MojoKid writes: AMD got the attention of PC performance enthusiasts everywhere with the recent launch of its Ryzen 7 series processors. The trio of 8-core chips competitively take on Intel's Core i7 series at the high-end of its product stack. However, with the extra attention AMD garnered, came significant scrutiny as well. With any entirely new platform architecture, there are bound to be a few performance anomalies — as was the case with the now infamous lower performance "1080p gaming" situation with Ryzen. In a recent status update, AMD noted they were already working with developers to help implement "simple changes" that can help a game engine's understanding of the AMD Zen core topology that would likely provide an additional performance uplift with Ryzen. Today, we have some early proof-positive of that, as Oxide Games, in concert with AMD, released a patch for its game title Ashes Of The Singularity. Ashes has been a "poster child" game engine of sorts for AMD Radeon graphics over the years (especially with respect to DX12) and it was one that ironically showed some of the worst variations in Ryzen CPU performance versus Intel. With this new patch that is now public for the game, however, AMD claims to have regained significant ground in benchmark results at all resolutions. In the 1080p benchmarks with poweful GPUs indeed a Ryzen 7 1800X shows an approximate 20% performance improvement with the latest version of the Ashes, closing the gap significantly versus Intel. This appears to be at least an early sign that AMD can indeed work with game and other app developers to tune for the Ryzen architecture and wring out additional performance.
MojoKid writes: Consumer WiFi router products are classified by three major performance characteristics: overall throughput or bandwidth, multi-client performance, and range. Although throughput and multi-client bandwidth has scaled-up over the years, range hasn't improved quite as robustly. Even the most powerful WiFi routers, with active antennas, can still leave dead spots in large home or office installations. That's where the recent crop of mesh router technologies, that startups like Eero and Google with Google WiFi, are making significant advancements. By spreading out multiple, interconnected router access points across a WiFi network, you blanket the area with a stronger, more contiguous signal. If you need to go the distance, mesh WiFi routers are the new way to go and Netgear is now entering the fray with a 3Gbps tri-band setup called Orbi. Where the Orbi is different from recent mesh networking products is its 5GHz, 1733Mbps backhaul connection between its satellite and the base router. A combined two unit system offers a 2X2, 866Mbps, 5GHz AC connection and a 2x2, 400Mbps, 2.4GHz link. However, in between, including Gig-E wired devices that you can plug into a satellite, there's a 4x4, 5GHz dedicated backhaul link that lets client connections stretch their legs. Tested against a powerful standard AC5300 router, the Orbi mesh setup delivered consistent performance well north of 130Mbps, through multiple floors, and upwards of 300Mbps at longer distances, up to 4,000 square feet, with the Orbi satellite on the same level as the client PC.
MojoKid writes: Samsung announced a new family of 960 EVO and 960 Pro NVMe PCI Express M.2 Solid State Drives today. Built on Samsung's 3D V-NAND technology and employing the new Samsung Polaris SSD controller, the 960 Pro is Samsung's highest performance, high endurance drive and the successor to last year's 950 Pro. The 960 EVO is the lower cost model and a follow-on to last year's Samsung 950 EVO drive. The 960 EVO is also powered by the same Samsung Polaris controller but employs more cost-efficient Samsung TLC NAND memory. Both drives arrive in standard M.2 gumstick form factors with PCI Express Gen 3 X4 interfaces and utilizing the NVMe (Non-Volatile Memory Express) protocol for lightning-fast speeds and low latency. Specifically, the 960 Pro offers up to 3.5GB/sec and 2.1GB/sec of sequential read and write throughput respectively, with endurance rated at up to 1200TB writes per day. The 950 EVO's specs drop in at a peak 3.2GB/sec and 1.9GB/sec for reads and writes respectively, with a top-end endurance rating of 400TB written per day. The 960 Pro will come in 512GB, 1TB and 2TB capacities starting at $329, while the 950 EVO comes in 250GB, 500GB and 1TB capacities starting at $129. Samsung will be shipping the drives in October this year.
