inference

Design software company Canva has lowered its annual revenue growth forecast by one-third to 20% due to AI feature delivery costs far exceeding expectations. CEO Melanie Perkins stated that user demand for AI features "significantly exceeded expectations," but the company has decided to postpone the large-scale launch of related products until it optimizes its architecture, reduces unit costs, and improves its business model. Since the release of Canva AI 2.0 in April, the cost per task has decreased by nearly 90%, but the volume of designs generated by AI users is three times that of previous versions, and economic pressures remain.This case reflects a common dilemma in the software industry: the cost of AI inference is breaking the traditional software "zero marginal cost" profit logic. Pitchbook analysts pointed out that Canva and Figma are the "most obvious signals" of this trend. Figma's free cash flow profit margin in the second quarter has dropped from 27% in the first quarter to 14%, and revenue growth in the third quarter is expected to slow from 48% to 36%. Canva previously sold employee shares at a valuation of $42 billion, and the market had anticipated its IPO in 2026, but analysts now believe the IPO may be delayed until next year.

Research institution SemiAnalysis released an analysis stating that SpaceX is expected to build approximately 10GW of AI data center capacity by the end of 2027. If 50% of this is used for inference services, with annual revenue exceeding $10 billion per GW, the annualized revenue could reach $300 billion. SpaceX CEO Elon Musk stated in the first earnings report that a "conservative estimate" suggests an additional 6-8GW will be added in 2027, with the actual figure possibly exceeding 10GW.SemiAnalysis's inference simulator shows that when running on the GB300 cluster at current startup cloud prices (about $3/GPU hour), leading model companies like OpenAI and Anthropic could generate annual inference revenue exceeding $10 billion per GW, with annual costs around $12 billion. Microsoft, with full access to OpenAI models and without bearing training costs, can also capture revenue of the same scale. The analysis points out that Microsoft has signed contracts for 10GW of data centers (total value exceeding $300 billion) since 2026, with a 90-day cancellation clause, significantly reducing signing risks.Regarding SpaceX's construction progress, SemiAnalysis believes Musk will significantly shorten the construction cycle by using onsite gas power generation, bypassing large power transformers, parallel construction, and shortening the debugging process. The Southaven plant in Tennessee expanded from 27 turbines (approximately 495MW) in February to 69 turbines (1.7GW) in July, and the "MiniHard" project can be completed in about 5 months with 450-500MW. However, the 10GW target still faces multiple challenges such as land approvals, gas supply, and equipment delivery. This analysis is based on model simulations, and actual implementation still carries uncertainties.

According to a report by Reuters, AMD announced the acquisition of AI inference chip startup Taalas, with the transaction amount yet to be disclosed.The purpose of this acquisition is to enhance AMD's technological capabilities in the field of artificial intelligence inference chips, and it plans to integrate Taalas's technology into its Instinct GPU accelerator roadmap, further advancing the development of system-level solutions.

Jensen Huang announced on the X platform the launch of Alpamayo 2 Super, a cutting-edge open-source inference model developed by NVIDIA for the autonomous driving field. Jensen Huang stated that Alpamayo 2 Super can not only "see" the environment but also understand and reason about the complex world, thinking before acting, and can serve as the core model for robot taxis, autonomous trucks, shuttles, delivery vehicles, agricultural machinery, and future large-scale mobile robot systems.The model will be based on OpenMDW-1.1 and made available for commercial use, allowing developers and enterprises to inspect, fine-tune, and deploy the model. Jensen Huang mentioned that open models help promote the safety and reliability of autonomous driving systems. He stated that the next wave of artificial intelligence will belong to robotics, and autonomous vehicles will be an important starting point for this trend.

