NVIDIA Logo
Hardware

Rooting for your favorite graphic card giant is a primary part of any PC Enthusiast, but while it is almost customary for a tech enthusiast to do so, it is less becoming for the giants themselves to engage in low blows and PR Battles. The recent tit-for-tat PR Battle being fought between AMD and Nvidia is about Mantle Titles and the Game Works Library. Both Red and Green are trying their level best to gain the upper hand in being seen as the champion of PC Gaming, but they only thing they are currently achieving is making the trolls reach for popcorn.

NVIDIA GameWorks

Nvidia and AMD’s PR Team Exchanges Low Blows to be seen as the Champion of the PC Gaming World

Although it would seem unintuitive, I guess basic information about the spat is in order. Sometime ago AMD accused Nvidia of using the Game Works library in a negative way. Basically they said that Nvidia was making Game Works completely optimized to run on Nvidia Geforce cards while doing absolutely nothing about Radeon cards. The result is obviously that Game Works titles will perform superiorly on Geforce cards. Nvidia denies this accusation but fairly speaking there is no way to verify Nvidia’s claims because Game Works is a proprietary library and its code is not available to the public. Therefore any optimization has to be done by Team Green themselves. An admittedly tricky situation with no actual way to verify anything, apart from Game Works becoming open, which isn’t happening. However when AMD’s Robert Hallock accused Nvidia of optimizing Watch Dogs primarily for Geforce cards, Nvidia revealed a counter statement which said that the source code in question wasn’t actually necessary for driver optimization and AMD’s accusation s were just PR attacks. Here is the detailed response taken from Forbes.com:

Intel-Logo
Hardware

We just received a major leak that updates the original Intel Skylake leak published some time ago. We have now learned more details about the Skylake’s different variants especially the desktop ‘S’ variant including the TDP, eDRAM and Memory configurations. Needless to say this is a very big leak and is going to be a detailed post, so buckle up.

Intel Skylake

Intel 14nm Skylake Details Leaked -Configurations, TDP, Memory and eDRAM Detailed

Alright, let me begin by saying that since this is an aggregate post which has the latest leak included the sources are numerous. We have our Own Sources, Sweclockers and CPU World. You may recall that Skylake is coming in 4 flavors, namely the U/Y/H and S variants. The ‘S’ variant is the variant intended for desktops while the U, Y and H variants are all BGA packaged variants intended for the mobile platform. We received a very extensive description that updates our original leak. Before we had a hazy idea about the different TDP ranges but now we can assign those TDP ranges to the specific eDRAM and Graphics Configurations. Basically one of the biggest points of Skylake is that it will feature support for  DDR4 Ram and will be based on the 100 Chipset ‘Sunrise Point’ of the ‘Sunrise Bay’ platform.

  • Now Variant ‘U’ and ‘Y’ will integrate this PCH (Platform Control Hub) on the die while the ‘S’ and ‘H’ variants will not and require the chip separately.
  • The DMI 2.0 is getting replace with the DMI 3.0 Interface that promises speeds of upto 8 GT/s.
  • Variants ‘U’ and ‘Y’ will support 1 DIMM Slot per Channel while the ‘S’ and ‘H’ Variants will support 2 DIMMs per channel.
  • The IVR (Internal Voltage Regulator) is to be abandoned for Skylake Processors.
Tegra K1 Logo
Hardware

Mobile chipsets have bragged on many an occasion of being more powerful than consoles and old generation  GPUs. Most of the time, however, that only deals with the raw graphics horsepower and not the actual capability of the chip. Nvidia at the Google I/O Event demoed its Tegra k1 chip running a beautiful demo of the Unreal Engine 4.0 , and best of all, this wasn’t some gimped down mobile version, it was the full desktop renderer.

GameScreenAinlineOne word: wow.

