The First Xbox Handheld. Can Intel afford to lose the console/handheld market?

[soAsian]

It will use the AMD Ryzen AI Z2 Extreme processor. Both current-gen Xbox and PlayStation consoles use AMD Zen 2.

ARM has taken over the mobile/tablet/phone market. Cloud providers like AWS are coming out with their own ARM-based servers.
Qualcomm is getting into Windows PCs/laptops.

Intel faces AMD and ARM for PC and server. Is the console/handheld profit margin just too low for Intel? Can Intel afford not to be in the console/handheld market?

 

[Maximus]

Intel faces AMD and ARM for PC and server. Is the console/handheld profit margin just too low for Intel? Can Intel afford not to be in the console/handheld market?

The console and handheld markets are neither large enough nor sufficiently profitable for Intel. The gross margin falls well below the 50% threshold recently established for new product launches. While this sector was crucial for AMD due to its scale at the time, it does not offer the same strategic value for Intel.

 

[Daniel Nenni]

The console and handheld markets are neither large enough nor sufficiently profitable for Intel. The gross margin falls well below the 50% threshold recently established for new product launches. While this sector was crucial for AMD due to its scale at the time, it does not offer the same strategic value for Intel.

Agreed, I don't think this market is the best use of Intel resources. Mobile (low power) devices historically have been a challenge for Intel. This monster will probably keep you warm at night.

 

[siliconbruh999]

Agreed, I don't think this market is the best use of Intel resources. Mobile (low power) devices historically have been a challenge for Intel. This monster will probably keep you warm at night.

Lunar Lake is one such product that Excels at low power and the challenge is mainly due to their methodology of the raw performance that they like over anything else.

 

[hist78]

The console and handheld markets are neither large enough nor sufficiently profitable for Intel. The gross margin falls well below the 50% threshold recently established for new product launches. While this sector was crucial for AMD due to its scale at the time, it does not offer the same strategic value for Intel.

I can’t say for certain whether the profit margin would be attractive enough for Intel to re-enter the game console market, but in terms of volume, is the console market really too small for Intel?

For example, the Sony PlayStation 5, released in 2020, sold approximately 18.5 million units in the fiscal year 2024 alone.

By comparison, a total of 262.7 million PCs were sold worldwide in 2024, with Apple accounting for 22.9 million units of its Mac PCs and notebooks.

Considering that the PlayStation 5 sold 18.5 million units in a single year, four years after launch, this volume for a single console product line doesn’t seem too small to me.

2024 Worldwide PC Shipments

PC worldwide shipments are in millions of units

Source: https://my.idc.com/getdoc.jsp?containerId=prUS53061925#:~:text=For the full year, PC,outlook and difficult demand planning.

 

[Maximus]

I can’t say for certain whether the profit margin would be attractive enough for Intel to re-enter the game console market, but in terms of volume, is the console market really too small for Intel?

For example, the Sony PlayStation 5, released in 2020, sold approximately 18.5 million units in the fiscal year 2024 alone.

By comparison, a total of 262.7 million PCs were sold worldwide in 2024, with Apple accounting for 22.9 million units of its Mac PCs and notebooks.

Considering that the PlayStation 5 sold 18.5 million units in a single year, four years after launch, this volume for a single console product line doesn’t seem too small to me.

AMD Reports Fourth Quarter and Full Year 2024 Financial Results

SANTA CLARA, Calif., Feb. 04, 2025 (GLOBE NEWSWIRE) -- AMD (NASDAQ:AMD) today announced financial results for the fourth quarter and full year…...

ir.amd.com

Thanks for the volume reference. My focus is primarily on revenue.

Based on AMD's financial reports, its gaming segment revenue for 2024 was $2.6 billion, reflecting a 58% decline from the previous year, mainly due to a drop in semi-custom revenue.

Considering Intel's current annual revenue of approximately $50 billion, AMD's entire gaming segment accounts for about 5% of Intel's total revenue.

AMD Reports Fourth Quarter and Full Year 2024 Financial Results

SANTA CLARA, Calif., Feb. 04, 2025 (GLOBE NEWSWIRE) -- AMD (NASDAQ:AMD) today announced financial results for the fourth quarter and full year…...

ir.amd.com

 

[hist78]

Thanks for the volume reference. My focus is primarily on revenue.

