cuda
Latest
NVIDIA GeForce GTX 480M reviewed: fastest mobile GPU to date
It's one thing to have a product called the world's fastest on paper, but it's another thing entirely to have the benchmarks confirm it. NVIDIA's GeForce GTX 480M has been swooned over for months now, but it just recently hit the hardcore review sites in a big way. Frankly, there's not a whole lot to say about the thing at this point: it's simply the fastest mobile GPU to date, with Hot Hardware finding it to be "significantly faster in nearly all gaming benchmarks," with just one title showing the Mobility Radeon HD 5870 as the champ by only a few frames. If you've been searching for the fastest mobile GPU in town, you're wasting your time looking any harder; 'course, all of that power consumes an insane level of energy, so true road warriors will certainly want to look elsewhere. Critics pointed out that energy consumption and excess heat were real issues, though both of those are easily overlooked when you're able to take a beastly laptop to a LAN party rather than your desktop. Give those links below a visit if you still need proof.
NVIDIA GeForce GTX 480M: 'world's fastest' mobile GPU now official, landing in June
NVIDIA has just announced that the GTX 480M, the mobile re-spin of its extravagantly overpowered Fermi desktop parts, will be arriving in the middle of next month. Touted as the world's fastest mobile graphics processor, this chip will bring 352 CUDA cores and a 256-bit memory interface to up to 2GB of GDDR5 RAM. These are clear and convincing advances over the incumbent Green Team leader, the 360M, but things start to look a little worrying when we check the 480M's clock speeds. The processor speed is nearly halved from the desktop GTX 480, at 850MHz, the memory does only 1,200MHz, and the graphics run at 425MHz -- we didn't know anything worthwhile even operated below 500MHz these days. Either way, you're getting a computing powerhouse, with the 480M's 897 gigaflops comfortably dwarfing its predecessor's 413 and promising almighty tessellation performance. What it all means with regard to keeping your frame rates up while traversing the Terminus Systems, we can't yet say. We'll let the benchmarking gurus figure it out -- go past the break for the full press release and spec sheet.
NVIDIA GTX 465 detailed ahead of June 1 launch, GTX 460 also rumored
Hey, you there, wanna check out some unreleased new hardware from NVIDIA? Donanimhaber, the same site that brought us early (and accurate) specs of the GTX 470 and 480, is back with the above shot of an ASUS-built GTX 465 that is intended to entice the more mainstream crowd into the Fermi party. Essentially a GF100 with its wingtips trimmed, this'll offer the same 607MHz graphics and 1,215MHz processor clock speeds of the 470, but comes with a narrower 256-bit memory interface, a tamer 3.2GHz effective memory clock, 352 rather than 448 CUDA cores, and -- happily -- lower power requirements. This word is also corroborated by Bit-tech, whose scribes expect a launch at the start of Computex. The Turkish site also has news of a GTX 460, which they say will show up in the middle of July with 768MB of GDDR5, while also being "cheaper and slower" than the 465. Hit the Hexus link below for some early benchmarks of the latter card.
NVIDIA VP says 'Moore's law is dead'
NVIDIA and Intel haven't been shy about their differing respective visions of the future of computing in the past year or so, but it looks like Team GPU just upped the rhetoric a little -- a Forbes column by NVIDIA VP Bill Dally argues that "Moore's law is dead." Given that Moore's law is arguably the foundation of Intel's entire business, such a statement is a huge shot across the bow; though other companies like AMD are guided by the doctrine, Intel's relentless pursuit of Gordon Moore's vision has become a focal point and rallying cry for the world's largest chipmaker. So what's Dally's solution to the death of Moore's law? For everyone to buy into parallel computing, where -- surprise, surprise -- NVIDIA's GPUs thrive. Dally says that dual, quad- and hex-core solutions are inefficient -- he likens multi-core chips to "trying to build an airplane by putting wings on a train," and says that only ground-up parallel solutions designed for energy efficiency will bring back the golden age of doubling performance every two years. That sounds fantastic, but as far as power consumption is concerned, well, perhaps NVIDIA had best lead by example.
