Try Free

The Hidden Challenge of AI Data Centers: Power, Cooling & What's Next — Season 4 Ep 7

Nusantara Academy July 26, 2026 56m 8,454 words
▶ Watch original video

About this transcript: This is a full AI-generated transcript of The Hidden Challenge of AI Data Centers: Power, Cooling & What's Next — Season 4 Ep 7 from Nusantara Academy, published July 26, 2026. The transcript contains 8,454 words with timestamps and was generated using Whisper AI.

"if you talk about the cooling i want to let you look back like a motorcycle you see those small motorcycles right the engine only use air cool because the engine is small so the air still good enough to bring the heat out from the engine and you see if the big motorcycle like a big bike you need..."

[00:00:00] Speaker 1: if you talk about the cooling i want to let you look back like a motorcycle you see those small motorcycles right the engine only use air cool because the engine is small so the air still good enough to bring the heat out from the engine and you see if the big motorcycle like a big bike you need the water the radiator you need the water to cool down the engine similar thing to a data center so in the past when we talk about 5 kilowatt 10 kilowatt per rack efficient enough to blow through the server the fan right to grow and then to bring the heat out but when the heat density come to like i mentioned 200 kilowatt per rack the density of the air is not enough to bring the heat out right that's why the water come [00:01:17] Speaker 2: hi everyone welcome to the santara academy podcast my name is nila and today i'll be your host and today our topic is inside the rack the relationship between power cooling and sustainability in the ai era before we begin don't forget to like comment subscribe to nusantara academy's youtube channel to stay updated on the latest digital infrastructure trends and today i'm joined by two distinguished guests our first speaker is mr sakda sayong regional data center business platform head and regional director for southeast asia and oceana at data delta electronics welcome pa sakda and our second speaker is mr nawijaya chairman of the indonesia data center ecosystem council or edcec thank you for joining us finally sakda sakhda is the regional data center business platform head and regional director for southeast asia and oceania at delta electronics throughout his careers sakda has contributed to the development of power supply technologies and has led regional business expansion across southeast asia australia and new zealand he currently oversees delta's data center platform business covering power infrastructure cooling solutions software and integrated data center solutions for enterprise telecom and channel markets patnawijaya is the chairman of the indonesian data center ecosystem council or idcec the association has a mission to advocate for stronger collaboration between global and local shared stakeholders to promote best practices facilitate knowledge exchange and advocate for policies that support the growth of indonesia's digital infrastructure ecosystem panawi is currently active in inviting corporates and professionals to become a part of the association to promote indonesia's digital ecosystem with data center as the foundation all right so i guess now we can begin to the question and answer session of this podcast and so you know how ai is currently the main driving force of growth on the computing demand and also power density and i guess my first question would be to pass up the um you know how is ai changing the way data centers are designed and built today especially you know in comparison to about five to ten years ago [00:03:56] Speaker 1: the ai requirement will change the data center infrastructure design in probably two or three aspects the first thing is that ai require a very high density per rack you know so in the past we we talked about 10 15 kilowatt per rack right but as of now the ai will have to support like 100 to 200 kilowatt per rack so that makes a lot of change in terms of the power distribution the cooling requirement and then ai will also change how we supporting the the network equipment the fabric right because that the accelerated computing from the ai will require more bandwidth in terms of the networking and in the very near futures ai will change another step from the ac to be the dc supply to the data set that's also quite challenging in the industry at the moment yeah in terms of the design [00:04:58] Speaker 2: yeah i think that's very interesting because you know like ai the workload it requires a lot of gpu and then the gpu requires a lot of power and also the backup power right and all of this power generates a lot of heat and so you need more cooling as well and so i feel like the power and cooling as you have mentioned it will become like the backbone of the infrastructure so now we looking at you know indonesia how ready is our ecosystem to support the growth of this digital ai infrastructure well if we look at indonesia now i [00:05:33] Speaker 3: believe indonesia is on the right track uh over the past few years we have seen like many data centers like hyperscale uh built in indonesia significantly right then um they they invest a lot in indonesia and build a big data centers like more than 100 megawatts or or more up to 500 megawatts right i think it's is is a good good chance for indonesia to become like a hub of uh data centers industries yeah [00:06:09] Speaker 2: certainly i mean the market is there like the demand is also there but what do you think are the [00:06:14] Speaker 3: challenges ah okay about the challenges i think that uh we have challenge in like power availability now and then because every data centers hyperscalers come indonesia request i need 100 megawatts i need 200 megawatts yeah that's right yeah yeah they book all the