Based on AMD's financial reports, its gaming segment revenue for 2024 was $2.6 billion, reflecting a 58% decline from the previous year, mainly due to a drop in semi-custom revenue.

Considering Intel's current annual revenue of approximately $50 billion, AMD's entire gaming segment accounts for about 5% of Intel's total revenue.

AMD Reports Fourth Quarter and Full Year 2024 Financial Results

SANTA CLARA, Calif., Feb. 04, 2025 (GLOBE NEWSWIRE) -- AMD (NASDAQ:AMD) today announced financial results for the fourth quarter and full year…...

ir.amd.com

If we are looking into the revenue, should we compare one AMD console product revenue to one Intel product revenue?

If 5% of the revenue is the threshold to decide if an Intel product deserves to keep, Intel can only sell up to 20 products. If the threshold move up to 10%, then Intel is limited to sell up to 10 products.

Because Intel has so many products, I guess some of them may not contribute revenue big enough to surpass 5% of the Intel total sales at all.

 

[Maximus]

If we are looking into the revenue, should we compare one AMD console product revenue to one Intel product revenue?

If 5% of the revenue is the threshold to decide if an Intel product deserves to keep, Intel can only sell up to 20 products. If the threshold move up to 10%, then Intel is limited to sell up to 10 products.

Because Intel has so many products, I guess some of them may not contribute revenue big enough to surpass 5% of the Intel total sales at all.

Yes, the volume and revenue of a single console or handheld product could surpass many of Intel's products. However, AMD supplies all PS/Xbox CPUs (while Switch uses Nvidia's CPU), and its gaming segment revenue still amounts to only 5% of Intel's total revenue, indicating a limited total addressable market (TAM).

The cost constraints in the console and handheld markets restrict CPU pricing, which is why most of these processors are manufactured using TSMC's N-2 or even older nodes. If Intel were to enter this space, it would likely need to outsource production to TSMC or Samsung, meaning this segment wouldn't contribute to Intel's foundry business either.

Overall, key challenges include:

1. 1. Limited TAM
2. 2. Low profit margins
3. 3. Long product generation cycles (typically 5-7 years per generation)
4. 4. Established partnerships—Microsoft and Sony have collaborated with AMD for over a decade. Unlike cutting-edge semiconductor advancements, console and handheld development relies on long-term customization and stability. Given Intel's recent trend of discontinuing non-core or less profitable projects, it's unlikely that Microsoft and Sony would feel confident working with Intel for future console and handheld generations.

 

[hist78]

Yes, the volume and revenue of a single console or handheld product could surpass many of Intel's products. However, AMD supplies all PS/Xbox CPUs (while Switch uses Nvidia's CPU), and its gaming segment revenue still amounts to only 5% of Intel's total revenue, indicating a limited total addressable market (TAM).

The cost constraints in the console and handheld markets restrict CPU pricing, which is why most of these processors are manufactured using TSMC's N-2 or even older nodes. If Intel were to enter this space, it would likely need to outsource production to TSMC or Samsung, meaning this segment wouldn't contribute to Intel's foundry business either.

Overall, key challenges include:

1. 1. Limited TAM
2. 2. Low profit margins
3. 3. Long product generation cycles (typically 5-7 years per generation)
4. 4. Established partnerships—Microsoft and Sony have collaborated with AMD for over a decade. Unlike cutting-edge semiconductor advancements, console and handheld development relies on long-term customization and stability. Given Intel's recent trend of discontinuing non-core or less profitable projects, it's unlikely that Microsoft and Sony would feel confident working with Intel for future console and handheld generations.

"1. Limited TAM
2. Low profit margins
3. Long product generation cycles (typically 5-7 years per generation)
4. Established partnership"

For "#3. Long product generation cycles", Isn't it good for a semiconductor company to have some of its products to enjoy a longer product cycle? It can give them longer time to recoup the high cost of development and manufacturing.

 

[Maximus]

"1. Limited TAM
2. Low profit margins
3. Long product generation cycles (typically 5-7 years per generation)
4. Established partnership"

For "#3. Long product generation cycles", Isn't it good for a semiconductor company to have some its products to enjoy a longer product cycle? It can give them longer time to recoup the high cost of development and manufacturing.