According to official news, Marvell Technology announced the launch of a new generation of memory solution portfolio for AI infrastructure, covering server-level AI storage, rack-level CXL memory expansion and pooling, as well as multi-rack optical interconnect shared memory, aimed at addressing the growing memory capacity and bandwidth bottlenecks in Agentic AI inference processes.Marvell stated that as AI model sizes increase, context windows extend, and KV Cache demand grows, traditional tightly coupled architectures of computing and memory are limiting AI inference efficiency. Through memory disaggregation, memory resources can be made more independent of computing resource expansion, improving GPU utilization and reducing data movement latency. The products released include:Bravera SC6 PCIe 6.0 SSD controller: Designed for AI inference storage scenarios, it helps cloud service providers migrate more KV Cache to high-performance SSDs, enhancing infrastructure efficiency. This product features an architecture compatible with multi-vendor NAND and is expected to begin sampling in the fourth quarter of 2026.Marvell Structera X memory expansion solution: Based on CXL technology, it supports rack-level memory expansion and resource pooling, helping data centers share and allocate memory resources more flexibly, reducing AI infrastructure costs.Marvell Photonic Fabric optical interconnect memory solution: Constructs a shared memory architecture across multiple racks using optical interconnect technology, supporting up to 32TB warm KV Cache offloading and helping AI inference clusters enhance throughput capacity.Marvell stated that the Photonic Fabric solution can achieve a 2 to 3 times increase in Token throughput under existing data center space and power consumption constraints, supporting larger scale models and longer context AI applications.Marvell executive Will Chu stated that AI infrastructure is transitioning from a single server architecture to a system where computing, memory, and connectivity operate in synergy, and in the future, memory needs to expand more independently to enhance resource utilization and Token efficiency.As the demand for AI Agents and large model inference continues to grow, memory capacity, bandwidth, and data transfer efficiency are becoming new focal points of competition in AI infrastructure, following computing power.

According to Jinshi reports, the CEO of Amazon Web Services (AWS) stated that the artificial intelligence business is transitioning from the training phase to the inference phase.

According to reports, three informed sources revealed that the Chinese AI startup DeepSeek is developing its own AI chip. This chip is primarily designed for the inference stage of AI computing, rather than for training new models. This move aims to reduce the company's reliance on chips from Nvidia and Huawei.The report noted that the chip development plan began about a year ago and is still in the early stages. DeepSeek is discussing with external partners such as chip design firms, foundries, and storage companies, and has increased its recruitment of chip design engineers through private channels in recent months. Additionally, following the news, Nvidia's pre-market stock price fell slightly by about 1.6%.Analysts believe that if the development is successful, it will mark a significant strategic shift for DeepSeek, aligning with the global trend of leading AI developers seeking hardware autonomy, and may also pose a challenge to Huawei's position in the domestic AI chip market. However, the project still faces challenges such as long design cycles, high funding requirements, and manufacturing barriers like U.S. export restrictions on advanced processes and high-bandwidth memory (HBM). This chip development initiative also aligns with recent reports about DeepSeek's plan to raise $7 billion in its first round of financing.

The founder of the AI agency platform Bankr, deployer, tweeted that he has raised about $100,000 to provide free LLM inference services for Bankr ecosystem projects. Projects currently being built in the Bankr ecosystem will soon be able to access free inference credits. However, before officially starting the distribution, the team will first launch a new feature that "the entire industry will love," with specific details to be announced later.

According to official disclosure from 0G Labs, 0G Private Computer has officially launched, providing privacy inference services based on TEE (Trusted Execution Environment) for AI agents and developers. Users can access the service through pc.0g.ai.The official statement indicates that the product utilizes a hardware isolation environment composed of Intel TDX CPU and NVIDIA H100/H200 GPU, with all inference requests completed within the TEE. Input content is transmitted in an encrypted manner, and output results will undergo signature verification, ensuring that service providers cannot directly view user data.Currently, the platform supports models such as DeepSeek DeepSeek Chat V3, Alibaba Qwen3.6 Plus, Zhipu AI GLM-5-FP8, Qwen3-VL-30B, OpenAI Whisper-large-v3, and z-image, and provides an interface compatible with the OpenAI API, allowing developers to complete migration by modifying the base URL.

AI Smart Routing and Infrastructure Network DGrid AI officially announces the launch of its AI inference API integrated with the x402 payment protocol. This API seamlessly merges payment logic with AI model invocation for the first time, allowing developers and AI Agents to complete authorization, inference, and payment within a single API request lifecycle without managing cumbersome API Keys or requiring centralized account pre-funding.It is reported that this API currently supports BNB Chain (BSC) as the underlying settlement network. With the micropayment features of the x402 protocol, the system can achieve extremely low-friction on-chain real-time settlement while ensuring that invocation costs remain absolutely controllable. Additionally, this API supports streaming responses and real-time usage feedback while being compatible with mainstream AI invocation methods, making it suitable for various application scenarios, including model selection for AI Agents, inference billing, intelligent agents, and multi-task execution.DGrid states that this initiative aims to completely break the prepaid barriers of traditional large model invocation and provide a programmable underlying payment infrastructure for building a fully automated AI Agent economy (machine-to-machine transactions).