powers right yeah power availability and then uh we have also difficulties to get uh renewable energy right how to get the green energy to become like um more uh carbon footprint or more green energy in data centers okay and then we also have talents in skill engineer because now we are like of uh engineers who really focus and know well known about the data centers industry and and technical okay and the fourth one um the speed of the infrastructure's development in indonesia is still a bit slower compared with other countries yeah the expansion right [00:07:26] Speaker 2: yeah okay so you mentioned a lot of really good factors there that are important in building the digital infrastructure but i think that firstly we should focus on power first um for this discussion and so yeah power is really important right reliability and also its availability of course and maybe can you expand a bit more why is it so important what happens if we don't have enough power or reliable power what happens to the data centers [00:07:56] Speaker 3: yeah every data centers are running based on power ai cloud or gpus everything's need power power is everything's like in human power is like like the heart without power you can cannot do anything even you have very expenditures uh gpus or any kind of technology but without power you can get it out [00:08:26] Speaker 2: yeah i think it's very very important yeah i totally agree and sometimes where a data center built is determined based on the availability of the power so like where the land is you follow where the power is and you know uh mr sagda so what are the biggest um power challenges in facing the industry today [00:08:52] Speaker 1: both outside and also inside the racks yeah from from outside right if we look at actually we we have done some research in our company in terms of what should be the the obstacle of a data center business to grow right and the first thing we actually look about the power generation whether we we can't generate the power or not but actually it's not true the the main the major obstacle actually on the power delivery or the distribution yeah because the utility company right when they want to invest the delivery or the power transmission to the point where we would like to construct our data center right they they still hesitate to invest that kind of distribution yeah because they're also unsure when they invest this distribution to that area they can have more customers more usage on that area or not and what we see nowadays that the growth of data center is around like 36 to 40 percent growth in terms of the power usage right but the growth of the power distribution it really you know they cannot catch up yeah they cannot check out and actually i look at the regional scale at the moment basically we are already at the to reach at that equality of the demand and supply already yeah at the moment so they're gonna be a very big uh challenge for from the outside to a data center even especially the new data center construction right but in terms of the internal data hall itself uh let's let's think this way right in the past when we design uh three or four kilowatt per rack right the the current bring to the rack is just like 16 20 m right but now when we talk about 100 or 200 kilowatt per racks we need to bring like 400 amp the current density to just single rack inside our data hall so how we can handle such a very high uh current in our data hall yeah in a safety manner right you you think about like a 400 amp breaker to each of the rack into the data halls not include the the cooling right you have the piping of the liquid go to your racks that also creates some of the challenge in terms of the the maintenance or reliability proven yeah that there's also this outside and inside uh challenge into the data center yeah it's it's a lot of things to really [00:11:30] Speaker 2: consider and it's really shaping the way we designed our data center as well yeah and it seems like the demand is growing more and more and so maybe the key is to just plan ahead instead of reacting to the [00:11:44] Speaker 1: demand okay right in both inside and outside yeah we also need to convince to those the power company to invest in terms of the distribution like i mentioned when when we ask them to expand or to invest in terms of more substation more power distribution to the the area because you you you also in the data center industry right to find the land plot to to build a data center also not easy right but we look at several factors right whether the flooding whether that's transportation with yeah that's true yeah right so one we find a very good land plot but we just find out yeah that the power company doesn't have enough distribution to the area and when we ask them whether can you supply more power to my area they say yeah only for you or not and then any other customer that i can serve yeah that i can invest more distribution so that you know forward and back discussion between you know like it was chicken and egg yeah chicken and egg yeah so i believe like some country they they really trying to convince the power company in terms of investing the the distribution there they even like have a zoning right so that yeah and make sure the they're gonna be a multiple uh customers over there so they can invest and then they can expect on the sensitivity on the on return of the investment all right uh but now we [00:13:11] Speaker 2: looking at indonesia how ready is our ecosystem to support the growth of ai infrastructure and what [00:13:17] Speaker 3: are the challenges that we might face yeah the good news is that indonesia has a major advantages of a large digital economy and a strong demand and plenty room to grow right then so the question is no longer whether indonesia can build uh data centers or not but the question is whether we can