Nvidia Tegra K1 Shows Off the ‘Rivals’ Demo – DirectX 11 to Android AEP (ES 3.1) Port Shows Great Results

Nvidia has revealed that the Tegra K1 64 Bit version that will be based on the ARMv8 Architecture is already being worked on by Google and will support the 64 Bit Android L from the get go. Not only that but Epic demonstrated the ‘Rivals’ Demo running on a Tegra K1 chip with a renderer identical to one used in full  fledged desktop games. The amount of eye candy is simply breathtaking although I have a feeling that  we are looking at the thermal envelope of bigger devices than slim smartphones and mobiles. One thing is for certain, the demo was absolutely gorgeous considering the fact that its running on a mobile chip. See for yourself:

AMD R9 Series Logo
Hardware

I have a very interesting article today. Normally I wouldn’t post simulated benchmarks from any source but the man behind the curtains is so authentic that he is literally, single handedly, responsible for the entire R9 295X/Hawaii XTX leak. You might remember DG Lee from some months ago whence he successfully reverse engineered the existence of the Hawaii XTX Core and therefore the alleged R9 295X. DG Lee goes through multiple variations of the simulation with wildly varying results and even a variation with stacked DRAM where he uses my HBM table.

AMD Hawaii XTX GPU

AMD’s R9 295X with the Hawaii XTX Core gets Simulated Benchmarks with Varying Results – 5% Improvement without HBM

Alright lets dive right in. First of all the source page is here. But if his history is any indicator it could be pulled at any second. He uses his home made algorithm that employs the Stream Processor Count, Texture Mapping Unit Count, Raster Operator Count, GPU and RAM Clock Speed and the Bus Width to predict (very accurately) the performance of a Radeon GPU (the R9 295X) in this case. This simulated benchmarks also unveils the probable performance of the complete and fully uncut Hawaii XTX Core. Without any further ado the first benchmark simulates the GPU with the unlocked core ‘as is’ and clocks of the R9 290X.

NVIDIA Logo
Hardware

Nvidia unveiled a relatively new design two days ago, a design which green claims is a world’s first and it probably is. This venture marks the entry of the ARM Cores into the HPC Market for the first time. Basically Nvidia has coupled two ARM processors both with 8 cores with 2 Tesla K20 GPUs for a server ‘block’ style HPC Solution. Tesla K20

Nvidia Pairs 64 Bit ARM Octacore with the Tesla K20 GPU for World’s First ARM Powered HPC Solution

This information comes from the folks over at Computerbase.de who received the information in a press deck fo the International Supercomputing Conference 2014. Here Nvidia presented the ARM HPC with GPU Acceleration ‘block’ which you can see below. The block is actually from Cirrascale in a shallow 1U rack with two sections each containing one ARM + GPU couple. After some time 3U systems are planned as well and the company Eurotech promises to bring Liquid Cooler variants as well. These solutions can be used for Public or Private Cloud, HPC, and enterprise applications. Along with your usual seismic signal  and image processing, video analysis, trace analysis and web applications. The list is endless, and this combo can pack a deadly punch. The GPU Accelerator will handle all the grunt work while the ARM fires off the orders (only Intel’s Co-Processors can task themselves). The Micro X-Gene ARM 64 CPU will be clocked at 2.4 Ghz and will be coupled to a Tesla K20. The server style is allegedly micro-ATX. The motherboard will feature support for the X-Gene ARM CPU (duh) along your usual suspects such as the PCI E 3.0 slot with 8 lanes.

