27 posts
The Hot Take: If this can even compete with HBM4 then intel has a HUGE win.
Intel and SoftBank, through their joint venture Saimemory, are working on a new memory architecture aimed at addressing bandwidth limitations in AI accelerators.
Read the full articleThe Hot Take: And you get Ai, you get Ai and you get Ai!
Apple's new CEO is very hopeful for how AI can transform the company.
Read the full articleThe Hot Take: HBM competitor would be good bringing Ai accelerator competition for RAM. Intel bringing it should help Intel to catch up in the Ai game for sure.
Japan’s SaiMemory, a SoftBank subsidiary collaborating with Intel, has secured NEDO funding to develop Z-Angle Memory (ZAM), a next-gen DRAM architecture addressing HBM limitations
Read the full articleThe Hot Take: We all desperately need this to not be vapor-ware!
A relatively unknown U. S. -based startup, Bolt Graphics, has announced that it has successfully taped out the first test chip of its Zeus GPU architecture. The company positions Zeus as a disruptive compute accelerator, previously claiming performance up to 150% higher than an NVIDIA GeForce RTX 5090 while consumi ...
Read the full articleThe Hot Take: Getting cozier with your previous manufacturing division in a previous life.
AMD will be leveraging GlobalFoundries for the development of its MRM Co-packaged Optic solution for the next-gen Instinct MI500 AI accelerators. GlobalFoundries & AMD Working Together on Co-Packaged Optics Hardware For Instinct MI500 Accelerators CPO or Co-Packaged Optics (Silicon Photonics) is the next-generation solution that reduces reliance on copper and harnesses light to transfer signals. These CPOs are packaged alongside hardware accelerators such as GPUs and will be a key solution for next-gen AI factories, offering improved interconnect latency and creating high-bandwidth connections between CPU and GPU. Both AMD and NVIDIA will be leveraging these technologies for their next-gen AI […]Read full article at https://wccftech.com/amd-taps-globalfoundries-for-mi500-co-packaged-optics/
Read the full articleThe Hot Take: Seems it was only time with these chips.
AMD's 3D V-Cache CPUs deliver a huge boost versus the Non-X3D part in AI benchmarks, showcasing why they are best suited for RAG pipelines. AMD 3D V-Cache Vs Non 3D V-Cache CPU Benchmarks in AI Showcase a Massive Uplift For RAG Pipelines We know that there are two ways to do AI: the first is LLM, which is currently the most popular model. LLMs are AI models that have been pre-trained on a large set of data and feature various parameter sizes. But LLMs' limitations can be seen when it needs to generate responses on data it wasn't trained on. […]Read full article at https://wccftech.com/amd-3d-v-cache-turns-ryzen-into-surprise-rag-ai-weapon-88-percent-boost-vs-non-x3d/
Read the full articleThe Hot Take: Interesting shift in markets....
If you know the name Allbirds, it's probably for the company's longstanding stated commitment to "sustainable shoes and apparel." Going forward, though, the corporate entity wants to be known for its "long-term vision to become a fully integrated GPU-as-a-Service (GPUaaS) and AI-native cloud solutions provider." In a news release Wednesday morning, Allbirds announced that it has secured a $50 million convertible finance facility to help power this unexpected "pivot ... to AI compute infrastructure." If all goes to plan, the company will soon be known as NewBird AI, by which point it will presumably change the image of a spandex-clad hiker that still sits atop its News Release page. Just weeks ago, Allbirds announced the $39 million sale of the "Allbirds brand and footwear assets" to American Exchange Group, owner of Aerosoles, Ecko Unlimited, and other fashion brands. Today's AI pivot announcement certainly casts that sale in a new light. But Allbirds also announced a new line of colorful Canvas Cruiser shoes just last week, so it's unclear how much long-term planning went into this new AI-related direction.Read full article Comments
Read the full articleThe Hot Take: Apple can't compete so it's going the ASICs route from the looks of it.
Apple generally tries to keep its cards hidden, preferring a grand unveiling, replete with mega-flourishes, to a gradual trickle of information and unvarnished product launches. However, when you have a supply chain as expansive as Apple's, leaks abound nonetheless. And today, we've received a juicy tidbit regarding the tech giant's intentions for its upcoming ASIC, dubbed Baltra. Apple intends to move the production of its upcoming AI ASIC in-house According to a South Korean publication, Samsung Electro-Mechanics (SEMCO) - a company that specializes in core electronic components, multilayer ceramic capacitors, and chip substrates - has provided samples of its glass […]Read full article at https://wccftech.com/apple-shows-its-cards-plans-to-move-the-production-of-its-upcoming-baltra-asic-in-house/
Read the full articleThe Hot Take: Google, Nvidia and not Intel all the suddenly make this amazing new tech at around the same time? Not buying it.
Intel is advancing texture compression techniques with its newly introduced Texture Set Neural Compression (TSNC) technology, a neural network-based approach designed to significantly reduce the size of texture assets used in modern graphics workloads.
Read the full articleThe Hot Take: Interesting.
Nvidia has just demoed its Neural Texture Compression technique again at a GTC talk, where it showed VRAM usage dropping from 6.5 GB to just 970 MB in a scene. NTC uses a neural network to decompress textures instead of standard block-based compression, reducing texture size and VRAM usage while also improving final image quality.
Read the full article