TechPowerUp announced the latest version of GPU-Z, the popular graphics system information, monitoring, and diagnostic utility. Version 0.8.3 adds support for new GPUs, updates support for existing ones, adds new features, and addresses some bugs. To begin with, GPU-Z adds a new feature that tells you if the video BIOS embeds a UEFI module or not, letting you use some of the newer OS features such as Secure Boot and Fast Boot.

GPU-Z 0.8.3 comes with support for new and upcoming GPUs, such as NVIDIA GeForce GTX 980 Ti, GTX 965M, GTX 950M, NVS315, and GT 750 (GK106). On the AMD front, it adds support for AMD "Fiji" GPU, with its new memory technology; and "Mullins" APU (Radeon R2 and R3 series). It also adds support for the integrated graphics cores inside several Intel Core "Broadwell" CPUs. OpenCL detection code is improved, and a missing PerfCap sensor bug is fixed.DOWNLOAD: TechPowerUp GPU-Z 0.8.3 | GPU-Z 0.8.3 ASUS ROG Themed

The complete change-log follows.

Here is the first picture of a couple of NVIDIA GM107 silicons in a tray, ahead of graphics card assembly. The packages appear to be as big as those of the GK106 from the previous generation, however, the die itself is estimated to be smaller, at roughly 156 mm², compared to the 221 mm² die of the GK106, and the 118 mm² of the GK107. The best part? All three chips are built on the same 28 nm silicon fab process. So what makes the GM107 die smaller than that of the GK106 despite having a similar feature-set? Narrower memory bus. The GM107 is said to feature a 128-bit wide GDDR5 memory interface, in comparison to the 192-bit wide interface of the GK106.

Apart from the 128-bit wide GDDR5 memory interface, the GM107 is said to feature a total of 960 CUDA cores, 80 TMUs, and 16 ROPs. The CUDA core count is identical to that of the GK106. The GM107 is built on NVIDIA's next-generation "Maxwell" GPU architecture. It will form the foundation of two SKUs, the GeForce GTX 750 Ti, and the GeForce GTX 750. The former features the full complement of 960 CUDA cores; while the latter is slightly cut down, and features just 768. The TDP of the GTX 750 Ti is approximated to be around 75 Watt. If true, the GTX 750 duo will set new standards on the performance-per-Watt metrics. NVIDIA is expected to launch both, later this month.

With AMD detailing its Radeon R9 and R7 series, especially at some very attractive sub-$299 price-points for the most part, there are jitters being felt at NVIDIA. The company is expected to unveil one or two new sub-$250 GeForce GTX SKUs around mid-October, 2013. The company is also expected to introduce price-cuts across its entire lineup, to make it competitive with AMD's. NVIDIA could tap into its existing GK104 and GK106 silicons to carve out the two new SKUs ranging between $149.99 and $249.99. The idea here would be to topple Radeon R9 270X. Price-cuts could be directed at the likes of GeForce GTX 760 and GTX 770, to make them competitive with the Radeon R9 280X, while in anticipation of the $599 pricing of the R9 290X, NVIDIA could rethink pricing of its $650 GeForce GTX 780, and $1000 GTX TITAN.

PNY launched two GeForce GTX 650 Ti Boost graphics cards under its XLR8 (accelerate) brand. The base model (VCGGTX650TBQXPB) sticks to NVIDIA reference clock speeds of 980 MHz core, 1033 MHz GPU Boost, and 6.00 GHz memory; while an OC variant (VCGGTX650TBQXPB-OC) features factory-overclocked speeds of 1006 MHz core, 1072 MHz GPU Boost, and 6.00 GHz memory.

Both cards appear to be based on designs originally made by Palit. Based on the 28 nm GK106 silicon, the GeForce GTX 650 Ti Boost packs 768 CUDA cores, 56 TMUs, 24 ROPs, and a 192-bit wide GDDR5 memory interface, holding 2 GB of memory. The base model PNY XLR8 GTX 650 Ti Boost is priced at US $169.99, while the OC variant goes for $179.99.

GIGABYTE announced a pair of custom-design graphics cards based on NVIDIA GeForce GTX 650 Ti Boost, the N65TBWF2-2GD, which sticks to NVIDIA-reference clock speeds of 980 MHz core, 1033 MHz GPU Boost, and 6008 MHz memory; and the factory-overclocked N65TBOC-2GD, which ships with 1033 MHz core, 1098 MHz GPU Boost, and 6008 MHz memory.