AMD Mantle logo
Industry

Interesting developments are now taking place regarding graphic APIs today, related specifically to AMD’s Mantle API. AMD has developed Mantle as a replacement to Direct 3D and Open GL. AMD hopes to eliminate bottlenecks created in application and game speed by allowing developers to access the hardware more ‘closely’. It aims to do that by allowing more effective scaling on CPU cores and allowing greater control over the graphics pipeline. Mantle will allow GCN GPUs to better utilize TressFX and TrueAudio programs. AMD is to bring Mantle technology to Linux and Linux based gaming devices. AMD has also prioritized its relationship with developers to develop apps and games that include Mantle support, which will also help drive sales of its GPUs. Mantle is currently in the development stage, ith beta testing taking place.

amd-mantle-1

Intel Asks AMD For Mantle Access Yet Continues Commitment For Other Standards. AMD Denies Request Claiming That Mantle Is Still In Testing Stages

Mantle was launched by AMD last year as a low level API which provides a low level code path for developers. PCWorld now reports that Intel has approached the company, expressing interest at an experimental level in the API. Coming towards market share and other gaming libraries, AMD’s main competition is Microsoft’s DirectX and Open GL and Nvidia’s own PhyX engine called Gameworks. Intel currently controls two-thirds of GPU market share and has not committed itself specifically to any API. An Intel spokesman further elaborated: “At the time of the initial Mantle announcement, we were already investigating rendering overhead based on game developer feedback,” an Intel spokesman said in an email. “Our hope was to build consensus on potential approaches to reduce overhead with additional data. We have publicly asked them to share the spec with us several times as part of examination of potential ways to improve APIs and increase efficiencies. At this point though we believe that DirectX 12 and ongoing work with other industry bodies and OS vendors will address the issues that game developers have noted.”

KFA2HOF_4-pcgh
Hardware

Taiwanese manufacturer Galaxy microsystems has launched today its own version of Nvidia’s GeForce GTX 780. Nvidia’s GeForce GTX 780 itself has some very impressive specifications. The card features a base clock frequency of 863 MHz with a boost frequency of 900 MHz. This is coupled with a total of 2304 CUDA cores. The GeForce GTX 780 comes with 3 GB of GDDR5 with a speed of 6.0 GB/s and 384-bit width. The card also features one 8-pin and one 6-pin power connectors. Galaxy today launched their own Hall of Fame + OC version of the card, and the product has several gains over Nvidia’s reference design.

KFA2HOF_2-pcgh

Galaxy Hall Of Fame OC GTX 780 To Feature Hybrid Cooling System And 20% Increase In Clock Speed.

Galaxy’s GTX 780 Hall Of Fame + OC has several important features. The card has a base clock speed of 1019 MHz. This is a 20 percent increase from the frequency of the GeForce GTX 780. The base frequency on the GeForce GTX 780 is 863 MHz. In addition, the card can also be overclocked to a turbo frequency of 1091 MHz. The GeForce GTX 780 can be overclocked to 900 MHz on the other hand. In the case of memory however the GTX Hall of Fame + OC stays the same. The card has the same 3 GB of GDDR5 graphics memory, which is the same as the GeForce GTX 780. The clock speed of the memory on the Hall of Fame GTX 780 is 3004 MHz with a transfer rate of 288.4 GB/s and an interface with of 384 bits.

AMD FirePro W8100 GPU
Hardware

AMD is expanding their latest FirePro series with the new FirePro W8100 series professional graphics card based on the Hawaii GPU architecture. The Hawaii GPU leverages performance of the FirePro graphics cards versus the previous generation which utilized the Tahiti GPU architecture.AMD FirePro W8100 GPU

AMD Unleashes Hawaii Based FirePro W8100 Professional Graphics Card

The AMD FirePro W8100 is the next generation Hawaii GPU based professional graphics card which features over 4.20 Tera Flops of Single-Precision and 2.10 TFlops of Double-Precision performance. Previously spotted at AMD’s Computex 2014 booth, the FirePro W8100 was confirmed to feature the Hawaii-GL Pro chip which means the card features 2560 Stream processors, 160 TMUs and 64 ROPs compared to its bigger brother the FirePro W9100 which includes the Hawaii-GL XT chip with 2816 Stream processors, 176 TMUs and 64 ROPs.