MojoKid writes: AMD is now shipping a new Polaris-based mainstream graphics card, known as the Radeon RX 470, which will retail in the $149 — $179 price range. The Radeon RX 470 is built around the same Polaris 10 GPU as the Radeon RX 480. However, two CUs have been disabled, which results in fewer stream processors. There are 2048 stream processors active in the RX 470 versus 2304 on the RX 480. The 470's clocks are somewhat lower as well, with base and boost clocks of 926 MHz and 1206 MHz respectively. The end result brings peak compute performance for the RX 470 down to 4.9 TFLOPs (compared to 5.8 TFLOPs for the 480). The RX 470's memory is clocked slightly lower as well, which results in a peak 211GB/s, 13GB/s lower than the RX 480. Considering its sub-$200 price point, the Radeon RX 470 puts up respectable numbers in the benchmarks, falling in right behind the more powerful Radeon RX 480, but typically well ahead of AMD's previous-gen Radeon R9 380X. In comparison to NVIDIA's offerings, the Radeon RX 470 outpaces the GeForce GTX 960 and GTX 950 across the board, and even manages to sneak out in front of the GeForce GTX 970 on a couple of occasions but can't catch the more expensive GeForce GTX 1060.
MojoKid writes: NVIDIA took the wraps off its new Pascal-based, flagship graphics card, dubbed the Titan X, a couple of weeks ago, but only pictures and specifications were available. NVIDIA Titan X cards are now shipping in limited quantities, however, and the benchmark numbers are in. The new Titan X (the company kept the same branding as its previous Maxwell-based Titan), based on NVIDIA's Pascal architecture, is some 60 percent faster than its older, Maxwell-based Titan counterpart and 20 – 30 percent faster than the new GeForce GTX 1080. On board Titan X are 1024 more CUDA cores versus a GeForce GTX 1080 (3584 versus 2560) and a wider 384-bit GDDR5X memory bus versus the 1080's 256-bit interface. Though the Titan X has the same memory clock as a GTX 1080, it has 12GB of GDDR5X memory, versus 8GB on the 1080. In testing, nothing can touch it and there's still additional headroom for overclocking. The new Titan X's $1200 price point, however, will give even hardcore gamers sticker shock. For graphics professionals and deep learning applications it could be a reasonably good value though, versus pro graphics GPUs.
MojoKid writes: Google has been teasing its Project Tango augmented reality (AR) platform for years but no OEMs have stepped up to the plate to deliver Tango-enabled hardware until now. Lenovo just came out with its PHAB2 Pro 6.4-inch phablet smartphone which packs a full-fledged AR experience. The PHAB2 Pro will be the first commercially available Lenovo smartphone in the US and it leverages Tango AR technology in three ways. The smartphone's "eye" uses motion-tracking to determine its location in 3D. Area learning can also feed location information to the phone, and depth perception allows the phone to analyze the world around it. The PHAB2 Pro is also huge with a 6.4" QHD display covered in 2.5D curved glass. Powering the PHAB2 Pro is a Snapdragon 652 processor with 4GB of RAM, a generous 64GB of storage and a microSD slot. There's also a 16MP rear camera, 8MP front camera and a 4050 mAh battery. Lenovo's Motorola Mobility division also announced the Moto Z and Moto Z Force, which are next generation Android flagships. The Moto Z is the standard model and measures just 5.2mm thick and comes with a 5.5" QHD AMOLED display, a Snapdragon 820 processor with 4GB of RAM and up to 64GB of storage. Its 13MP rear camera features optical image stabilization and laser autofocus, while its 5MP front camera with wide-angle lens takes care of selfies. Then there's the new Moto Z Force, which ups the ante with a 3500 mAh battery, a 21MP rear camera and a shatterproof screen. But what truly makes the Moto Z and Moto Z Force stand out are Moto Mods. These are modular accessories that attach to the back of the smartphones via four magnets and a 16-pin connector. It's much more elegant than what LG has employed with the G5 (which requires you to remove the bottom of the smartphone). Instead, Moto Z users can simply attach an accessory, like the JBL SoundBoost Mod which brings high-end sound, with a quick snap.