Long product generation cycles are a double-edged sword. As you pointed out, securing a win means enjoying its benefits for an extended period. However, the key assumption is "if you win."
The combination of long product cycles and a highly limited customer base—essentially just Microsoft and Sony—results in very few opportunities. Unlike CPU/GPU sales, which provide consistent revenue with yearly chances to adjust pricing and boost sales, the console market operates on an all-or-nothing model. It's either a win or zero, and the next opportunity won’t come around for several years.

 

[siliconbruh999]

Long product generation cycles are a double-edged sword. As you pointed out, securing a win means enjoying its benefits for an extended period. However, the key assumption is "if you win."
The combination of long product cycles and a highly limited customer base—essentially just Microsoft and Sony—results in very few opportunities. Unlike CPU/GPU sales, which provide consistent revenue with yearly chances to adjust pricing and boost sales, the console market operates on an all-or-nothing model. It's either a win or zero, and the next opportunity won’t come around for several years.

For Intel it's a win imo even If the product is not big on product side the foundry side is huge.

The PS5 Die is 260mm2 N6 based on this image I was able to get the die size of 21.2*12.3 mm2.

https://twitter.com/x/status/1829881334497694103

At 0.1 D0 it would be roughly 155 DPW I am only counting good die as there is no cut down version

Based on this Sony sold 18.5 Million units last year and 1 wafer gives 155 Dies.
It would require 116K Wafers and dividing it by 12 it would be roughly 9.6K WSPM.

For Intel the wafer volume will go to fabs and help with utilization
Here is the unit sold data

Sony PlayStation 5 sales worldwide by quarter 2025| Statista

In the fourth fiscal quarter of 2024, Sony Corporation sold approximately 2.8 million PlayStation 5 gaming consoles.

www.statista.com

 

[Maximus]

For Intel it's a win imo even If the product is not big on product side the foundry side is huge.

The PS5 Die is 260mm2 N6 based on this image I was able to get the die size of 21.2*12.3 mm2.

https://twitter.com/x/status/1829881334497694103

At 0.1 D0 it would be roughly 155 DPW I am only counting good die as there is no cut down version

Based on this Sony sold 18.5 Million units last year and 1 wafer gives 155 Dies.
It would require 116K Wafers and dividing it by 12 it would be roughly 9.6K WSPM.

For Intel the wafer volume will go to fabs and help with utilization
Here is the unit sold data

Sony PlayStation 5 sales worldwide by quarter 2025| Statista

In the fourth fiscal quarter of 2024, Sony Corporation sold approximately 2.8 million PlayStation 5 gaming consoles.

www.statista.com

N6 is part of the 7nm family. Based on Intel's own slide, the wafer cost of Intel 7 is at a significant disadvantage compared to N7.

 

[siliconbruh999]

View attachment 3252
N6 is part of the 7nm family. Based on Intel's own slide, the wafer cost of Intel 7 is at a significant disadvantage compared to N7.

I was talking about business as a whole and for your consideration for PS6 the process would likely be N3P so 18AP will be a good comparison.

 

[Maximus]

I was talking about business as a whole and for your consideration for PS6 the process would likely be N3P so 18AP will be a good comparison.

https://www.reuters.com/technology/how-intel-lost-sony-playstation-business-2024-09-16/

PlayStation 6 SoC is Design Complete and In Pre-Si Validation Phase; GPU is an Early Fork of Gfx13 - Rumor

The PlayStation 6 SoC is design complete and in pre-silicon validation phase, with A0 tapeout planned for this year.

wccftech.com

Multiple rumors suggest that Intel has lost the bid for the PS6 SoC, with the console expected to launch in 2027. If the opportunity to secure PS6 business has passed, the next available window would be for PS7, projected for a 2034 release—potentially an A14 vs. 14A scenario.

If the PS6 SoC bidding took place in 2022, as reported by Reuters, then following the same cadence, the PS7 SoC bidding would likely occur in 2029.

