Take-Two's CEO predicts streaming will 10x gaming's install base within three years

[Daniel Sims]

Looking ahead: As the RAM crisis makes high-end hardware increasingly unaffordable, it's unsurprising to hear companies float streaming as a possible alternative for gaming. Despite the technology's past failures, the head of Grand Theft Auto publisher Take-Two is bullish on its future.

During Take-Two's first-quarter fiscal 2027 earnings call, CEO Strauss Zelnick put a number on it: he expects cloud gaming to reach commercial, low-latency viability within three years, potentially multiplying the effective installed base tenfold. He made the comments while addressing the rising cost of gaming hardware.

Due to tariffs, rising costs of TSMC's semiconductors, and memory shortages stemming from AI data center construction, game consoles and PC components have gotten more expensive rather than cheaper over the past several years. The Xbox Series consoles and the PlayStation 5 launched at $500, but their successors might double that price.

Zelnick believes the popularity of Take-Two's major franchises, likely referring to Grand Theft Auto and NBA 2K, will cushion the impact of hardware price hikes, though he acknowledged the price hikes could still limit the company's addressable market.

Cloud gaming, meanwhile, has long been floated as a route for bringing high-end games to low-cost devices, and Zelnick believes its moment is finally near.

The CEO argued that advances in networking technology and infrastructure will make cloud gaming more viable than ever for a wider swath of users, potentially growing the installed base by an order of magnitude. While he doesn't expect revenue to climb by a similar amount, he sees it as a way to reach players beyond Take-Two's traditional core market.

According to Zelnick, those advances are what will separate this next wave from the technology's most spectacular recent failure, Google Stadia. The major cloud gaming players today are Amazon's Luna, Nvidia's GeForce Now, Microsoft's cloud functionality for Xbox Game Pass, and Sony's PS5 cloud streaming through PlayStation Plus Premium.

The latter two have proven that streaming can succeed as a complementary feature for games users already own, but success via other business models remains elusive.

Strauss Zelnick noted that 90% of Take-Two's business is already digital, though he left the door open for occasional physical releases going forward, comparing the practice to the enduring demand for vinyl records.

Zelnick also weighed in on Grand Theft Auto VI's upcoming digital-only launch and the broader industry shift away from physical disc games on PlayStation. He noted that 90% of Take-Two's business is already digital, though he left the door open for occasional physical releases going forward, comparing the practice to the enduring demand for vinyl records.

Unlike vinyl (and music CDs), though, no single company controls the pressing of physical games; console manufacturers alone manufacture their platforms' discs, which makes a comparable resurgence less likely.

Grand Theft Auto VI is set to launch on November 19 for PlayStation 5 and Xbox Series consoles, with a PC version expected sometime later. An extended early look at the game will premiere exclusively on Netflix on August 27.

Permalink to story:

Take-Two's CEO predicts streaming will 10x gaming's install base within three years

 

[Alfonso Maruccia]

"This time is different, trust me" - OnLive's corpse speaking through its breathing puppet.

 

[NumberNine]

The only way streaming can increase any install base is if the user is forced to buy the game and install it on the server themselves. If you're streaming the game you're not installing it, are you?

 

[NumberNine]

"This time is different, trust me" - OnLive's corpse speaking through its breathing puppet.

By the ghost of Stadia, they'll never learn Alfonso.

 

[brucek]

"advances in networking technology and infrastructure will make cloud gaming more viable"

Such as? The primary issue is latency. Latency from distance is not changing anytime soon. Latency from delayed server processing was not previously characterized as the main problem; and in any event, in a current era where server hardware for the same capacity is getting more vs. less expensive, this doesn't seem like it's going to solve a gaming platform affordability problem either.

The only part I maybe agree on is that consumers who simply could not afford a gaming device will at least be more motivated to give streaming a try. Whether they'll like it, or be able to afford it at current market prices for the capacity required, I don't know.

 

[Loadedaxe]

"advances in networking technology and infrastructure will make cloud gaming more viable"

Such as? The primary issue is latency. Latency from distance is not changing anytime soon. Latency from delayed server processing was not previously characterized as the main problem; and in any event, in a current era where server hardware for the same capacity is getting more vs. less expensive, this doesn't seem like it's going to solve a gaming platform affordability problem either.

The only part I maybe agree on is that consumers who simply could not afford a gaming device will at least be more motivated to give streaming a try. Whether they'll like it, or be able to afford it at current market prices for the capacity required, I don't know.

You are not completely wrong about latency, but I think you are defining the problem too narrowly.

Nobody is going to invent faster than light networking. Distance will always impose a latency floor. But cloud gaming doesn’t improve only by making signals travel faster. It improves by putting data centers and edge nodes physically closer to users, improving routing and peering, reducing wireless last mile latency, improving codecs and encode/decode times, and reducing the server side time between receiving an input and producing the next frame.

That’s why “advances in networking technology and infrastructure” actually does mean something here. Research on cloud gaming already shows that network quality, server placement, bandwidth stability, wireless conditions and congestion management all materially affect the experience.