build a complete ecosystems that support ai growth today right then ai infrastructures like sakda said is very different from traditional data centers traditional data centers can be like 5 to 10 megawatts but now each rack can up to 120 even to 200 racks correct yeah yeah i think i think uh i can take a note that the country is that success in ai will be not the the one with the biggest [00:14:11] Speaker 2: building but the strongest ecosystem yeah i think i i find that very true as well that it needs to be an overall strength not just you know on one factor right and i guess you mentioned about power so i guess um let's go let's dig deeper more on that yeah and you know power is definitely like a data center cannot function without power so would you like to explain a bit more about why power is so important [00:14:41] Speaker 1: because like like we just mentioned right for the ai uh requirement they they need a lot of more power it's even look at about like 100 times from the past right and then the utility company might not be able to support that kind of such a high demand in terms of the power right like i mentioned earlier we we have done some study in terms of what really i create the obstacle to grow a data center business and we see this the distribution is a major thing that create really are obstacles so if the country would like to grow these data centers we have to prepare enough power and then don't forget that the ai data center will also use the same power source to supply to the people as well so in some countries they also have like a discussion between pro and con right with the data center going to use the power from the people right when we need the people for other purposes uh we we don't have the power to supply us so yeah so that that's how important and then um data center especially in ai now if you have heard about the term ai factory right because the way of data center change how to say um right now our trend the power trend in the part we think about the data center is to store data or to keep to process data right but now ai change this way to be ai factory mean one kilowatt of the power can generate how many tokens right that's become kind of the factory to change the energy to be a token to serve the ai purpose purpose so if we just we don't have the power we don't have the token generation and that is mean those data center operator doesn't have the money right is leaving you to to get back into their uh [00:16:44] Speaker 2: investment yeah wow and yeah and the ai is like it's like central to our you know economy right now so if there is any disruption or any outage it could have a widespread effect right hey yeah and so um maybe i'll ask where about that like what are the biggest power challenges both inside and outside of [00:17:05] Speaker 1: the data halls yeah the i mean from the outside that still like i mentioned on the power distribution it is not the power generation as we have the study done and what we see now uh if you look at the power consumption totally in the region it already reached to the demand and supply point where the power company they they can distribute and then when we talk to the utility company like like i say they also quite hesitate to invest more the distribution or the power transmission to the area you you are in the industry of the data center right to find a land plot to build a data center it's not easy right you you need to avoid many factors like the flooding you need to have ensure the logistics you need to make sure that the cable or the connectivity from the submarine cable will reach to that particular land plot right but at the end if you find out that particular land you can't find a power or the utility company say they cannot allow you to use power because they want to reserve power for other purpose let's say like the industrial zone or those things so i think to talk to the the utility company or to government you know in terms of the investment in terms of this power distribution is also quite important and then if we can have like specific zoning to make sure them when they invest in terms of the power distribution they can get return of money that could be the the other thing that we can look at that um talk about inside data hall right um in the past when we have a data center like this two or three kilowatt per rack we just bring like uh 10 20 amp to the rack but but now when we talk about 200 kilowatt you need like uh almost uh 400 amp to the rack so yeah that's a lot yeah that's a lot so you think about the cable right if 400 them require one inch of cable go to each rack how hard of the cable yeah it's not include the circuit breaker or those protective device right so that's all the challenge how to operate that in the safety manner right and not include the the liquid yet the water that get into the rack how to avoid the leakage and then the power inside that make things more complicated in a data hall that also kind of actually really challenge our day uh in terms of the data that the operators yeah yeah there's actually a [00:19:38] Speaker 2: a lot of things that needs to be done behind the scenes right whether it's in the field or actually inside the data hall and yet we must deliver all of these power seamlessly to the customers when actually you know in in in practice there's a lot of things that are going on that needs to be a certain right and i guess the one of the key highlights of what you were saying earlier just now is that it requires a lot of collaboration between the parties right and so yeah this brings me to my next question for panel we actually um so you know from the ecosystem perspective so what needs to happen in order to ensure efficient and reliable power infrastructure so what needs to happen and who needs to be [00:20:22] Speaker 3: involved in making that happen yeah talking