38 posts
The Hot Take: Interesting, is the bubble about to pop?
The latest quarterly figures show a recovery in PC graphics shipments, although the improvement is uneven across the market. Published on September 3, the report puts second-quarter 2026 shipments at 75.
Read the full articleThe Hot Take: Broadcom making bank from Ai ASICs? Is Nvidia's domination of the Ai market about to slip?
Broadcom's third quarter, ended August 2, 2026, is the point at which custom AI silicon stopped being an adjacent product line and became the shape of the company. Revenue reached US$29.6 billion, up 86% from a year earlier and 33% from the previous quarter. AI semiconductor revenue of US$16.7 billion ā up 221% year on year, and 54% sequentially ā is now larger than everything Broadcom billed, across every product and both segments, in the same quarter of 2025.
Read the full articleThe Hot Take: Please let GPU prices drop soon....
OpenAI seems to have used its current spate of models that run on NVIDIA GPUs to design the Jalapeno chip, and then paired it with its own Gluon kernel programming language, thereby daring to threaten NVIDIA's legendary CUDA moat. How the tables have turned! The architecture of OpenAI's Jalapeno chip SemiAnalysis has just provided a detailed breakdown of OpenAI's Jalapeno ASIC, describing a somewhat novel architecture that delivers substantial economies of scale. As such, the Jalapeno pairs a single reticle-sized compute die - built on TSMCāsā¦
Read the full articleThe Hot Take: We're entering the twilight zone.
Mini human brains are being grown in labs all over the world. Soon, they could outthink neural networks.
Read the full articleThe Hot Take: That sounds alot like an ASIC...
In AMDās latest bid to upset Nvidia's dominance in AI hardware, the House of Zen has acquired AI chip company Taalas, which bakes model weights directly into silicon in a process that promises to boost inference performance by an order of magnitude or more. The deal, announced at market close on Thursday, appears to be framed in much the same context as Nvidiaās $20 billion licensing deal with Groq last December: make high-performance āpremiumā inference services prized for AI agents, like code assistants, faster and cheaper to run. AMD didnātā¦
Read the full articleThe Hot Take: Still not buying the ARM SoC's are better than discrete system CPUs. We'll have to see how Intel/AMD ACE instructions far against this.
NVIDIA has started using clusters of its own Vera processors to develop future CPUs, GPUs, and AI accelerators. The company says initial electronic design automation testing shows performance improvements of up to 1.
Read the full articleThe Hot Take: Will have to take this with a bit rock of salt until we see the benches and real world performance. Just a price grab or actual performance?
AMD is apparently no longer satisfied with being perceived as a perennial underdog, and is reportedly choosing to make a bold statement by pricing its Helios rackscale solution at a hefty premium to NVIDIA's Rubin. What's more, AMD has reportedly managed to rope in Microsoft as the product's first confirmed client, demonstrating the utility that Helios offers even at its supposedly elevated price point. Futurum believes AMD will price Helios at $5 million to $5.5 million per rack, versus an expected retail price of $3.5 million to $4 millionā¦
Read the full articleThe Hot Take: Competition is GOOD.
Intel has published a new patent on its XBM memory, which is proposed as a replacement for HBM4, offering much higher bandwidth capabilities. XBM vs HBM: Intel's New Proposed DRAM Solution Extends To 32 GT/s Speeds, While Reducing Costs Through UCIe Links HBM continues to be the standard for AI accelerators, but more recently, we have seen LPDDR memory being used to overcome shortages, prices, and power associated with the standard. Intel's past attempts at DRAM, such as HMC (Hybrid Memory Cube) and MCDRAM, faced various issues and never cameā¦
Read the full articleThe Hot Take: I really wonder if this will deflate the Ai bubble even a little bit.
The x86 Ecosystem Advisory Group led by Intel and AMD recently firmed up the AI Compute Extensions (ACE) specification for optimizing x86 for AI computation tasks around matrix multiplication and the like for machine learning workloads. The cross-vendor ACE extension is ultimately a successor to Intel's Advanced Matrix Extensions (AMX). Posted to the GCC mailing list today by Intel engineers are the initial patches in preparing the compiler support for ACE...
Read the full articleThe Hot Take: Interesting, Risc-V shop from rockstar architect Jim Keller.
Jim Keller isn't bothered by Cerebras's recent IPO and says that he welcomes it, but Tenstorrent will still beat them on everything. Tenstorrent CEO, Jim Keller, Signals Deal With Intel or Qualcomm While Promising To Beat Cerebras "on everything" Tenstorrent recently introduced its latest BlackHole Galaxy server, a system with which it can disrupt the entire AI segment, with performance levels that crush the competition. We covered the announcement last month when the company demoed its Blackhole server undercutting a NVIDIA GB300 with up toā¦
Read the full article