Both cards are based on a blue GIGABYTE custom-design PCB that uses an UltraDurable VGA construction; and a WindForce 2X parallel-inclined cooling solution, which uses a pair of 90 mm fans to ventilate a heatpipe-fed aluminum fin array. Based on the 28 nm GK106 silicon, the GeForce GTX 650 Ti from NVIDIA features 768 CUDA cores, and a 192-bit wide GDDR5 memory interface, holding 2 GB of memory. The N65TBWF2-2GD is expected to be priced around US $170, while the N65TBOC-2GD could scrape the $190 price point.

MSI kicked off its GeForce GTX 650 Ti Boost lineup with two no less than four models, two of which are based on its 100 mm Propeller Blade cooling solution; and two on its Twin Frozr III cooler, which are branded under its Gaming Series. The series begins with the N650Ti-2GD5 BE, which features NVIDIA-reference clock speeds of 980/1033/6008 MHz (core/GPU Boost/memory), and the N650Ti-2GD5/OC BE, which ups clock speeds to 1006/1072/6008 MHz, while retaining the Propeller Blade cooling solution. The Gaming Series variants consist of the N650Ti TF 2GD5 BE, which sticks to NVIDIA-reference clock speeds; and the N650Ti TF 2GD5/OC BE, with a significant factory-overclock of 1033/1098/6008 MHz.

Based on the 28 nm GK106 silicon, the GeForce GTX 650 Ti Boost features 768 CUDA cores, 64 TMUs, 24 ROPs, and a 192-bit wide GDDR5 memory interface, holding 2 GB of memory. The company did not release pricing.

Galaxy announced its sole GeForce GTX 650 Ti Boost graphics card model, which makes use of custom-design cooler and PCB. The cooler uses an 80 mm fan to ventilate an aluminum fin heatsink. Galaxy's card sticks to NVIDIA-reference clock speeds of 980 MHz core, 1033 MHz GPU Boost, and 6.00 GHz memory. Based on the 28 nm GK106 silicon, the GeForce GTX 650 Ti Boost features 768 CUDA cores, 64 TMUs, 24 ROPs, and a 192-bit wide GDDR5 memory interface, holding 2 GB of memory. The card is expected to be priced at an aggressive US $169.99.

ASUS today announced the GeForce GTX 650 Ti BOOST DirectCU II, which features an overclocked 1085MHz NVIDIA GTX 650 Ti GPU paired with 2GB GDDR5 on a 192-bit interface. The card offers a combination of DirectX 11.1-compatible power and value, expanding the range of ASUS graphics cards that cater to gamers and DIY enthusiasts.

ASUS-exclusive features improve GeForce GTX 650 Ti BOOST DirectCU II performance. They include the DirectCU II cooler, which maintains 20% lower temperatures than a reference GTX 650 Ti BOOST, long-lasting Super Alloy Power components, and user-friendly GPU Tweak graphics card tuning.

NVIDIA launched the GeForce GTX 650 Ti Boost, its "wild card" for the sub-$200 market. Based on the same GK106 silicon as the original GTX 650 Ti and GTX 660, the card acts as an in-between, featuring the CUDA core and TMU count of the GTX 650 Ti, but the memory bus width and ROP count of the GTX 660. It features 768 CUDA cores, 64 TMUs, 24 ROPs, and a 192-bit wide GDDR5 memory interface, holding 2 GB of memory. As its name suggests, the card features GPU Boost. The core is clocked at 980 MHz, which can dynamically overclock itself to 1033 MHz. The memory runs at 6.00 GHz, churning out 144 GB/s memory bandwidth. NVIDIA is looking to disturb AMD's pack in this segment with an aggressive MSRP of $169.99, a 1 GB variant priced at $149.99 is also on the cards. Check out our review here.

Leadtek rolled out the WinFast GeForce GTX 660 Hurricane II graphics card. Based around the same cooler design as its recently-launched WinFast GTX 650 Ti Hurricane II, the card uses a non-reference design PCB, augmented with the company's dual-fan Hurricane II GPU cooler.