about power i think uh without reliable power we can cannot have a ai cannot have cloud cannot have uh uh what we call it uh big hyperscale data centers right and uh talking about responsibility who should take this responsibility right then i think we cannot uh uh uh request this only to one party like pln we have in indonesia now pln is our our local uh power company or state-owned company right but i think uh government should uh take responsibility also and then the data centers operator also need to take responsibility right you you can request more power but when uh by when you can utilize it i mean like within how many years right so pln or government can plan oh this data centers request 500 megawatts but they can utilize it maybe like volume up up to five another five years or another two years or something then this this data centers operator also need to take responsibility also the the technology company also like delta right they need to um um well because the rnd need to produce power generation or or ups somethings to to adapt to the new new technology right i think i think that's this part of uh should we take responsibility on that okay yeah so [00:22:03] Speaker 2: you mentioned earlier that it seems like there is a lot of collaboration and a lot of planning that needs to be done in order for the power demand to be met and i think another point that i catch you um saying just now was about cooling and how important it is is it is very significant because the ai data data centers they generate a lot of heat and you know and especially indonesia is like a hot climate as well so i guess that makes it a very important point uh can you um explain to us why cooling is such a critical um component to the infrastructure and also perhaps walk us through the different types of cooling okay yeah i if you talk about the [00:23:10] Speaker 1: cooling i i want to let you look back like very close experience or example to artists like a motorcycle right you see those small motorcycle right the engine only use air cool because the engine is small so that the air still good enough right to to bring the heat out from the engine and you see if the big motorcycle like a big bike right you need the the water right the radiator you need the water to to cool down the engine similar thing to a data center right so in the past when we talk about five kilowatt ten kilowatt per racks right the air still efficient enough to blow through the server yeah the fan right to grow and then to bring the heat out actually from data center perspective we we think about air or liquid that it actually the the heat mover it actually to move the heat from the gpu and then bring this heat out to the air outside a data hall right so when the density is small the air is efficient enough to like just the the aircon at our home right to bring the air go through the heat sink to the gpu and then go to the outside right but when the heat density come to like i mentioned 200 kilowatt per racks the density of the air is not enough to to bring the heat down right that's why the water come because water is very density right it's very it's liquid right so water is much more efficient to go to the gpu or go to the server and then bring out the all the heat and then flow out to the the room that's why when the power density increased right that would be the usage of the liquid cooling come to a data hall right to help to move those heat out from the server yeah [00:25:07] Speaker 2: so um you know like right now in southeast asia so you are you saying that cooling technologies in the form of liquid cooling is the most suitable one now for um ai data centers i mean for for the server [00:25:21] Speaker 1: level right when when the density come to like 100 kilowatt or 200 kilowatt yes the liquid is the best way to bring out the heat but they also are a different type of the liquid as well if you have heard that may be the uh direct to chip right it means you you put the liquid flow to the the gpu and it flow out and the other one would be the immersion cooling where you you put or you drip the whole uh server into the the liquid right and then lift the liquid to to bring the heat out from the room in the i have both kind of customer you know some of them they have started using the immersion cooling even like five years ago okay but finally i i saw they also changing to those uh the direct to chip because that seemed more common and then more i mean easier to maintain and then we can find more and more supplier or the skilled worker that can work with this kind of technology so in terms of the the cooling technology for uh soviet asia i see those are liquid to chip right direct to chip or those things will be more will be gain more momentum in terms of the that the usage yeah but if we talk about from the rack to outside right or or to the the the service hall right there's still uh someone we also find the cloud operator still using the air type it means we they use so-called liquid to air solution it's become that easier to maintain in a data hall and then they can try to re-utilize the the existing data hall right they don't need to modify to put the cdu to put the pipe in the data hall they can still keep using the fan wall or the the room cooling and then they connect the gpu server to the uh so-called the the air to uh the liquid to air uh cooling solution and then blow the heat through the the room right but we also see more some of the gpu they use the liquid to liquid which it requires the cdu right so you bring the coolant that the coolant right coolant and if they have another coolant loop they primarily go to the shiller system but that also create quite a lot of the complication in terms of the design in terms of the maintain in terms of those things that's why that this pro and con of of these two yeah yeah [00:28:03] Speaker 2: wow that's a lot of different types of um cooling that you were mentioning i mean my knowledge on cooling systems aren't that deep but yeah i mean from what i've heard like mostly it's direct to chip and that immersion cooling is still relatively niche but you said that you have experience in that even since five years ago so that's that's very interesting so based on your explanation just now it seems that more um data center companies they are willing to adopt more of uh liquid cooling and so what are the changes that needs to happen inside the data center and so in terms of its design and operations in order to support these technologies so nowadays it's that's even the the argument between [00:28:48] Speaker 1: just to bring the pipe into a data hall right whether we should bring to the red under the red floor so just in case if that's a leakage happen right all those leak water can not to go to the racks right and someone say my data hall doesn't have the red floor anymore right because people also try to avoid the red floor and how right have to bring the bike on the top yeah and when they bring a bike on the top if that is the leakage happen right how to prevent that kind of leakage right and then we also talk about all the the coupling the connection right when we want to connect the water to the rack how to do like hot swap to plug right to add that without any drop of the leakage that's also a kind of challenge as well and the equipment or the supplier of these kind of the the couple is also quite limited in in the industry and also all those piping we talk about the lifetime uh talk about the contamination corrosion or those things right it's it's not the piping for like our garden water CRI this for data center we talk about temperature as well right yeah temperature yeah and then we need to make sure if you're not like strain or like have some losing during next 10 years so all those material the assembly work the installation work need to be uh precise really precise really uh really well uh prepared all those things so that that a lot of challenge for the data center operator now so they in terms of bringing into [00:30:27] Speaker 2: this liquid cooling into data hall yeah i mean all of these designs like some of them are like changes to the system not even just to a component right yeah because the system is tough to change when you have installed it already because it's like it's like your vein basically you can't change your vein but you can change your organs right yeah okay wow that's that's really interesting um i guess my question would be is it common then to have like a hybrid cooling system like some air and then some [00:30:56] Speaker 1: liquid um actually nowadays we should have hybrid already because if we look at the the ai iraq right it's about 80 to 85 percent gonna be heat from the liquid however that is still about 10 to 15 percent from the air because inside the server there's still some the the air cool component let's say like those hard drive or the lamp or those kind of some the network equipment so basically we need to prepare the hybrid solution even we have the liquid solution already but some component in a data horse still require the air cools yeah that need to be prepared but the question gonna be how many percentage right okay have to prepare right because someone say yeah if like that i prepare only 20 percent air cool and then the other 80 percent for liquid cool right uh the question also comes to my mind because like two years ago when nvidia released the gb100 right that time they also say it doesn't have the air cool option so it's just liquid yeah people move to the liquid but just like one or two year later they released the air cool version so they're also adopting yeah right so yeah that i think that's quite challenge on on the data center operator as well yeah someone talk about now they're preparing 50 50 just in case you know they have the new generation of the gpu that not require the liquid anymore the air goes too much easier to maintain right to to to take care of those things yeah yeah i guess it would all [00:32:38] Speaker 2: you know depend on like the ai needs in the air center right so you know on top of like these design considerations maybe a question to partner with them so how can like these operators and you know their technology partners strike the right balance between the performance the energy efficiency and also the operation cost when deploying all of these cooling solutions yeah i think uh based on my experience [00:33:06] Speaker 3: there is no cooling solutions that works for uh every data centers right we need to uh when when we design the data centers we need to think about these uh three things uh how efficient is the equipment and then um how about the long term um when you invest some things of course the investors want uh if when is the the payback time right we need to consider these things also and then uh how about the performance about the cooling right they also worry uh if i put it in my data centers then the next two years oh change generation again again my investment will be uh like so for what i need to invest so much [00:33:55] Speaker 1: yeah that's also the fact that i have a cash you know that if you see that the growth of the the gpu generation right is quite quick as well yeah we are already out there there right and i haven't touched on the gpu and tpu here you know because nowadays our data center are using for the training right we train we we use for ai training okay the load profiles of training it's different from the load profile of inference wow okay yeah and then when we talk about inference that actually all the requirement in terms of power and cooling is totally different oh is that even higher um it it actually lower in terms of the power densities but the load