OpenClaw released version 2026.4.7, adding OpenClaw inference capabilities, music and video editing, conversation branching/recovery, Webhook-based TaskFlows, and support for Arcee, Gemma 4, and Ollama visual models, while also launching the memory-wiki persistent knowledge system.

Tether CEO Paolo Ardoino revealed that the Tether AI team has released a new version of QVAC Fabric, which integrates the cross-platform BitNet LoRA framework, enabling the training and inference of billion-parameter large models on consumer-grade GPUs and smartphones.The new version of QVAC Fabric LLM has achieved cross-platform operation of BitNet LoRA fine-tuning and inference on AMD, Intel, Apple Metal, and mobile GPUs for the first time. On flagship devices, GPU inference speed is improved by 2 to 11 times compared to CPU, and memory usage is reduced by up to 90% compared to full precision models. The Tether team has completed fine-tuning of models with up to 3.8 billion parameters on flagship smartphones such as Pixel 9, S25, and iPhone 16, and has achieved fine-tuning of models with up to 13 billion parameters on the iPhone 16. The related code has been open-sourced to GitHub.

AI quantitative trading company Inference Research announced the completion of a $20 million seed round financing, led by Avenir Group. The funds will be used to expand the company's self-developed infrastructure to integrate traditional finance and digital assets, while accelerating talent recruitment. It is reported that the company is forming a team of quantitative analysts, engineers, and researchers.

Decentralized AI inference routing network DGrid AI has announced that it has received seed round investment support from institutions such as Waterdrip Capital, IoTeX, Paramita VC, Cacher VC, 4EVER Research, Zenith Capital, and Abraca Research.DGrid's core relies on the "AI RPC + LLM inference + distributed nodes" architecture, aiming to break the monopoly of centralized AI platforms and provide open, trustworthy, and low-cost AI capabilities for the Web3 ecosystem. The funding will primarily be used for building the node network and expanding ecological partnerships.In addition, DGrid officially announced that to build a truly community-driven decentralized network, its node pre-sale plan will soon be officially launched. This plan will allow community members to participate in the construction of the decentralized AI network in advance by operating nodes and share in the early growth dividends of the network. Specific rules and timing will be announced through official channels.DGrid AI is a decentralized AI inference routing network that provides developers and applications with low-cost, verifiable, and trustless AI capabilities.

ChainCatcher news, open-source AI provider Inference has completed a $11.8 million seed round financing, led by Multicoin Capital and a16z CSX, with participation from Topology Ventures, Founders, Inc., and angel investors.This funding will expand its R&D efforts into new areas of model and infrastructure performance, while enhancing its ability to serve more companies. Inference.net enables enterprises to train and deploy custom AI models, having trained and deployed custom language models for AI-native companies worldwide.

ChainCatcher news, according to official sources, the decentralized AI inference network DecentralGPT has announced the completion of a $7 million strategic investment, led by AGICrypto.DecentralGPT stated that all investor tokens are locked for 1 year.ChainCatcher previously reported that the Web3 AI project DecentralGPT announced the completion of a $3 million equivalent asset angel round and Series A strategic financing recently. This round of financing was led by AGICrypto Capital, with participation from Cherry Ventures, BTR Capital, DePINX, Aethir, and others.As a fully decentralized large language model (LLM) infrastructure, DecentralGPT has revolutionarily achieved decentralized deployment of multiple top open-source large models, including DeepSeek R1 and Llama 4.0, through an innovative distributed AI inference network. The project's core architecture breaks the traditional AI monopoly pattern, truly implementing the next-generation AGI development philosophy of "secure and verifiable, data privacy, democratic governance, and fully open-source code," creating a trustless risk-free intelligent interaction ecosystem for global developers and users.