The WinFast GTX 660 Hurricane II features reference clock speeds of 980 MHz core, 1033 MHz GPU Boost, and 6.00 GHz memory (GDDR5-effective). It features 2 GB of memory across a 192-bit wide memory interface. Based on the 28 nm GK106 silicon, the GeForce GTX 660 packs 960 CUDA cores. The card draws power from a single 6-pin power connector, display outputs include two DVI, and one each of HDMI and DisplayPort.

PNY Technologies, a global storage, upgrade and graphics solutions manufacturer, recently unveiled the GeForce GTX 650 Ti. Replacing the GTX 550 Ti, this new GPU makes NVIDIA's Kepler architecture available for a price lower than £150. The card allows enthusiasts to immerse themselves in their favourite games with an incredible level of detail, harnessing recently introduced NVIDIA technologies such as Adaptive Vertical Sync, PureVideo HD or NVIDIA 3D Surround. To allow gamers to experience its performance potential, it comes bundled with Ubisoft's Assassin Creed III!

Pauline Monin, Product Manager PC Components at PNY EMEA, said: "An exciting adventure game and an exciting graphics card, this is the perfect combination for the long autumn evenings that await! It's time to take advantage of our limited edition GeForce GTX 650 Ti and Assassin's Creed III bundle. Prepare not to leave your PC for some time!"

NVIDIA added three new high-performance mobile GPUs to its lineup, the GeForce GTX 680MX, GeForce GTX 675MX, and GeForce GTX 670MX. Based on the 28 nm GK104 silicon, the GTX 680MX topples GTX 680M as NVIDIA's (and perhaps, the world's) most powerful GPU for notebooks. While the GTX 680M features 1344 CUDA cores, with 720 MHz core and 3.60 GHz memory, the new GTX 680MX features 1536 CUDA cores, with the same 720 MHz clock speed, but faster 5.00 GHz memory. 2 GB of GDDR5 memory is standard issue for this chip.

Moving on, NVIDIA created a new SKU that takes position between the Fermi architecture-based GTX 675M, and the GTX 680M. Called the GTX 675MX, the new SKU is also based on the GK104 silicon, but configured with 960 CUDA cores, and 2 GB of GDDR5 memory across a 256-bit wide memory interface. The core is clocked at 600 MHz, and memory at 3.60 GHz. The last of the three is the new GTX 670MX, based on the 28 nm GK106 silicon. Like the GTX 675MX, it is configured with 960 CUDA cores (the full compliment of GK106), but with a narrower memory interface, that's 192-bit wide. Its core is clocked at 600 MHz, and memory at 2.80 GHz. 2 GB GDDR5 is the standard memory amount. Expect a wave of performance notebooks and all-in-one desktops to feature these new chips.

ELSA announced the GeForce GTX 650 Ti SAC (silent air cooling) graphics card. While the card bears striking resemblance to the GTX 650 SAC, it uses a different (NVIDIA-reference design) PCB, since the ASIC is different (GK106-220-A1). The card features GPU clock speed of 928 MHz, with 5.40 GHz memory. It draws power from one 6-pin PCIe power connector, display outputs include two dual-link DVI, and a mini-HDMI. The SAC cooler combines an monolithic aluminum heatsink with an easy to detach cooler shroud on which the fan is suspended, which makes the cooler easy to clean. Slated for October 25, the GTX 650 Ti SAC from ELSA is priced around 16,800¥ (US $210).

PNY announced its XLR8 GeForce GTX 650 Ti graphics card. Based on an identical board design to Palit's offering, the card is likely originally manufactured by it. The XLR8 GTX 650 Ti features near-reference clock speeds, of 928 MHz core, and 5.40 GHz (GDDR5-effective) memory. It packs 1 GB of GDDR5 memory across a 128-bit wide memory interface. The card uses an aluminum fin heatsink that's ventilated by an 80 mm fan. Based on the 28 nm GK106 ASIC, the GTX 650 Ti packs 768 CUDA cores, 64 TMUs, and 16 ROPs. PNY's offering includes download coupons to five movies, and a 3-month subscription to E-Sports Entertainment (ESEA) Premium Video Game Network. It is priced at US $179.99.