profile is different yes yeah and uh the the reliability requirement is different yeah we haven't touched that part yet also create the challenge on [00:34:54] Speaker 2: data center operator yeah okay yeah i mean of course there's no one size fits all solution so i mean it requires really a lot of research beforehand and a lot of consideration and really you know constructing it in order to make it reliable and efficient at the same time all right i mean we talked about power and also you know cooling systems which also uses a lot of water and so i guess the next topic that we should talk about now is about sustainability you know growing with responsibility right and you know with all of these increasing concerns around the energy consumption the water usage carbon emissions so what role is delta playing in helping data center industry [00:35:45] Speaker 1: grow um to grow more sustainably i would start from our founder philosophy mr bojang he actually realized when we delta do the power supply business or power electronics right he really think about delta should take responsibility in terms of how to create our product to use energy as efficient as possible so if you look at the our product actually we really emphasize on the energy efficiency yeah in terms of like how is the power supply the power supply is the power supply or the ups or those power electronic things generate lesser heat than the the competitor in the market that's what we emphasize and if you look at delta actually delta joining the cop 21 which is we talk about the the environment where i love to now sustainability yeah sustainability until now the cop 30 delta also have the energy infrastructure because we we also look at in the future that's gonna be really the silency energy shortage yeah so a data center operator might have to look at how to generate power by yourself yeah so i don't know how to do that yeah so that that's what we are preparing uh in terms of from from delta uh technology preparation yeah to support how the uh data center can more uh sustainability in terms of the operation of your data centers in the future [00:37:20] Speaker 2: which is yeah the potential shortage is something to think about definitely and also you know like the hyperscale customers and all of these big data center players they're also increasingly aware of the importance and are requesting for you know clean energy yeah like where the power is sourced from they request for renewable energy yeah and so yeah we all need to take that into consideration when we are constructing the data center and of course now to panawi so uh what can like the data centers and um technology partners what can they do to ensure that the growth can be achieved without compromising [00:38:03] Speaker 3: all of these system sustainability goals yeah i think first for those companies like delta and delta has did the right things uh first you can still continue to your rnd team right to develop good efficiencies products and then um in power and cooling technology okay and then the second things uh work with the community i mean like the association or or the ecosystems how to develop the right solution or the good good efficiencies products and then i think the third one is the most important things um trying to uh invest in talent development uh yeah yeah because you have such good products or some things but if you don't have talent to to run this yeah how to install or or with with good knowledge it's also a a big challenge yeah i think that's the the approach and that's yeah i mean as you have correctly pointed it out um you know having a strong talent pipeline is really important [00:39:21] Speaker 2: because all of these innovations and these technological developments they're driven by the knowledge of the people right that's why i think that we in southeast asia and indonesia as well like we need to focus on building you know a good local um talent actually in order to create all of these innovations so i see that actually delta has been actively contributing to talent development like i think that um if i'm not mistaken they have scholarship programs and also the delta electronic foundations could you tell us a little bit more about that and the impact that these efforts are hoping to create [00:40:06] Speaker 1: yeah but basically we we see um right now the when the student right graduate from the institution right um the way of the the the university or the school in in our region still have some gap when they come out to work with the real industry yeah so we we have some program where we try to engage with the university to get the students you know to be trained or to be work on the real projects that very relevant to the industrial requirement and then we even build uh some of the the lab in the new university yes to let the student practice what really they would have to experience or to face when they graduate and go to the the real uh the real jobs like example we have the the program that we let all the university to build a robot you know not only in particular country they have to make like uh the contest in the the country to compete each other and we find the champ the champion of each country to have the another confidence yeah and then we we find out who is the best in term building that the robotic as well [00:41:24] Speaker 2: yeah yeah is do they have that in indonesia the competition not yet in indonesia but we we can [00:41:32] Speaker 1: also consider yeah it's a consider but for a data center as well like i say data center is the very complex even when we talk about the ai right it even more complex and to design data center or to operate data center right that is multiple discipline it's not only power anymore right you have to have the knowledge of power of cooling of monitoring system of even how to you know to do like uh