NVIDIA GTX 480M will bring Fermi to laptops this June, crazy power requirements and all
We had an inkling NVIDIA wouldn't keep the Fermi goodness just to the desktop and here's our first pseudo-official confirmation. Rushing in ahead of any announcements, Eurocom has started listing a GeForce GTX 480M part, replete with 2GB of GDDR5 memory and a $345 markup relative to ATI's Mobility Radeon HD 5870. It's not clear whether the 100W number refers to the TDP or power requirements of NVIDIA's new GPU, but it's safe to expect both to be pretty high. The MXM 3.0b interface provides a 256-bit linkup between the GPU and CPU, lending plenty of bandwidth, but it also demands plenty of PCB real estate. As a result, Eurocom is offering the GTX 480M on its 17-inch Cheetah and Panther and 18.4-inch Leopard desktop replacements, but not on its 15.6-inch Cougar. Man, no love for the Cougars. According to the listing, we're only a month or so away from release. [Thanks, Jacob]
NVIDIA unleashes GeForce GTX 480 and GTX 470 'tessellation monsters'
Let's get the hard data out of the way first: 480 CUDA cores, 700 MHz graphics and 1,401MHz processor clock speeds, plus 1.5GB of onboard GDDR5 memory running at 1,848MHz (for a 3.7GHz effective data rate). Those are the specs upon which Fermi is built, and those are the numbers that will seek to justify a $499 price tag and a spectacular 250W TDP. We attended a presentation by NVIDIA this afternoon, where the above GTX 480 and its lite version, the GTX 470, were detailed. The latter card will come with a humbler 1.2GB of memory plus 607MHz, 1,215MHz and 1,674MHz clocks, while dinging your wallet for $349 and straining your case's cooling with 215W of hotness. NVIDIA's first DirectX 11 parts are betting big on tessellation becoming the way games are rendered in the future, with the entire architecture being geared toward taking duties off the CPU and freeing up its cycles to deliver performance improvements elsewhere. This is perhaps no better evidenced than by the fact that both GTX models scored fewer 3DMarks than the Radeon HD 5870 and HD 5850 that they're competing against, but managed to deliver higher frame rates than their respective competitors in in-game benchmarks from NVIDIA. The final bit of major news here relates to SLI scaling, which is frankly remarkable. NVIDIA claims a consistent 90 percent performance improvement (over a single card) when running GTX 480s in tandem, which is as efficient as any multi-GPU setup we've yet seen. After the break you'll find a pair of tech demos and a roundup of the most cogent reviews.%Gallery-89115%
Digital Storm's Davinci workstation gets down with Core i7-980X, Quadro graphics
Creative professional. Hear that much? If that's how this all-too-structured world views you, you just might be due a new rig -- particularly if you're thinking of stepping into 4K territory. Digital Storm is offering up a rather unique solution in its Davinci, which opts for a 3.33GHz Core i7-980X Extreme Edition (yeah, that new Intel chip) instead of a more traditional Xeon. You'll still get an NVIDIA Quadro FX 1800 (768MB) GPU, 12GB of DDR3 memory, Windows 7 Processional and one of the nicer liquid cooling systems that we've seen, which may or may not be enough to sneak a little Crysis in between edits. Too bad the base price base rings up at $4,995, but look, that next indie film you're producing is totally hitting it big.
NVIDIA GeForce GT 340 highlights introduction of 300-series cards, none are powerful enough to matter
Is there a tribunal where you can bring up marketing teams for crimes against common sense? NVIDIA's epic rebranding exercise knows no bounds, as the company has now snuck out its very first desktop 300-series cards, but instead of the world-altering performance parts we've always associated with the jump into the 300s, we're getting what are essentially GT 2xx cards in new garb. The GT 340 sports the same 96 CUDA cores, 550MHz graphics and 1,340MHz processor clock speeds as the GT 240 -- its spec sheet is literally identical to the 240 variant with 1,700MHz memory clocks. To be fair to the company, these DirectX 10.1 parts are exclusively for OEMs, so (hopefully) nobody there will be confused into thinking a GT 320 is better than a GTX 295, but we'd still prefer a more lucid nomenclature... and Fermi graphics cards, we'd totally like some of those too.