With 1/2 FP64 rate the card churns out impressive compute performance which enables content creators and professionals to harness the full power of the Hawaii GPU. The core clock for the FirePro W8100 graphics card is maintained at 824 MHz as opposed to 930 MHz on FirePro W9100 and also comes with a lower VRAM of 8 GB versus 16 GB GDDR5 memory on the W9100 clocked at 1250 MHz (5 GHz effective) memory clock churning out 320.0 GB/s memory bandwidth.

AMD R9 Series Logo
Hardware

As you may remember, we posted about a leak by the reliable Dr D.G.Lee which claimed that the Hawaii XT core was actually a cut down variant of the actual die. From thence speculation started about the existence of the full core which is called Hawaii XTX. The existence of said core was kept under wraps for AMD to have a competitive strategy against any intermediary cards from Nvidia. However another site, Videocardz.com has now confirmed its existence so we can safely say that this is now pretty much a done deal.AMD Hawaii XTX GPU

The real Hawaii eruption never happened, so buckle up. Courtesy UDteam.Tistory!

Fully Unlocked Hawaii XTX Could Feature 3072 Stream Processors and 48 CUs – Is this the Flagship AMD Radeon R9 295X?

So things have just gotten much more interesting. Technically AMD’s hasn’t released its flagship yet and this new flagship with the Hawaii XTX core will be pitted directly against the GM204 Maxwell in all probability. Now D.G. Lee pulled the leak nearly instantly (and I am not going to link a dead URL here) but if you go through to our original post you can still find it. There is no word on the exact launch date but AMD’s Hawaii XTX GPU should be launching at about the same time as the Maxwell GM204 for maximum impact. It will come down to a battle between the GTX 880 and the AMD R9 295X, sadly a war that would be fought on the 28m node. Here are the specifications as revealed by DG Lee:

Radeon Logo
Hardware

Reports from Overclockers.Co.Uk have emerged which seem to indicate that successor of AMD’s flagship Radeon R9 290X graphics card might possibly in-development. The new graphics card is being positioned to tackle NVIDIA GeForce GTX 780 Ti which still holds the single-chip performance crown whereas the dual-chip throne is held by AMD’s Radeon R9 295X2.AMD Radeon R9 290X Successor Hawaii XTX

Image Credits: TechSpot

AMD Could Already Have A Radeon R9 290X “Hawaii XT” GPU Successor

In comments posted by overclocker 8Pack over at OCUK forums which were spotted by Videocardz, the user tells that the Radeon R9 290X successor is definitely on the way. Surprisingly, if you take a closer look at the comment, 8Pack seems to confirm the existence of not one but two separate chips. In the original thread, 8Pack mentions that AMD already have their next-generation high-end part ready for a while and that AMD also have a Hawaii chip fully enabled, in-development under their sleeves. We might have not taken this seriously if this was a post from a random forum but this coming out from 8Pack who’s a renowned overclocker and has links with several tech companies is quite surprising indeed.

Now you may remember my old article which explicitly mentioned that the Hawaii GPU may have extra compute units left disabled? The article being a rumor was quickly debunked by AMD’s Dave Baumann (Technical Marketin Manager at AMD) at Beyond3D Forums. The report which was posted by industry insider Dr.Lee or DGLee  unveiled some critical details of the fully enabled Hawaii chip and mentioned that the Hawaii GPU actually holds 48 compute units which amount to 3072 stream processors. If AMD does have this chip prepared for launch, then this could be a major attack on the GeForce GTX 780 Ti which has no successor in sight planned till Q4 2014 as hinted in the recent rumors. The upcoming GeForce GTX 880 and GeForce GTX 870 which are expected to launch with the latest Maxwell core and at lower cost than the GeForce 700 series cards will arrive in either October or November time-frame.

