Advanced Micro Devices has unveiled unloaded a bevy of product roadmap details for its upcoming Fusion processors—though; much of the information had already been leaked last month.
AMD's Fusion chips are the culmination of the chip maker's blending of x86 central processor technology with the GPU instruction set it acquired when it bought ATI Technologies in 2006. By putting both key computing functions on a single processor die, the company thinks it has a more compelling processor package for makers of mobile computers and light-footprint desktops than rivals Intel and Nvidia.
AMD calls these chips "accelerated processing units" or APUs. The company's A-Series APUs, formerly codenamed Llano, are currently shipping to computer makers and are expected to appear in more than 150 desktops and notebooks set to hit retail shelves throughout the second quarter of this year, AMD said.
The company isn't shy about talking up the advantages of its A-series processors, which combine up to four x86 CPU cores with up to 400 Radeon GPU cores with DirectX11 support, and dedicated HD video processing on a single chip of silicon.
Samsung may be developing phones that run so fast, they'll leave today's most cutting-edge dual-core phones in the dust. The company apparently has dual-core 2GHz smartphones in the works, which would theoretically be capable of 4GHz speeds, says Korea's Maeli Business Newspaper.
"We are planning to release a 2GHz dual-core CPU-equipped smartphone by next year," the paper reports a "high-ranking official" from Samsung said. "This product will have the data processing capacities of a regular PC."
The report comes as dual-core 1GHz phones such as the Motorola Atrix and LG Optimus 2X, which was the world's first dual-core phone, are just beginning to become available. A phone running with two 2GHz chips would technically be capable of running at speeds up to 4GHz, though real-world tests almost never match specs.
We caught up with Pat Moorhead of AMD and had a chat with him about the new Turion X2 chipsets and the new Hybrid Graphics technology, which allows dynamic switching between integrated and discrete graphics cards, with no restart required. The technology promises better battery life and enhanced rendering performance (although, unfortunately, not necessarily at the same time).
Stay tuned for more news from Computex 2008.
We take a close look at AMD‘s recently announced “Puma” notebook platform. Puma is a tightly-knit system for notebook suppliers comprising of the chipset, CPU, GPU and wireless chipset. Similar in some ways to their Spider desktop platform, Puma takes it to the next level by allowing for a hybrid integrated and discreet graphics solution. While Intel has something similar, AMD is apparently the first to offer a dynamically switchable system, allowing your notebook to automatically switch to integrated graphics when it detects you’ve gone on battery, or to allow you to switch manually. I asked if they plan on making it automatic, based on load, which they seemed to think was a pretty great idea and relatively easy to achieve as well. What’s more, despite the fact that the integrated and discreet GPUs are chips of differing abilities and specs, they’re still able to operate in Crossfire/hybrid mode, providing an extra boost in power and achieving some very impressive framerates.
AMD shows us their reference bench system and two other identically-configured systems, the only difference between them an Intel versus AMD integrated graphics chipset. (And CPU, naturally.) With a price difference of around $25 to AMD’s favor, the AMD integrated chip actually performs considerably better in their demo, grain of salt included. The reference system shows off their hybrid processing and lets us see what framerates we might be able to hit on an entry- to mid-level hybrid Puma system.
Puma also brings about something that’s been announced with Intel as well—the ability to deactivate CPU cores and speeds, instead of just stepping down speed. This, coupled with an intelligent HD decoder will bring about several hours of additional battery life.
IBM today announced that they have made the first shipment of “Broadway” processors for Nintendo’s Wii processor, according to a report on Gamasutra. The shipments come out of IBM’s East Fishkill fabrication plant. The CPU utilizes IBM’s Silicon on Insulator technology that provides a reduction in power consumption of over 20 percent over the “Gekko” processor in the Gamecube. The news from IBM suggests that Nintendo is well on the way towards making their production goals for the Wii.
Read More | Gamasutra
Intel has revealed its new dual-core processors for high-end servers using more than four chips, the Intel Xeon 7100 Series. The units, although based on the outdated NetBurst microarchitecture, still offer relatively strong performance as well as power efficiency. The chips, built on the 65nm process, includes all of the features expected on Intel processors such as hyperthreading, virtualization, and Intel cache safe technology. The clock speed vary on different models, ranging from 2.60Ghz to 3.40Ghz, though all feature 2MB of L2 cache. The Xeon 7100’s are directed at enterprise customers. Admittedly, Intel has done an excellent job lately with their new technologies and processors. Looks like it’s time for AMD to play catch up.
Read More | X-bit Labs