NVIDIA Tesla 20-series GPUs promise to dramatically cut supercomputing costs
Sure, you've been hearing NVIDIA toss around names like CUDA, Fermi and Tesla for what seems like ages now, but we're guessing this is the sort of thing that'll get most folks to really take notice: a promise to cut supercomputing costs by a factor of ten. That rather impressive feat comes courtesy of the company's new Tesla 20-series GPUs, which come in the form of both single GPU PCI-Express Gen-2 cards and full-fledged GPU computing systems, and promise a whole host of cost-saving benefits for everything from ray tracing to 3D cloud computing to data analytics. Of course we are still talking about "cheap" in supercomputing terms -- look for these to run between $2,499 and $18,995 when they roll out sometime in the second quarter of 2010.
EVGA GeForce GTX 275 co-opts a GTS 250 for PhysX duties
Ready for some more dual-GPU madness, only this time in the resplendent green of NVIDIA? EVGA has gone and concocted a special Halloween edition of the GTX 275, which has sprouted an entire GTS 250 appendage solely for PhysX gruntwork. Dubbed a new form of Hybrid SLI, EVGA's latest combines -- for the first time, from what we can tell -- two different GPUs and assigns them with specific and mutually exclusive tasks. Whether this concept takes off will depend to a large extent on the effectiveness of PhysX acceleration and whether it can show more efficient scaling than regular old SLI with two boards or more conventional dual-GPU setups like the GTX 295. Color us intrigued, either way. P.S. - That's what the actual card will look like, we're not making it up.%Gallery-77092%
NVIDIA launches Fermi next-gen GPGPU architecture, CUDA and OpenCL get even faster
NVIDIA had told us it would be accelerating its CUDA program to try and get an advantage over its competitors as OpenCL brings general-purpose GPU computing to the mainstream, and it looks like that effort's paying off -- the company just announced its new Fermi CUDA architecture, which will also serve as the foundation of its next-gen GeForce and Quadro products. The new features are all pretty technical -- the world's first true cache hierarchy in a GPU, anyone? -- but the big takeaway is that CUDA and OpenCl should run even faster on this new silicon, and that's never a bad thing. Hit up the read links for the nitty-gritty, if that's what gets you going. Read - NVIDIA Fermi siteRead - Hot Hardware analysis Read - PC Perspective analysis
ATI Stream goes fisticuffs with NVIDIA's CUDA in epic GPGPU tussle
It's a given that the GPGPU (or General-Purpose Graphics Processing Unit) has a long, long ways to go before it can make a dent in the mainstream market, but given that ATI was talking up Stream nearly three whole years ago, we'd say a battle royale between it and its biggest rival was definitely in order. As such, the benchmarking gurus over at PC Perspective saw fit to pit ATI's Stream and NVIDIA's CUDA technologies against one another in a knock-down-drag-out for the ages, essentially looking to see which system took the most strain away from the CPU during video encoding and which produced more visually appealing results. We won't bother getting into the nitty-gritty (that's what the read link is for), but we will say this: in testing, ATI's contraption managed to relieve the most stress from the CPU, though NVIDIA's alternative seemed to pump out the highest quality materials. In other words, you can't win for losin'.
Maingear's eX-L 18 grabs for "world's most powerful gaming laptop" title
You know, this whole "most powerful laptop" game is pretty hilarious. Back in '05, CompAmerica's (who?) Orca 9098 held the title with a cutting-edge 3.8GHz Pentium 4, and over the years, we've watched outfit after outfit pull the award back and forth, much like those tugging games we used to play as tots. Regardless, it seems as if Maingear's down for the fun, today rolling out the planet's all new "world's most powerful gaming laptop" in the eX-L 18, which arrives with an undisclosed Intel Core 2 Duo Extreme processor, twin NVIDIA GeForce GTX 280M GPUs, up to 8GB of DDR3 memory, up to three 2.5-inch SATA or SSD drives, an optional Blu-ray drive and a screen that's probably larger than your mother's desktop LCD. And by that, we mean 18.4-inches with a 1,920 x 1,080 resolution. If you're down with lugging around this beast, you can get one headed your way provided you've got at least $2,999 you're willing to see off.
NVIDIA's GT300 specs outed -- is this the cGPU we've been waiting for?