Gigabyte's GeForce GTX 650 Ti lineup includes a total of three models, the GeForce GTX 650 Ti 1 GB base (GV-N65T-1GI), the GeForce GTX 650 Ti 1 GB OC (GV-N65TOC-1GI), and the premium GeForce GTX 650 Ti 2 GB OC WindForce 2X (GV-N65TOC-2GI). The two 1 GB variants use a compact PCB inspired by NVIDIA reference design, and a single 100 mm fan Triangle-Cool heatsink. The 1 GB base model sticks to NVIDIA-reference clock speeds of 925 MHz core and 5.40 GHz memory, the 1 GB OC version ups the core speed to 1033 MHz core, and the WindForce 2X OC version doubles its memory to 2 GB. Based on the 28 nm GK106 silicon, the GeForce GTX 650 Ti packs 768 CUDA cores, 64 TMUs, 16 ROPs, and a 128-bit wide memory interface. Prices remain under wraps.

Gainward announced its GeForce GTX 650 Ti series with two models based on a custom design. The two include a base-model that sticks to NVIDIA-reference speeds, and the GeForce GTX 650 Ti GS, a factory-overclocked model. The two use a compact full-height PCB with an aluminum fin heatsink and single 80 mm fan. Both models pack 1 GB of GDDR5 memory. Based on the 28 nm GK106 silicon, the GeForce GTX 650 Ti features 768 CUDA cores across four SMX units, 64 TMUs, 16 ROPs, and a 128-bit wide GDDR5 memory interface. Prices start at US $159.

Colorful's kicked off its GeForce GTX 650 Ti lineup with the iGame GTX 650 Ti. Based on a custom board design, Colorful's card makes use of the iGame AresX cooling solution. The cooler consists of a dense aluminum fin stack to which heat from the GPU is fed by two 8 mm-thick nickel-plated copper heat pipes. Ventilation is care of a single 90 mm fan. Another innovative feature is that all electrical contacts on the PCB are silver-plated. The PCB features dual-BIOS, and a factory-overclocked profile can be set at the push of a button. Among the two profiles, one sticks to NVIDIA-reference clock speeds (925/5400 MHz core/memory), while the OC profile could up core clock speed by 7-10%. Based on the 28 nm GK106 silicon, the GeForce GTX 650 Ti packs 768 CUDA cores, 64 TMUs, 16 ROPs, and a 128-bit wide memory interface. Colorful could begin shipping its card in the Greater China region first, before moving on to Europe.

ASUS rolled out its first GeForce GTX 650 Ti graphics cards, which include the GeForce GTX 650 Ti Dual-Fan (GTX650TI-1GD5), the GeForce GTX 650 Ti DirectCU II OC (GTX650TI-DC2O-1GD5) and the premium GeForce GTX 650 Ti DirectCU II TOP (GTX650TI-DC2T-1GD5). The GTX 650 Ti Dual-Fan sticks to near-reference clock speeds of 928 MHz core with 5.40 GHz memory, while the DirectCU II OC ups the core speed to 968 MHz, and DirectCU II TOP to 1033 MHz.

The Dual-Fan card features a slightly trimmed-down version of its DirectCU cooler, which features a monolithic aluminum heatsink with a copper core. The DirectCU II, on the other hand, is ASUS workhorse performance-segment VGA cooler, which uses three copper heat pipes that make direct contact with the GPU, to convey heat to a compound aluminum fin heatsink, which is ventilated by a pair of fans. All three models feature 1 GB of GDDR5 memory. Based on the 28 nm GK106 silicon, the GeForce GTX 650 Ti packs 768 CUDA cores, 64 TMUs, 16 ROPs, and a 128-bit wide memory interface. Prices remain under wraps.

NVIDIA announced its newest (and probably last) addition to its GeForce Kepler desktop GPU family, the GeForce GTX 650 Ti. Based on the 28 nm GK106 silicon, the GPU packs 768 CUDA cores, 64 TMUs, 16 ROPs, and a 128-bit wide GDDR5 memory interface (exactly half that of the GTX 680). The GTX 650 Ti lacks GPU Boost, its core is clocked at 925 MHz, and memory at 1350 MHz (5.40 GHz GDDR5-effective). 1 GB is the standard memory amount, though partners could release premium 2 GB variants. The reference-design board implements a 2+1 phase VRM, which draws power from a single 6-pin PCIe power connector. Display outputs include a pair of dual-link DVI, and HDMI 1.4a. NVIDIA's partners are allowed to release custom-design and factory-overclocked models from the word go. Prices start at US $149.99.

