2025-12-30 12:20:25

The chip industry is迎来 a critical turning point. According to the latest industry analysis, we can see how ambitious several tech giants are in the ASIC chip field.

From Broadcom's order volume, Google's demand growth is the most aggressive. It is expected that revenue from Google will double in 2026 compared to 2025, and may double again in 2027—this means the intensity of computing power competition is continuing to heat up.

Meta's ASIC layout is relatively lagging behind, but it is expected to truly ramp up by 2027, at which point revenue may begin to grow significantly. In contrast, ByteDance's ASIC shipments, although starting later, are expected to be quite small in scale by 2026, but the entire production chain will only be fully operational by the end of 2026.

What’s more interesting is the participation of new players. Anthropic's order volume in 2026 may be close to Google's, but there will be a noticeable pullback by 2027. XAI's ASIC chips will truly ramp up in 2027, with huge growth potential. OpenAI's roadmap is clearer—mass production is expected to start by the end of 2026, with large-scale deployment beginning in 2027.

Not to be overlooked, Broadcom has already secured important orders from AWS and Microsoft, and this portion of revenue is expected to see significant growth in 2027. Behind these data reflects that the global AI infrastructure arms race has moved from the conceptual stage into a real investment phase. Whoever can master the maturity and capacity of autonomous chips faster will be able to secure greater influence in the upcoming AI era.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

5 Likes

Reward
5
6
Repost
Share

Comment

0/400

staking_gramps

· 10h ago

Google doubles down, doubles down, this pace... really is a chip arms race --- Wait, Anthropic's order volume is close to Google? When did this guy become so fierce --- Confident in xAI's breakout in 2027, but it depends on whether Musk can really ramp up production --- Broadcom made a huge profit this time, almost dominating the entire scene... the winner of the computing power arms race --- ByteDance won't be able to run until the end of 2026? That's a bit slow, brother, this pace can't keep up with those foreign guys --- Whoever controls the chips in 2027 will dominate the entire AI ecosystem, now I understand --- OpenAI's roadmap is the clearest, this statement is a bit arrogant, but it is indeed much more reliable than Meta --- Everyone is burning money and competing in chips, this is not the same as the crypto circle's虚的 stuff, this is a real infrastructure war

View OriginalReply0

Liquidated_Larry

· 10h ago

Google is really throwing money around like crazy this time, doubling down repeatedly, by 2027 they will have to spend a fortune. I’m optimistic about xAI; starting late actually presents greater opportunities, just worried about production capacity keeping up. ByteDance is still sharpening its tools; it won’t be operational until the end of 2026? That timeline is a bit tight. OpenAI’s roadmap is indeed clear, unlike some projects that shout slogans but make no progress. Broadcom is about to make a killing, all orders coming in. Meta won’t ramp up until next year? Feels a bit slow, how will they ensure competitiveness? This is the real AI arms race—burning money to buy influence, there’s no other way. Independent chip manufacturing has truly become a new necessity; anyone who doesn’t compete will be out. Anthropic’s orders are close to Google’s? No way, that data sounds a bit exaggerated. Who will have the most stable production capacity in 2027? That’s what everyone is betting on now.

View OriginalReply0

AlwaysAnon

· 10h ago

Google's move... doubling in 2026 and doubling again in 2027, this is about fully understanding the entire industry chain. ByteDance started late, but once the production chain is running smoothly, it might take off directly. This is something to keep an eye on. xAI is waiting in the wings, only scaling up in 2027. We can just watch the show then. OpenAI's roadmap is quite clear and aggressive, with a full-scale mass production deployment. That's really about pouring money in. Broadcom is really making a killing, with Google, Meta, Microsoft, AWS all placing orders, and chip development is heating up again. The competition for computing power is heating up—sounds nice, but essentially it's a money-burning race. The deeper your funds, the louder you can speak. Anthropic's recent pullback is a bit interesting; it feels like they are running out of steam. This wave of infrastructure investment is really uncertain; there will definitely be new moves before 2027.

View OriginalReply0

MaticHoleFiller

· 11h ago

Google doubles again, the pace is really fierce... Broadcom is making a killing ByteDance is still sharpening its tools, will it really succeed? OpenAI's large-scale deployment won't happen until 2027, the window of opportunity is a bit tight... Why does xAI have such huge growth potential? What exactly is Musk playing in this chess game? Meta has been lagging behind for so long, it feels like it's falling behind... Anthropic's order volume is close to Google's? That number sounds unbelievable The arms race is heating up, whoever controls the chips has the say, it's just this logic Broadcom has really succeeded in positioning itself this time, benefiting from hardware dividends... By the end of 2026, OpenAI will mass produce, will it be another PPT dream? Let's see Can production capacity really keep up, or is it just another round of false hype...

View OriginalReply0

just_another_wallet

· 11h ago

Google's tactics are too ruthless, doubling down repeatedly. The computing power race in 2027 might just blow everyone's minds. Is xAI's growth potential huge? Elon Musk is up to something again. Will this succeed this time? Broadcom must be making a fortune, landing contracts with Google, AWS, and Microsoft—raking in easy profits. ByteDance's supply chain won't be fully operational until the end of next year? Their efficiency is really lacking. Anthropic is approaching Google in 2026, but then pulls back in 2027, feeling like just following the trend. OpenAI's roadmap is actually the clearest, with mass production by the end of 2026 and expansion in 2027. They’re pacing it well. Ultimately, who controls the capacity in chip manufacturing will determine the outcome; this is how influence is gained. Google's demand growth is so aggressive; ASICs are definitely burning money. It seems Meta is about to fall further behind. Launching in 2027 might be a bit late.

View OriginalReply0

HodlTheDoor

· 11h ago

Google's move is brilliant, doubling in 2026 and doubling again in 2027... I feel like this is the real AI arms race, not just hype about concepts. xAI and Anthropic are also eyeing the prize. OpenAI seems to have the clearest roadmap, and large-scale deployment by 2027 is quite promising. ByteDance's progress is still a bit slow; by the time their supply chain is up and running, the season for yellow flowers might be over... But honestly, the real winner will be whoever can lock down production capacity in the end. Chips are much more reliable than tokens.

View OriginalReply0