 

[siliconbruh999]

https://www.reuters.com/technology/how-intel-lost-sony-playstation-business-2024-09-16/

PlayStation 6 SoC is Design Complete and In Pre-Si Validation Phase; GPU is an Early Fork of Gfx13 - Rumor

The PlayStation 6 SoC is design complete and in pre-silicon validation phase, with A0 tapeout planned for this year.

wccftech.com

Multiple rumors suggest that Intel has lost the bid for the PS6 SoC, with the console expected to launch in 2027. If the opportunity to secure PS6 business has passed, the next available window would be for PS7, projected for a 2034 release—potentially an A14 vs. 14A scenario.

If the PS6 SoC bidding took place in 2022, as reported by Reuters, then following the same cadence, the PS7 SoC bidding would likely occur in 2029.

Yeah likely i am 100% sure it is due their margin or Sony not wanted to port their stuff to other archirectures.

 

[Markwrob]

Intel Foundry will have to hope they can win some deals sized like these console deals, since no customers are likely to risk any bigger deal with IFS for the first round. IMO.* That said, from the console makers' point of view, they must be very careful because each such deal could make or break their product line over the next 6 years or so. That puts the thumb on the scale towards awarding the deal to whom they already know, while using the IFS bid to lower the final pricing of course.

* : every '5%' deal IFS can bag is good since it gives IFS some fab volume to drive improvement and signals to other potential customers that IFS may have something for them too.

IFS has to bid competitively for each of these '5%' deals imo because that's the largest deals they will see from customers for now. IFS must be especially patient and committed to earning trust in this phase of the business.

 

[TICin852]

1. 4. Established partnerships—Microsoft and Sony have collaborated with AMD for over a decade. Unlike cutting-edge semiconductor advancements, console and handheld development relies on long-term customization and stability. Given Intel's recent trend of discontinuing non-core or less profitable projects, it's unlikely that Microsoft and Sony would feel confident working with Intel for future console and handheld generations.

Here's the thing: Intel's software support teams are massive and Intel has contributed significantly to many open standards. I do think that Intel would be willing to provide extended support and that they would point directly to their ecosystem support - Microsoft and Sony shouldn't really be as worried about this. Additionally, the products discontinued historically were hardware and not software - I wonder who would be working with the console company - Intel products, or are they working with the foundry directly?

For handhelds, though, Intel has some products it could offer - Lunar Lake is the first I can think of, and they certainly can offer some extreme low power x86 CPU to Microsoft if they want. A custom PTL APU for low power could definitely work, depending on how PTL works out - especially because 70% of silicon is in-house.

 

[siliconbruh999]

Here's the thing: Intel's software support teams are massive and Intel has contributed significantly to many open standards. I do think that Intel would be willing to provide extended support and that they would point directly to their ecosystem support - Microsoft and Sony shouldn't really be as worried about this. Additionally, the products discontinued historically were hardware and not software - I wonder who would be working with the console company - Intel products, or are they working with the foundry directly?

I absolutely agree with you here part of AMD advantage in DC CPU is due to Intel.
( Autocorrect Corrected AMD to and thanks autocorrect for being so useful.)
Intel has spent years optimizing their CPU SW Stack not to mention much of the part is open source.

For handhelds, though, Intel has some products it could offer - Lunar Lake is the first I can think of, and they certainly can offer some extreme low power x86 CPU to Microsoft if they want. A custom PTL APU for low power could definitely work, depending on how PTL works out - especially because 70% of silicon is in-house.

They can increase the % of inhouse silicon.

 

[semiman]

The other problem Intel faces in console bidding is that most of GPU designs are already designed for TSMC. AMD have been designing their own CPUs and GPUs in TSMC, while only GPUs Intel foundry manufactured were their in-house GPUs. That's part of reason why Samsung 8nm won Nintendo console bidding. New Nintendo console uses NVIDIA Ampere microarchitecture, which was designed for SF8 thanks to RTX 3000 series and NVIDIA Odin SoCs in early 2020s. IP ecosystem matters a lot.
So Intel has to either win business using their in-house designed console SoC (18A + ARC + Intel CPU, in this case Intel "Product" team has to fight against AMD products) or get some design win first(AMD GPU working better in Intel foundry?) in leading edge nodes and wait for process to mature...

 

[TICin852]

MJ's comment about not going after products unless they get a 50% gross margin seems like this would preclude Intel from doing consoles. Not the best idea, at least in my view.