FirePro Logo
Hardware

AMD is increasing focus on the FirePro and there fore GPGPU side of things. It already has major design wins in the form of the Mac Pro where AMD cards are used but now not only does it want a more fair share of the cake that Nvidia is eating but it has also introduced the FirePro S4000X and FirePro S9150 (which is the passive version of W9100).AMD S4000X MXM GPU

AMD Wants to Invest More on the FirePro Side of Things – Want’s a Piece of the Pie Nvidia is Eating

Currently Nvidia has a major lead in GPGPU sector whose high end is dominated by Tesla and Quadro graphics (the video industry has an almost perfect monopoly for Nvidia Cards). But AMD is looking to change all that, infact in a niche presentation given on Computex (which we have just now received thanks to some friends) they revealed that they are going to increase investment in the FirePro side of things. This investment will be both on the software and media level, which means that they will not only increase money on R&D but on Marketing as well. Also the AIB Sapphire has pledged its support and since it is highly active in Europe thats a good sign.

Lisa Su and David Cummings mentioned the OpenCL way that is the path of choice for FirePro cards and how the open compute platform provides better flexibility than CUDA and Quadro/Tesla counterparts. It was at this presentation that two new models, that have only been heard off on paper before were shown. The first of these models were the FirePro S4000X which is an MXM Module based GPU designed specifically for Blade Servers.  It has a peak single precision capacity of 992 Gflops and houses 640 processing units while sipping just 45W of juice. The GPU is Cape Verde. The second card shown was the AMD FirePro S9150, which is just the passively cooled model of AMD’s influential W9100 Professional Graphic Card.

AMD R9 Series Logo
Hardware

[RUMOR] Wait, Stop! Notice the rumor tag. Alright so heres the deal, a couple of weeks ago our very reliable colleagues at Videocardz revealed that Tonga GPU would be manufactured by GlobaloFoundries (GloFo) so you can imagine our surprise when we received a completely contradictory point of view from Russian Overclockers. They state that Tonga GPU will actually be manufactured by TSMC. Their report can be found here.

AMD Tonga GPU

AMD Tonga GPU will be Fabricated on TSMC’s 28nm Node and Land in Autumn Claims Russian Report

This is a rather short update, but it deserved posting. Now the primary reason I have put the rumor tag is that although Overclockers Russia is reliable, it directly contradicts VCZ, which I consider more reliable. Russian Overclockers did not state their sources but that is something to be expected when exclusive news sources are concerned. They also state that Tonga is going to land in Autumn. Tonga is basically the an island family (and a State) which consists of 176 Islands. AMD’s nomenclature has always been partial to volcanic islands and landscapes, and the Tonga Nomenclature is no different.

We have already heard reports of the R9 M295X and the expected GPU Counterpart is the R9 275X (since its performance is theoretically somewhere between the R9 270X and R9 280). Now a GPU is expected within 3 Months but once again everything we know is based entirely on leaks and no first hand information so a pinch of salt applies everywhere. TSMC has previously reported a GPU Market Share of 100% and this russian report would actualy verify that claim because if Tonga GPU is manufactured by GloFo then TSMC cannot really have a 100% GPU Market Share. Even though to be fair to TSMC the report was published when  Tonga GPU was still not (and still isnt) released.

NVIDIA Logo
Hardware

The air has been quite abuzz with Nvidia news these days. The most recent of these was information about a GM204 die featuring a 256 bit wide bus. This contradicted with previous speculation regarding Nvidia’s GeForce GTX 880, according to which the GPU was to feature a 512 bit wide bus. Before that we received some important details about the GeForce GTX 880. According to these details the GeForce GTX 880 is to be cheaper and more powerful that the GeForce 700 series. And lastly, yesterday we received some concrete information about the launch details of GeForce GTX 880 and 870. Well, we have more news for you today folks. Today Nvidia officially launches its entry level GPU, the Nvidia GeForce GT 730.

184a

Nvidia GeForce GT 730 Launched With Three Variants. Clock Frequency Starts At 908 MHz.

