The recent whitepaper from Jon Peddie on Multi GPU issues and opportunities is an interesting read. It’s a pretty even-handed analysis, as you would expect from someone of Jon’s experience. However I do wonder about the way in which it is being reported.
Related posts
Add-in BoardGPUGraphic Cards
Experimental GPUs from DIYers
Not multibillion transistors or tens of thousands of shaders.
3Dcharacter creationDazGamingGenerative AIYellow
An easy, colorful way to generate 3D characters
YellowSculpt uses AI and natural language processing to turn text into a pre-rigged 3D character model.
Recent products
AI Processors in Wearables Report
(Buyer will receive $300 coupon, good for 90 days, toward any other JPR AIP report)
Jon Peddie Research’s AI Processors in Wearables report is a supply-side report analyzing the emerging class of SoCs designed for AI inference at the edge.
The Wearables in AI report defines the SoCs used in wearables and how they are used.
The report also contains information about the companies, their technologies and products, and the approaches each company is taking.
SoCs used in wearables are not a replacement for GPUs or NPUs — they are typically highly efficient, low-power chips that sell in medium- to high-volume devices. This JPR report covers the architecture, competitive landscape, and commercial trajectory of wearable SoCs from Whoop 4.0 / 5.0, Fitbit Air, Oura Ring 4, Amazfit Helio, Polar Loop, and Garmin Cirqa.
FPGAs in AI Report
(Buyer will receive $300 coupon, good for 90 days, toward any other JPR AIP report)
Jon Peddie Research’s FPGAs in AI report is a supply-side report; it covers the companies building field-programmable chips for AI inference at the edge and in the cloud, and IoTs.
The FPGAs in AI report defines FPGA processors and how they are used.
The report also contains information about the companies in this segment and their technology and products, as well as the approaches each company is taking.
FPGA processors are not a replacement for GPUs or NPUs — they are typically used as test chips and low-volume devices where the time and cost of designing and building an ASIC don’t make sense. The JPR report covers the architecture, competitive landscape, and commercial trajectory of FPGAs for AI chips from Altera/Intel, AMD/Xilinx, Flex Logix, Lattice Semiconductor, and QuickLogic.
Neuromorphic AI Processors Report
(Buyer will receive $300 coupon, good for 90 days, toward any other JPR AIP report)
Jon Peddie Research’s Neuromorphic AI Processors report is a supply-side report; it covers the companies building neuromorphic compute processors.
The Neuromorphic AI Processors report defines neuromorphic AI processors and how they are used. The report also contains information about the companies in this segment and their technology and products, as well as the approaches each company is taking.
The broad neuromorphic landscape divides into three main categories. Digital SNN processors implement spike-based computation using CMOS logic. Analog and mixed-signal designs process signals in continuous domains to improve efficiency. A third group adopts neuromorphic principles without relying on spike timing, instead using digital, analog, or photonic methods to accelerate AI workloads.