Where I agree with you most is the economics. Streaming doesn’t magically eliminate expensive gaming hardware...it moves a large portion of that hardware into somebody else’s data center and adds bandwidth, power and infrastructure costs. And cloud gaming can require tens of Mbps continuously rather than the tiny amount of network traffic needed for locally rendered gameplay.

So I can absolutely see streaming expanding gaming to people who don’t want to buy a console or gaming PC. But Zelnick’s idea that the potential install base could become 10X larger within three years is a much bigger leap. Even his fuller statement makes clear he isn’t predicting 10X revenue, only a materially larger addressable audience.

There’s a legitimate technological trend here. There’s also a CEO doing what CEOs occasionally do...extrapolating that trend all the way to the moon.

 

[Axiarus]

No

 

[Axeia]

Interesting timing for such a statement.

he expects cloud gaming to reach commercial, low-latency viability within three years, potentially multiplying the effective installed base tenfold

Based on what?
All the AI servers are going to run games and be installed with a direct glass fiber link towards our homes that are nearby?

Because NVIDIA is already offering a 'commercial, low-latency' viable solution right now to quite a few people. Afaik there's no new tech on the horizon that would lower latency of the internet connection any further (other than simply having more datacenters available increasing the odds of them being geographically closer) and pretty much all other latency in the entire chain is of minimal influence / won't be improved that much any time soon.

The only thing that can make it more appealing than it is at the moment is higher prices of consoles and PCs still. But even then I don't see how it would 10x cloud gaming, we still have our old PC/Console and if no one is upgrading and you make your title so that it only runs on the latest hardware or cloud gaming a good chunk of people will simply ignore it.

The CEO argued that advances in networking technology and infrastructure will make cloud gaming more viable than ever for a wider swath of users.
....
According to Zelnick, those advances are what will separate this next wave from the technology's most spectacular recent failure, Google Stadia.

What advances?
WiFi 7 has matured since then but ehhh how many people actually have a device they want to game on with that along with a router that actually supports it? And even then what's the benefit over wifi 6 in that case?
Did that many more people get a faster/lower latency connection since then? I had glass fiber in 2023 and I still have it, no data centers that would host the hardware for this have been opened any closer to me. Think it's the same story for the vast majority of people, not that much has changed network/network infrastructure wise since 2023.

Perhaps he's one of the people that think 5G connections are some magical thing that now that the rollout is more complete and basically every phone ships with it that we can have low latency mobile cloud gaming?
If the game isn't natively available on mobile, I don't want to play it on my phone in the first place and I think that goes for most people. Besides that... 5G is flakey AF, what's worse than a bit of latency is inconsistent latency. I'll take a consistent 80ms over one that bounces between 10 and 70 constantly. I got pretty decent results even in multiplayer FPS with a consistent ~180ms, my brain can adapt to a constant delay but not to inconsistent behavior.

 

[Vixiy]

Where are these people who have been holding off playing on their phones until the games are laggier and fuzzier?

 

[Kirby1]

Maybe 2x with the chip shortages. 10x? I wish I still got that high.

 

[starfalls]

I dont care even if that tech had 1 latency. Id still rather own my computer/console. Still rather not use net if I have to. I dont have net 24/7, especially if I go to a vacation place/house. I live in the middle of a big city and my net dies/goes slow every 3-4 days. Just no. I cant imagine what people in villages would do

 

[bviktor]

Just as tone deaf as the AI CEOs

 

[user556]

Hehe, all those failed AI data centre deployments will need to recoup at least some of the losses. What better way than running games on overpowered GPUs? Certainly more productive than gimmicky AI.

 

[Theinsanegamer]

Bandwidth and latency restrictions would like to have a word.

DisplayPort pushed 80Gbps for a reason. It's not for the lulz. You're always going to have that 20-50 MS extra latency from cloud computing no matter what, plus all the compression.

It slike streaming. Even the best streaming services that claim 4k don't come close to the visual acuity, or more importantly, the audio experience, of a 1080p Blu Ray, let alone a 4k Ultra Blu Ray. And short of developing worm hole technology, that isn't gonna change.

 

[Axeia]

Hehe, all those failed AI data centre deployments will need to recoup at least some of the losses. What better way than running games on overpowered GPUs? Certainly more productive than gimmicky AI.

Not sure if that could be done in any cost effective way. The AI cards lack DirectX support, have no media encoders and are typically sold in rack setups that are seriously lacking in CPU power for gaming.
The buildings themselves would still serve a purpose but I don't think much of what is in them would.