Nvidia’s GeForce GT 730 is based on the same 28mm GK 107 as the GeForce GT 740. This essentially makes the card a toned down version of the GeForce GT 740, meaning that it features similar components but lower clock and DDR3 frequencies. Nvidia has launched the GeForce GT 730 in three variants. Two of the variants launched feature DDR3 memory. The difference between them lies in interface width, with one featuring a 128 bit interface and the other a 64 bit interface. The third GeForce GT 730 variant comes with GDDR5 memory, effectively making it the fastest of the three. The DDR3 variant with the 128 bit interface features 96 CUDA cores and has a core clock of 700 MHz. Memory clock for all the DDR3 models is at 900 MHz, translating into an effective clock of 1800 MHz.

AMD Mantle logo
Hardware

Alright, finally some really good news. AMD has just dropped a massive hint on their Blog here, which implies that AMD will start working on Linux or has started working on Linux. This is a very big development because just 3 months ago, not a single line of Mantle-Linux existed. I don’t want to go on a rant here but this is the way forward for Mantle API since it will soon face heavy competition on the Windows Side from DirectX 12.

Mantle API White Paper

Mantle-Linux API Is Coming – Could Make SteamOS and Linux a Competitive Gaming Platform

Just wait a second and think, Mantle API + Steam OS, both complimentary to each other and when combined a potential game changer. Steam machines are nothing more than a glorified PC in a fancy packaging running a linux distro (SteamOS). But if Mantle API actually comes to Linux then they could get just the right kind of edge to penetrate the console market. Anyways, here is the actual bombshell taken from AMD’s Blog:

On March 20 Microsoft announced DirectX® 12, the next major evolution of its own game API. This is terrific news because it really draws attention to the value of low-level programming and Mantle’s leading contribution. With DirectX 12 games still over 18 months away and no alternatives in sight for Linux gamers, Mantle’s future looks bright.

richland apu a10-6790k
Hardware

Our colleagues over at CPU World recently sent us some very interesting details regarding the PRO Series APU, which is a fancy name for OEM-Only Kaveri APUS with a locked multiplier. This is one of the reasons why we didnt cover the PRO Series in detail, since a name change and a locked multiplier didnt really warrant detailed coverage. However they have just updated their lineup with new APUs that are actually not present in the original Kaveri Series, infact they are almost certainly from the Richland Series.

AMD Richland Box

AMD PRO Series APUs Get a Kaveri and Richland SKU Update

Now the PRO-Series is exclusively for OEMs and B2B transactions, basically the ‘Business Editions’ of AMD’s Kaveri Lineup. I guess an intro is in order since we have not covered these in detail before. These APUs will have a much longer ‘life support’ package and will be installed in OEM systems. Uptill yesterday this series consisted entirely of 7 Kaveri SKUs (4 Desktop and 3 Mobile) that had their multipliers locked. All these Kaveri APUs consumed only 65W TDP with the exception of the flagship which consumed 95W. The flagship is the A10 PRO-7850B where the ‘B’ stands for Business Class. This APU is the obvious counterpart of the A10-7850K of the mainstream APUs. Now however there has been an addition of two SKUs and one of the these is not Kaveri and judging from the Graphic portion it is a Richland APU.

As CPUWorld states “Specifications of these models were spotted (PDF file) in QuickSpecs of HP EliteDesk 705 G1 business PC.” Now the SKU we are concerned about is A4 PRO-7300B. This APU actually has the highest clocks in the entire lineup with the base clock being 3.9 Ghz and Boost of 4.2 Ghz. It features 1MB of L2 Cache and DDR3 1866Mhz. This APU does not feature an R5 or an R7 series GPU, instead it features the 8470D graphics, which implies its nature as a Richland core. Here is the full lineup of the PRO-Series ‘Business Class’ AMD SKUs:



Creative Commons License This work by WCCF (Pvt) Ltd. is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
Permissions beyond the scope of this license may be available on the Terms and Conditions page. WCCF 2011 is designed by UzEE, Inc.
Ran 32 queries in 1.292  seconds.