kong-based provider.

  • licsonlicson Member

    @William said:

    licson said: Actually there are real ping improvements in HK side

    Ping to asia is only secondary to most, capacity is far more issue - you can take care of ping once there is a stable all around network as in Europe and the US.

    HE should also have capacity on this, or turn up shortly, to SG/HK - this makes sense as it is cheaper than existing options (or, for large players - so not HE Asia lol - offers easier usage by less waves).

    Yeah I agree with your point on capacity, AAE-1 did provide much higher bandwidth than previous solution s. This will benefit everyone as the cable goes alive.

  • licsonlicson Member

    @licson said:

    @William said:

    muratai said: Any decrease in pings? Or isps too slow to make the necessary changes?

    This system was not really designed to reduce latency (it is actually a bit longer than existing and the route from start-end does not offer much more ways to cut larger distances), it however adds a LOT of capacity.

    CU seems live on it from HK to Italy (seen via Aliyun in SZ-GD)

    Real improvements should show up for:

    • Yemen (only one ISP which is consortium member) avoids likely the African routes and the low capacity FALCON system with it

    • Vietnam, as VNPT had issues the last months with cuts and never really enough capacity EU bound but also HK bound (VNPT is a major HKIX user, and for Asia surprisingly peering friendly)

    • Myanmar (complicated right now, not much connectivity and land not too stable border wise yet) as they rely on SeaMeWe3/5 too much

    Global Transit (Thailand, not to be confused with EU one), VNPT and PCCW should be able to offer transport and transit immediately.

    Actually there are real ping improvements in HK side. It's known that Cogent Communications uses the new AAE-1 system to connect Hong Kong to Marseille. The latency of that particular link is 172ms which I think is the express routing that AAE-1 claims to avoid traditional areas to decrease latency.

    As I know, HE has chosen capacity on SMW5 instead. They currently have 2 * 100G from HK to SG and another 2 * 100G from SG to Marseille.

  • WilliamWilliam Member
    edited July 2017

    licson said: In political sense, yes but in a network sense, not really. Although Chinese operaters all have terabits of capacity to Hong Kong, they are all dark fibres and hardly utilized to serve the country

    You seem to confuse transit with transport POP, aside of this China Mobile certainly is in HK. I have them on HKIX.

    The others are there but mainly to:

    • Pick up Asian transport (especially to SG)

    • Pick up US/EU bound transport

    • Sell direct china transit (direct and to larger players as PCCW)

    They have no interest in anything else (especially HKIX peering) than this and no one else has any in them.

    Nobody wants to serve HK from CN and nobody serves CN from HK aside of direct China (no P2P, no waves) so most capacity goes in in various directions by HK being a large and notably - this is very important for the large cable owner conglomerate members as TATA and Telecom Italia - commercial judicative independent POP (as a cable landing station in China is Chinese, no matter who you are or what you do for them, it belongs to Beijing).

    licson said: Also their routing policy prefers sending traffic over to the US directly

    Traffic to US passes HK from CN often (however EU does more). You just do not see a HK hop because this is not a routed point; it's L2. The nearest CN hop - if any - is in Guangdong province.

    You can correlate this by knowing the latency of HK<->US (and US-CLS<->US-DC) and the latency from CN-location<->CN-Shenzhen, then just add them, which is your ping from China +/-10ms.

    The second path is cables ending further up and connecting Taiwan/US, but this is another route. Third are land-based routes, the cables to Korea (both North and South) and Russia notably.

    licson said: You'll start to understand more if you're a Hong Kong-based provider.

    I owned a HK based provider of medium size primarily selling VPS, maintain a HKIX presence for a customer, run my private BGP in HK, set up HK locations for various VPS providers (all highly successful which speaks for market knowledge and good pricing already) and have colo in multiple Asian countries as well as a transport link from Oceania to Asia.

    Maybe this adds to the point, as HK ISP you should have no issue to understand any, depending on age and years in business:

    • I ran into Sunnyvision (for colo) and had insane issues, same with their 36cloud (where i at times was one of the larger clients later also)

    • I saw the rise and fall of CUHK with the primary HKIX pop, then local loop issues resulting in HKIX2, 3 and 4 ultimately aside of CITIC

    • The Pacswitch Globe Telecom guy (no need to name him or current company name) tends to hate me/my companies for various reasons, from "abusing" paid power/BW to going around telling their fascinating history and disputes

    licson said: As I know, HE has chosen capacity on SMW5 instead

    This was before AAE-1 was finished; HE already has an Asian presence since some years, expanding locations since then (TW and KR as newer ones eg.).

  • Interesting, but makes sense if they lay in parts - i wonder what path CU turned up then as latency did reduce a lot but with their general no hops in trace hard to track down. Might be via India or Viet Nam.

  • OBHostOBHost Member, Host Rep

    It will make some good routes between Asia, Africa and Europe.

  • OBHostOBHost Member, Host Rep

    Already we notify our clients about it, 4th July and 8th July and 20 July to 24 July, Maintenance Time..

  • Anyone know if Telin (AS7713) use this cable too? Latency from Indo to Europe for some provider now under 200ms

Sign In or Register to comment.