Scott Wasson from The Tech Report is on the ball - he benchmarked an AMD Phenom processor affected by the translation lookaside buffer (TLB) and L3 cache bug which is said to be rare but could cause system crashed under virtualization mode.

AMD promised a fix by way of a motherboard BIOS update but the sad news for AMD fans is that patch causes a penalty hit which Scott was kind enough to test.

The folks over at BCC Hardware managed to get their hands on an upcoming Nvidia nForce 780i SLI reference motherboard.

As you may know, this is the upcoming chipset to replace the aging 680i SLI chipset which lacks proper support for new 45nm processors from Intel.

Thermaltake relentless in their endeavors for the ultimate product range today announced a couple dedicated graphics card power supplies - a 450-watt and 650-watt model.

With some high-end graphics cards these days chewing up as much as 300 watts of power, not all main system power supply units will be able to cope with the stress. If you're using a 400 watt power supply for instance and do not particularly want to upgrade it, this is where the VGA PSU from Thermaltake could come in handy.

Galaxy recently informed us about the release of their brand new 8800GT graphics card with native HDMI port for HD output up to 1080p resolutions.

As the Hong Kong based company likes to do, it comes with a customized cooler along with higher 3-phase power system compared to the usual 2-phase, SLI (but not 3-way SLI since it only has one SLI bridge connector) and all the usual GT trimmings.

Expreview have unravelled the workings of a unique locking mechanism spotted on one of ASUS's upcoming ROG series motherboards for the AMD AM2 platform.

Dubbed "G-Guardian", this feature is aimed to be a protective system used to stop unwanted intruders at a LAN or other publicly accessible location from stealing your expensive memory modules or keyboard/mouse.

To come with Intel's first batch of 45nm desktop CPUs early into the new year is the replacement chip for the very popular Q6600, Intel's entry-level quad core part.

Known as the Core 2 Quad Q9300, this processor will come clocked in at 2.5GHz (100MHz faster than the Q6600) but will carry only half the L2 cache (6MB).

On the upside, being a 45nm Penryn chip it's going to run a lot cooler and with a lot more overclocking headroom to boot; potentially making it the new chip of choice for many enthusiasts.

The folks from Hard Spell in China who call themselves a professional IT hardware product website managed to scoop up one of GIGABYTE's upcoming Intel X48 chipset based DQ6 motherboards for some early testing over in the mainland.

On an early production board from GIGABYTE which we believe is still being fine tuned according to our sources at the Taiwan motherboard company, they compared Intel QX9650 and QX9770 processors.

The folks over at Sapphire sent in a press release today announcing their passively cooled ULTIMATE Edition based on the ATI Radeon HD 3850 GPU.

They claim "to deliver the fastest performing passive cooled graphics card in the World - ever!". That might be the case for fastest passively cooled Radeon graphics card but not for graphics cards on the whole as surely the recently released passively cooled Sparkle 8800GT will beat it in most tests.

Following news of a passively cooled 8800GT to show up very soon from Sparkle, ECS are following suit with their new 8800GT Dual Turbo Version Graphics Card which sports Arctic Cooling's beefy S1 cooler.

This massive heatsink with large heatpipes does have a couple of fans sitting on it for really bringing down those temps to gain some more overclocking headroom, but the cooler is easily capable of managing the G92's heat levels passively as well.

A fresh Multi-GPU Performance Hotfix has become available for Windows Vista users which deals with a DMA latency issue of sorts between the two cards. Microsoft describe it in full detail below :-

In some scenarios, games and benchmarks perform more poorly than expected on a Windows Vista-based computer. This issue occurs when the games and the benchmarks use multiple display adapters in a Linked Display Adapter (LDA) configuration.

A graphics processing unit (GPU) or part of a GPU link may incur a small scheduling latency on a new Direct Memory Access (DMA) packet. This is true if the DMA packet is submitted to the GPU that has previously run out of work and if another GPU in the linked configuration is very busy. In some scenarios, significant starvation of one or more GPUs in the linked configuration reduces the expected performance of an application.