some kind of the the it skill right to to to check all those things so that really a shortage in the region not only in particular in indonesia but actually in the region right so we need to work together uh with the institution yeah to have more people trained to uh know how to design how to operate our data center some country we already start already to join the association or to join the institution and then form a team as what should be the how to say uh data center 101 we call data center 101 right start from 101 and then we start from how is the the country standard right to to design data center like tier one tier two three three or four and they start to have the training program with the local institution so that yeah we can train the people in the industry wow yeah yeah i think that it's important to also [00:43:07] Speaker 2: understand first how a data center works like for me as well it took me a while as well to really understand how a data center works and then after understanding how it works and also the industry then all the technical knowledge would come into play and it will all integrate and then it would make sense right i'm just curious so um you were mentioning about different disciples right so if we talk about universities and faculties what would be like the relevant majors or faculties that the [00:43:37] Speaker 1: students should enroll this good question that's why i say for data center it's multiple discipline yeah the one who only trained for power might not be able to cover a whole data in there similar thing if the one only trained for mechanicals it also not able to cover a whole data center right but it's a pity that right now it doesn't have any curriculum that can cover yeah because it's very specific yeah it's very specific so but some countries start to have this kind of the data center engineering curriculum already do we have that in indonesia not yet not really yeah we might need to [00:44:16] Speaker 2: consider and that yeah yeah yeah i think mainly the people who work at the data center their their [00:44:23] Speaker 1: background is like engineering yeah like electric or engineering computer or mechanical engineer yeah yeah yeah but they doesn't like i say it cover electrical mechanical civil computing right yeah the cloud yeah [00:44:38] Speaker 2: yeah or those it yeah yeah okay so that's from uh delta's perspective right maybe let's now move on to panawi maybe you know beyond all of these like classroom learnings what are other you know practical [00:44:53] Speaker 3: hands-on training opportunities that we need to develop same like sakda said before they put a laboratories i think this is a good idea in the universities and let the the student do some projects right and and let their knowledge more creative right and this data centers is unique same i agree with sakda that data centers not just mechanical electrical discipline it is it's from architect structure um civil um civil and then mechanical electrical ict security security security so many things it has all my fiber security right yeah it's also an important yeah yeah uh in indonesia we are not approaching to uh what we call it university yet because yeah even we don't have that the faculties or that that studies yeah exactly for data centers but luckily we have nusantara yeah nusantara we are approaching to um vocational school right then we work together with vocational school to do this this training train train the student become at least upscale their their their their their skill become oh i i can become a technician i become an operator operational guys in data centers then it will be uh contribute to to our countries then uh this lack of um what we've had talent yeah it can be filled in from this uh vocational schools okay the vocational school yeah learning in vocational school is not to become a welders become a world because anything in very late industries technology but now you need to learn something new right data centers built many many many uh hyperskills built in indonesia why we just become like [00:47:11] Speaker 4: uh uh uh uh what we call it yes oh yeah there's data centers why you don't need to be you you should [00:47:18] Speaker 2: contribute more yeah actually yeah but i guess maybe the the first step is being aware of what a data center is okay now thinking of the bigger picture now maybe a question to both um so what advice would you give to you know regulators industrial players academic institutions if indonesia wants to become a leading digital infrastructure in southeast asia [00:47:49] Speaker 1: i think like like like we we touch at the beginning of the podcast today right prepare enough power prepare enough power and then not only power now there's some lesson learned for some country in the region where they prepare enough power but they don't prepare enough water so actually data center consume these two things right water or the power and the water as well so prepare these two to be sufficient right and i would also uh i mean suggest to to form the ecosystem right in terms of the data center industry because you also can have more power to talk to those power or utility company yeah and convince them in terms of create more power distribution yeah right or we have some zoning that purposely for our data center in the indonesia so that they can make sure their investment to that particular area they can find the returns of investment or even some of the private power generation company want to join that they can do easier flexibility in terms of the investment right yeah and the last part could be on the talent preparation yeah we just touched right that's still the the major challenge not only for indonesia but actually the whole world yeah yeah because like i say data center require a very complex design and they make multiple disciplines yeah who