Sources among retailers confirmed what could be the finalized specifications of NVIDIA's upcoming GeForce GTX 650 Ti graphics processor. Some of these specifications were first leaked when Newegg.com accidentally listed Galaxy GTX 650 Ti GC. According to the sources, the GTX 650 Ti, which is based on the 28 nm GK106 silicon, will carry the ASIC label "GK106-220," it will be configured with 768 CUDA cores (and not 576, as earlier believed).

GeForce GTX 650 Ti will have a narrower 128-bit wide GDDR5 memory interface, holding 1 GB of memory. The source also revealed NVIDIA-reference clock speeds to be 925 MHz core, with 1350 MHz (5.40 GHz GDDR5-effective) memory, churning up 86.4 GB/s memory bandwidth. The chip's TDP is rated at 110W, and cards based on it feature one 6-pin PCIe power connector. According to older reports, the GTX 650 Ti is slated for October 9.

Galaxy's upcoming GeForce GTX 650 Ti GC (model: 65IGH8DL7AXX) factory-overclocked graphics card was listed (later deactivated) by American retailer Newegg.com for US $149.99. The sale page revealed quite a few details about the card. The specifications listed beat those which have been doing rounds for the past couple of weeks. According to the specs, the GTX 650 Ti in fact has 768 CUDA cores, and not 576, as previously believed. The GPU core is clocked at 966 MHz, with 5.40 GHz (GDDR5-effective) memory. The card packs 1 GB of memory across a 128-bit wide memory interface. Pictured below is Galaxy's GeForce GTX 650 GC (non-Ti, GK107-based), according to the source, the Galaxy GTX 650 Ti GC (GK106-based) looks very similar.

Post the September 13 launch of GeForce GTX 660 and GTX 650, a wide performance gap between the two is more than evident, and we've known for a few weeks now that NVIDIA isn't stopping at the GTX 660, at harvesting its GK106 silicon. There are more sections of the media with plausible specifications of what is being referred to as the GeForce GTX 650 Ti, an SKU designed to let NVIDIA fill the US $130-$220 price-range, between the GTX 650 and GTX 660.

According to these new set of specifications, the GTX 650 Ti is indeed based on the GK106 silicon, but with an entire GPC disabled, resulting in a CUDA core count of 576, and TMU count of 48. The memory bus width appears to have been reduced to 128-bit, with a standard memory amount of 1 GB. With a memory clock speed of 1350 MHz (5.40 GHz GDDR5-effective), a memory bandwidth of 86 GB/s is achieved. The GPU is clocked at 960 MHz, there is no GPU Boost feature. With the given configuration, the GPU TDP is expected to be no more than 85W.

MSI's flagship factory-overclocked graphics card based on the GeForce GTX 660, the N660GTX HAWK, was pictured and detailed by Expreview, earlier this week. According to the source, the card will be designed for about 15% higher core clock speeds than reference, backed by high-grade voltage regulation circuitry and the company's TwinFrozr IV cooling solution.

To begin with, the GeForce GTX 660 HAWK will ship with clock speeds of 1100 MHz core, 1176 MHz GPU Boost core clock speeds. The memory is left untouched at 6.00 GHz (GDDR5-effective). Under the hood is a 10-phase VRM that uses SSC chokes (low energy loss to noise), and IR DirectFETs. The card appears to use a high-end voltage controller that allows software voltage control, and consolidated voltage measurement points. The card draws power from two 6-pin PCIe power connectors.

Gigabyte's GeForce GTX 660 lineup takes advantage of the company's Ultra Durable VGA+ PCB technology with WindForce 2X cooling solution. The company, much like every other AIC partner, launched two models, a reference-clocked base model with 980/1033/6008 MHz clock speeds, and an OC model clocked at 1033/1098/6008 MHz. The base model is priced at $230, while the OC model goes for $260. Based on the 28 nm GK106 silicon, the GPU packs 960 CUDA cores and a 192-bit wide memory interface, holding 2 GB of memory.

MSI's GeForce GTX 660 lineup consists of a pair of custom-design graphics cards that take advantage of the company's trusty TwinFrozr III cooling solution. The series consists of a base-model, clocked at 980/1033/6008 MHz (core/Boost/memory), and an OC variant, clocked at 1033/1098/6008 MHz. The base model is priced at US $239, with the OC version going for $259. Based on the 28 nm GK106 silicon, the GPU packs 960 CUDA cores and a 192-bit wide memory interface, holding 2 GB of memory.