 

[godrilla]

This 10x figure seems to be a hallucinating one. Too many blind spots to count.
1) Streaming is not net zero cost both for the end user and the data server end. Cost of hardware is also going up on the server data center end and needs to be upkept via newer more efficient hardware, software management, network management etc. Power is also a limited resource even if you use renewable energy unless you get into nuclear generators which isn't going to be a 3 year project. Competing revenue streams have to taken into consideration especially if someone else is going to rent at a higher rate then gamers are willing to pay for those data centers.
2) Quality of Streaming is always worse, via artifacts, latency, softer images and likely lower bit rate as well as lower resolution content than Streaming locally.
3) Streaming single player games for a shorter period of time vs online multiplayer with checks gtav online stats with 1000s or even 10,000 of hours played doesn't make sense for Streaming. You will recuperate the cost of local hardware more like 10x at 10,000 hours on a enthusiast level local hardware vs Streaming the same online multiplayer game at the maximum quality allowed.
3)Imagine paying $100 for a game then paying a monthly subscription fee to access the game for Streaming. Also those subscription fees go up over time eg. Game pass and most all Streaming services in the past 5 years.
4) No unlimited access for binge gaming for a holiday weekend unless you use up your dedicated hours of allocation or wait In a line.
5) Servers can also be offline from attacks, updates, resetting servers. If all your streaming are in one basket then you have to wait for the servers to go back up. GeForce now has reported server shutdowns.
6) using a 0% interest credit card x 21 payments while costing more initially will often get you better hardware, better image quality and save cost long term while enjoying the game at higher quality with low latency without any wait in lines or waiting for servers to go back up.
7) Sure some might delve into streaming via desperation from current hardware inflation but given the opportunity via sales like Black Friday especially on yesterday's tech will likely prevent the 10x streaming in my outlook.


Streaming is something you want to try and then go ah that's why local dedicated hardware is so much better.

 

[starfalls]

AND I predict ill get GTA 6 for free on the PC, just like I did with GTA 5 (epic game store helloww)

I sure aint paying for an old spoiled to hell game. There will be a billion GTA 6 videos and spoilers when it comes out. NAAH.

 

[dangh]

The only way this could become true is even higher cost of ram and drives. They strongly don't want to to own anything. With a demise of physical gaming we already are not in control, making it all subscription and streaming will make everyone a prefect consumer.

 

[RudyBob]

Data Centers anyone? Or won't those be needed for games?

 

[Theinsanegamer]

The only way this could become true is even higher cost of ram and drives. They strongly don't want to to own anything. With a demise of physical gaming we already are not in control, making it all subscription and streaming will make everyone a prefect consumer.

And how will people play these streamed games then? One of the best streaming services is GeForce now, and you need a GeForce GPU to get the best experience on the user end.

So you'll still need a reasonable PC. Some 2 core 4gb shitbox won't be able to stream well at all.

 

[toooooot]

It would have to be fairly cheap. Xbox already stumbled on this rock. It is too expensive to keep that big catalog and constantly adding new titles at the price they have.
So, on one side, such services needs to be very cheap, at least for the first few years.
On the other, hardware costs are insane. This idea got a lot worse since AI started to use all of the chips.
I can see this thing working, especially if it is as simple as turning on a tv and begin playing, while paying a very modest sum every month.

But in current time, it is an insanely expensive business to do. So expensive that is is worth waiting till the prices stabilize.

 

[axiomatic13]

Cloud gaming implies a subscription. Networking advances aside... still no.

 

[dangh]

And how will people play these streamed games then? One of the best streaming services is GeForce now, and you need a GeForce GPU to get the best experience on the user end.

So you'll still need a reasonable PC. Some 2 core 4gb shitbox won't be able to stream well at all.

You can stream on any android device, nvidia shield, or a mini pc with 4gb ram which is way too little for any local gaming.
You don't need reasonable pc. 2 core 4gb shitbox will do, if there is hardware h264 or h265 decoder. Any intel n100 or n95 will do great.
You dont need geforce gpu either. People would quickly accept new normal, companies will adjust for a simple arm endpoints, actually geforce already did. Small improvements are simple to push.

 

[Theinsanegamer]

You can stream on any android device, nvidia shield, or a mini pc with 4gb ram which is way too little for any local gaming.
You don't need reasonable pc. 2 core 4gb shitbox will do, if there is hardware h264 or h265 decoder. Any intel n100 or n95 will do great.
You dont need geforce gpu either. People would quickly accept new normal, companies will adjust for a simple arm endpoints, actually geforce already did. Small improvements are simple to push.

Cloud gaming runs like **** on shitboxes. It runs worse on **** phones.

 

[Axeia]

Cloud gaming runs like **** on shitboxes. It runs worse on **** phones.

All the local device has to do is
1. Display the video stream (h264/h265/AV1). Phone will be fine.
2. Handle input. Phone kinda sucks for that, you might end up using Bluetooth which isn't the greatest for latency.
3. A stable high bandwidth internet connection. Phone kinda sucks for that as you're probably using WiFi.

Now if you have a phone dock providing a wired connection for input and internet and ideally output to a bigger display it could be just as good as a PC for cloud gaming.

I'm surprised non of the TV manufacturers have tried to get into bed with Nvidia yet. Integrate what the shield is ATM into a TV and add 6 months of GeForce now and advertise it as the gaming TV.