who know how to operate how to maintain this data center so preparing the talent preparing the those uh uh educational uh people who are ready to work in this industry is also a very crucial yeah yeah yeah it's it's a long game it's it's a long game it's a long game but in terms of investment i think probably what we also have to look at how the the country can support those overseas players to come to indonesia right what are the the benefit what are these uh the thing that if they come more to indonesia they can they can have uh a more benefit there right it's this to compete between multiple countries if you look at the region like malaysia or like thailand there are a lot of a program from the the government to support a digital economy yeah and then try to attract those a big digital guy from the america or from yeah you know right come to invest in the country that could be another thing that the government have to look at okay now that you [00:50:28] Speaker 2: mentioned other countries maybe do you mind sharing a bit based on your experience so how does indonesia compare to these neighboring countries like you know singapore malaysia or even other countries that [00:50:40] Speaker 1: you've worked in i would say like for singapore now the because the land is limited the power is limited right but but the point is that singapore is still the best connectivity hub for those things so now for singapore government yeah they try to like to be leader in terms of innovation mama data center right the program that they release it just to attract the one who really create a really efficient or really innovative data center to be a new build in singapore but if you look at like malaysia they prepare enough power they prepare enough land and try to attract the the investor to go there and try to make very investor friendly environment to attract more and more investment and thailand start to do so we bring in more connectivity there we have the zoning for data center and then we create more power distribution to support support that's how start to attract more investors and come to the country so i think for indonesia uh we we we might have to look at this and then make more competitive to attract more investors to come [00:51:51] Speaker 2: yeah okay uh so after we've heard the experiences from the other countries how do you think indonesia could capitalize our opportunities because we have more land in comparison to other countries we also have the market we have the scale so how do we capitalize on these opportunities in order to become more competitive to [00:52:11] Speaker 3: become a leading infrastructure infrastructure yeah uh following sakda said before every country facing the same problem power ability sustainability talent and others right but for me i think indonesia is the most important important things we need uh the regulatory certainty how to attract the investors coming to indonesia if you have this i think indonesia will be leading uh to attract the investors coming to indonesia the second one you need uh build power infrastructures like previous we said this power is not uh just a trn responsibility but the whole stakeholders responsibility right the third thing is um talent you need to create more talent in data centers right and the fourth thing i think is the sustainability yeah yeah you take it from like uh from from the the the the what we call it uh natural yeah you need to give it back in uh like like you said yeah yeah it really [00:53:29] Speaker 2: requires a lot of proper planning and then executing them in the proper manner and then start thinking about how we can give back to the nature after so many things that we have taken from them right yeah and yeah i guess i mean right now with you know ai changing how data centers are built and also the landscape in general then the goal or the challenge is no longer building more data centers but how reliable and sustainable we can build them right yeah so yeah i think this should be definitely or is already a you know a key national priority program right yeah yeah okay we unfortunately have reached the end of our discussion for today and i guess do you have any closing remarks from both pasak and panawi [00:54:20] Speaker 1: yeah i i could say the the ai gonna change a lot of things in a data center requirement and then we need to prepare to be ready in terms of to how to serve the ai data center to be in this country no matter in terms of the ecosystem population the sustainability population and also the tarant population yeah [00:54:47] Speaker 3: i see from my side i think uh we need to have more stronger collaboration right between uh government and then uh industry academia and also with the technology providers i think it should be have have [00:55:05] Speaker 2: uh have stronger with with that yeah yeah i think that the public and private you know partnerships could like propel us further into becoming or creating a stronger digital infrastructure all right thank you so much for all the insights today i myself find it very very you know useful and it adds more to my knowledge thank you so much thank you thank you all right uh before we conclude let's do the nusantara academy signature pose salam unity salam unity to our audience don't forget to like comment subscribe to nusantara academy's channel in order to stay updated on the latest trends in digital infrastructure thank you [00:55:49] Speaker 5: untuk terus mendapatkan insight dan mengikuti perkembangan terbaru seputar data center dan aktifkan notifikasi bagikan juga pendapat atau pertanyaan anda di kolom komentar dengarkan episode lengkap podcast nusantara di spotify sampai jumpa di episode berikutnya [00:56:11] Speaker ?: selamat menikmati

Transcribe Any Video or Podcast — Free

Paste a URL and get a full AI-powered transcript in minutes. Try ScribeHawk →