NVIDIA's been dabbling in the CPU space behind closed doors for years now, but with Intel finally making a serious push into the GPU realm, it's about time the firm got serious with bringing the goods. BSN has it that the company's next-generation GT300 will be fundamentally different than the GT200 -- in fact, it's being hailed as the "first truly new architecture since SIMD (Single-Instruction Multiple Data) units first appeared in graphical processors." Beyond this, the technobabble runs deep, but the long and short of it is this: NVIDIA could be right on the cusp of delivering a single chip that can handle tasks that were typically separated for the CPU and GPU, and we needn't tell you just how much your life could change should it become a reality. Now, if only NVIDIA would come clean and lift away some of this fog surrounding it (and the rumored GTX 380), that'd be just swell.[Thanks, Musouka]
ArcSoft's SimHD plug-in takes SD footage to "near HD"
Another month, another upscaling technology. Tapping into NVIDIA's versatile CUDA architecture, ArcSoft has just announced its SimHD upscaling tech, which scales SD and DVD content to "near high-definition quality." The app integrates with the company's heralded TotalMedia Theater, though you'll need an NVIDIA GPU in order to really take advantage. We're told that it's compatible with millions of NVIDIA GPUs from the GeForce 8 Series onwards, including GeForce and Quadro lines. The pain for better looking pixels? $19.95 and a 5MB download, which may or may not end up costing more than the code itself if you use Time Warner Cable.
ASUS N81Vg: first laptop with NVIDIA's GeForce GT 120M
Not quite an ultraportable, but not quite a behemoth -- the 14-inch ASUS N81Vg fits nicely between the two laptop extremes, and given that it's the first to house NVIDIA's GeForce GT 120M graphics card, even gamers can feel free to sneak a deathmatch or two in between conference calls. The rig itself can be ordered with one of many Core 2 Duo processors, up to 4GB of DDR2 RAM, upwards of 500GB of hard drive space, an optional Blu-ray burner, 1.3 megapixel webcam and a battery good for three to four hours. The newfangled 120M GPU features NVIDIA CUDA technology, 32 processing cores, DirecX 10 support, 1080p video playback and 110 gigaflops of computing power. Per usual, ASUS is keeping quiet when it comes to pricing and release details, but it ought not be long now, tiger.
NVIDIA now offering laptop drivers directly through website
NVIDIA has announced that it'll now offer laptop GPU drivers directly via its website -- long overdue, if you ask us. These drivers have traditionally been offered through the computer manufacturers since most mobile GPUs are customized to be compatible with the devices' specific hotkeys and suspend / resume functionality -- NVIDIA said it has found a way around with a new modular architecture. First on the menu are beta drivers for GeForce 8M and 9M series as well as Quadro NVS-series laptops that add CUDA and PhysX support, with Windows-certified drivers for all GeForce 7, 8 and 9 series and Quadro NVS series are due out early next year. Now, if only we could download hugs... [Via Notebook Review]
NVIDIA dishes about OpenCL
We spent some time on the phone with NVIDIA today in the wake of last night's official release of the OpenCL GPU-processing spec, and we learned some interesting things. NVIDIA thinks OpenCL is going to bring a lot more attention to general-purpose GPU computing, and it's planning on stoking the flames -- not only is it accelerating the CUDA release schedule, it's planning on working with Microsoft on DirectX 11 Compute. Hit the break for some more highlights!
NVIDIA's Quadro FX 5800 with 4GB graphics memory is 'the most powerful graphics card in history'
That's some serious boasting by NVIDIA, but this is some serious graphics horsepower. The Quadro FX 5800, already seen in NVIDIA's Quadro Plex D data cruncher, replaces the 5600 at the top of the NVIDIA heap with 240 CUDA-programmable parallel cores and the industry's first card with 4GB of graphics memory. MSRP? Just $3,499 for you big spender -- pennies for the companies who can harness the power for the purposes of oil and gas exploration, 4D modeling, and graphics design.
ASUS G50 / G71 laptops and ARES CG6155 gaming PC now available
It's been awhile (or a long while, in the case of the ARES CG6155) since we've heard about these machines, but ASUS has at long last decided to start shipping 'em to retailers. The aforesaid gaming desktop still doesn't have a publicly available price tag, but ASUS assures us that it's out there now for those who know where to look. Thankfully, it was a bit (and we stress "bit") more forthcoming with details on the G50 and G71 gaming notebooks, which are also available as we speak for $1,249.99 and take-your-best-guess, respectively.Read - ASUS ARES CG6155Read - ASUS G50 and G71