MojoKid writes: Yesterday, during his keynote address at GTC 2016, NVIDIA CEO Jen-Hsun Huang made a number of interesting announcements. We saw Apple co-founder Steve "Woz" Wozniak take a virtual tour of Mars and witnessed the official unveiling of NVIDIA's Tesla P100, which is based on the company's bleeding-edge, GP100 GPU. GP100 leverages NVIDIA's forward-looking Pascal GPU architecture and features 16GB of HBM2 memory, all built using TSMC's 16nm FinFET manufacturing process. GP100 is a massive chip with a 600mm2 die, which is about the size of current gen high-end Maxwell GPUs but when you consider it's built on 16nm process technology, you can see how NVIDIA was able to squeeze in excess of a 15 billion transistor design. What might be more impressive is that NVIDIA is also actually shipping the part now to select HPC (High Performance Computing) OEMs. Though NVIDIA's on DGX-1 Supercomputer was off limits at the show, an OEM by the name of QTC had Tesla modules racked up and exposed. Getting a closer look at the P100 shows just how large the chip complex is, with HBM2 modules just barely visible on either side of the die.
MojoKid writes: Intel let slip a mention of its forthcoming Core i7-6950X Extreme Edition processor this weekend and it's not an April Fool's Day prank. The new chip appears to be a beefy 10-core CPU. To be clear, Intel hasn't launched the Core i7-6950X, not yet anyway. But what the company did do is update its support website for its latest Management Engine software, listing the much anticipated Broadwell-E part as a compatible chip. Details show the flagship Broadwell-E part will feature a whopping 25MB of L3 cache and a boost clockspeed of 3.5GHz. Since Broadwell has always touted 2.5MB of cache per core, doing the math tells you this will be a 10-core processor. Toss in Hyper Threading support and you have 20 threads to throw at workloads. As for the base clockspeed, Intel lists the boost clock at 3.5GHz so a 3GHz base seems highly plausible. From what we already know about Broadwell-E, the 6950X will be compatible with the current crop of X99 Express chipset motherboards sporting LGA 2011-v3 sockets, though users will likely have to update their BIOS with the latest microcode.
MojoKid writes: Asus unveiled its latest addition to the Transformer series at CES in January, the Transformer Book Chi, which just recently began shipping. Available in three sizes, the new Transformer Book Chi Series features a 2-in-1 detachable design. The flagship Transformer Book T300 Chi offers a 12.5-inch screen, an Intel Core M processor, and a fanless cooling solution. The more compact T100 Chi is a 10.1-inch model that promises all-day battery life. The 2-in-1 detachable design employs a magnetic hinge that supports four usage modes: Attached, Detached, Flipped, and Tented. The T300 Chi measures about 0.65 inches when docked, making it slightly thinner than an Apple Macbook Air. Asus claims the T300 Chi is the world's thinnest Windows tablet, measuring just 0.28 inches thick. Most interestingly, perhaps, is that Asus built this machine with Intel's fastest Core M chip, the Core M 5Y71. In the benchmarks, it competes well even with full-sized ultrabooks, though battery life does take a hit due to the system's mechanical limitations and smaller 31Whr battery.
MojoKid writes: Intel is taking the wraps off of its latest processors for enterprise server and pro workstation applications today, dubbed the Xeon E7-8800 / 4800 v3. Like its high-end desktop processors, the Xeon E7-8800 / 4800 v3 product families are based on the Haswell-EX CPU core. These new Xeons, however, offer a plethora of other enhancements and are packing significantly more cores than any current desktop processor. The highest-end Xeon E7-8800 series processors, for example, are 18 core chips. Previous generation Xeon E7 v2 processors were based on the Ivy Bridge-EX core, while the new E7 v3 parts are based on Haswell-EX, though both are manufactured on Intel's 22nm process node. Next generation Broadwell-EX based Xeons will make the move to 14nm. Xeon E7-8800 / 4800 v3 series processors have 32-lanes of PCIe 3.0 connectivity per socket, TSX is enabled in all SKUs, they offer support for both DDR3 and DDR4 memory (though, not simultaneously), and can address up to 6TB of memory in a 4-socket configuration or 12TB in an 8-socket setup. Intel has also goosed the chip's QPI interface speeds to 9.6GT/s.
