ByteDance is reportedly negotiating to buy MediaTek AI chips for its data centers, though no deal has been confirmed. Bloomberg News reported on Oct 15, citing people familiar, that an agreement is close; the buyer's 2026 AI capex guidance is 160 to 200 billion yuan.
Perspective Coverage
4 publishers
- Builder
- Builder 15%
- Operator
- Operator 30%
- Investor
- Investor 55%
Reality
- Evidence40
- Adoption
- Not scored
- Hype gap+20
- Incentives
- Not scored
- Confidence55
buildOne report1 publisher d-Matrix models 100 TB/s per card for Raptor, a stacked-DRAM chip due in late 2027
d-Matrix's ISCA paper models up to 100 TB/s per card for Raptor, a stacked-DRAM inference accelerator. The figures come from modeling and the first cards are expected in Nvidia MGX racks in late 2027, so they have no bearing on inference hardware bought in the next year.
Reality
- Evidence38
- Adoption4
- Hype gap+35
- Incentives70
- Confidence45
buildConfirmed3 publishers NVIDIA is licensing the Fusion chiplet, its C2C CPU attach and NVHBM to MediaTek and buying $3.5 billion of its convertible bonds, so the quickest route to a custom accelerator now ends inside an NVIDIA rack.
Perspective Coverage
3 publishers
- Builder
- Builder 37%
- Operator
- Operator 22%
- Investor
- Investor 41%
Reality
- Evidence55
- Adoption20
- Hype gap+30
- Incentives80
- Confidence60
Nvidia's money buys bonds that can convert into MediaTek equity. The return that matters is NVLink Fusion, which routes every custom accelerator MediaTek designs for a hyperscaler back into an Nvidia rack. The bond terms were not disclosed.
Perspective Coverage
6 publishers
- Builder
- Builder 27%
- Operator
- Operator 24%
- Investor
- Investor 49%
Reality
- Evidence72
- Adoption25
- Hype gap+30
- Incentives70
- Confidence66
The saving comes off cache reads, so it lands unevenly across teams, and the model's most sensitive capability now reaches users through a credential issued by an access program, not a purchase order a budget holder can sign off alone. Both changes move the decision away from the budget holder.
Perspective Coverage
3 publishers
- Builder
- Builder 37%
- Operator
- Operator 25%
- Investor
- Investor 38%
Reality
- Evidence55
- Adoption30
- Hype gap+35
- Incentives65
- Confidence60
The Information reports two configurations, one with two chips and one with four, aimed at developers, businesses and governments, with a possible 2029 launch. Apple already builds servers for Private Cloud Compute and has turned partners away from them.
Perspective Coverage
7 publishers
- Builder
- Builder 32%
- Operator
- Operator 32%
- Investor
- Investor 36%
Reality
- Evidence35
- Adoption15
- Hype gap+20
- Incentives55
- Confidence40
Qualcomm puts tokens per watt at the centre of its AI infrastructure pitch
Qualcomm's Tony Pialis told the AI Infra Summit in Santa Clara that tokens per watt is the new key metric in the AI war. The vendors in the room disagreed about which memory architecture gets them there.
Reality
- Evidence46
- Adoption34
- Hype gap+33
- Incentives82
- Confidence52
buildOne report1 publisher Apple's reported M8 Ultra server would pick up NVLink where UltraFusion stops
The Information reports Apple is weighing Nvidia's NVLink Fusion for an inference server built on two or four M8 Ultra chips in 2029. The engineering question is which part of that package would join the NVLink domain.
Perspective Coverage
3 publishers
- Builder
- Builder 37%
- Operator
- Operator 23%
- Investor
- Investor 40%
Reality
- Evidence30
- Adoption18
- Hype gap+42
- Incentives55
- Confidence50
Apple weighs building its M8 Ultra servers around Nvidia's NVLink Fusion interconnect
Apple has not confirmed the plan and the M8 Ultra is years from release. Seven other companies have already licensed the same rack interconnect from Nvidia, and their dates are the ones a buyer can use.
Reality
- Evidence45
- Adoption50
- Hype gap+20
- Incentives70
- Confidence55
buildOne report1 publisher Lambda ran 19 Blackwell nodes on the power budget provisioned for 16
The one measured number behind NVIDIA's tokens-per-megawatt pitch comes from Lambda's Blackwell cluster, where power reclaimed from static provisioning ran three extra nodes. Amazon's Annapurna Labs and d-Matrix got a line each.
Reality
- Evidence33
- Adoption45
- Hype gap+44
- Incentives90
- Confidence60
buildConfirmed2 publishers NVHBM offers NVLink Fusion partners 30 percent more bandwidth and 15 percent less power, in exchange for NVIDIA owning the interface between their compute and their memory.
Perspective Coverage
3 publishers
- Builder
- Builder 45%
- Operator
- Operator 29%
- Investor
- Investor 26%
Reality
- Evidence34
- Adoption14
- Hype gap+38
- Incentives88
- Confidence66
buildOne report1 publisher Memory is 30% of hyperscaler capex, and the cheapest fix is the DDR4 you already own
Contract prices roughly tripled in two quarters. Marvell is selling recycled DDR4 behind CXL, but its reference customer, Meta, did the job on an ASIC of its own design.
Reality
- Evidence63
- Adoption38
- Hype gap+33
- Incentives74
- Confidence58