MojoKid writes: NVIDIA's Maxwell GPU architecture has has been well-received in the gaming world, thanks to cards like the GeForce GTX Titan X and the GeForce GTX 980. NVIDIA recently took time to bring that same Maxwell goodness over the workstation market as well and the result is the new Quadro M6000, NVIDIA's new highest-end workstation platform. Like the Titan X, the M6000 is based on the full-fat version of the Maxwell GPU, the G200. Also, like the GeForce GTX Titan X, the Quadro M6000 has 12GB of GDDR5, 3072 GPU cores, 192 texture units (TMUs), and 96 render outputs (ROPs). NVIDIA has said that the M6000 will beat out their previous gen Quadro K6000 in a significant way in pro workstation applications as well as GPGPU or rendering and encoding applications that can be GPU-accelerated. One thing that's changed with the launch of the M6000 is that AMD no longer trades shots with NVIDIA for the top pro graphics performance spot. Last time around, there were some benchmarks that still favored team red. Now, the NVIDIA Quadro M6000 puts up pretty much a clean sweep.
MojoKid writes: Intel first offered a sneak peek of their forthcoming Compute Stick HDMI dongle earlier this year at CES but today is officially announcing product availability and has lifted embargo on first tests with the device. The Compute Stick is essentially a fully-functional, low-power, Atom-based system with memory, storage, and an OS, crammed into a dongle much bigger than a USB Flash drive. There will initially be two compute sticks made available, one running Windows (model STCK1A32WFC) and another running Ubuntu (model STCK1A8LFC). The Windows 8.1 version of the Compute Stick is packing an Intel Atom Z3735F processor, with a single-channel of 2GB of DDR3L-1333 RAM and 32GB of internal storage, though out of the box only 19.2GB is usable. The Ubuntu version of the Compute Stick has as a similar CPU, but is packing only 1GB of RAM and 8GB of internal storage. All sticks have USB and MicroSD expansion capability.The device is packing a low-power Atom processor, Intel HD graphics and only a single-channel of DDR3L-1333 memory, so it's not going to burn through any benchmarks. For multi-media playback, basic computing tasks, web browsing, HD video, or remote access, the Compute Stick has enough muscle to get the job done and it's cheap too at $99 — $149.
totalcaos writes: YouTube announced today its plans for an ad-free, subscription-based service by way of an email sent out to YouTube Partners. The email details the forthcoming option, which will offer consumers the choice to pay for an “ads-free” version of YouTube for a monthly fee. The additional monetization option requires partners to agree to updated terms on YouTube’s Creator Studio Dashboard, which notes that the changes will go into effect on June 15, 2015.
MojoKid writes: Dell unveiled a new Android 2-in-1 today, the Venue 10 7000, which brings with it many of the same hardware features that we saw with their popular Venue 8 7000 8-inch tablet. It's powered by a quad-core Intel Atom Z3580 processor with 2GB of RAM, 16GB of internal storage, and a 2560x1600 10-inch display. You'll also find a microSD slot that supports up to 512GB of additional storage, 802.11ac, Bluetooth 4.0, Miracast, front-facing stereo speakers, a 2MP front-facing camera, and an 8MP Intel RealSense 3D camera on the rear. Where things get more interesting, perhaps, is with the design of the tablet. Whereas the Venue 8 7000 features a more traditional tablet form-factor, the Venue 10 7000 features a cylindrical "barrel edge" which Dell says makes the tablet easier to hold and carry. It's reminiscent of Lenovo's Android-powered Yoga Tablet family. In addition to providing a handy place for your hand to grip the tablet, the cylindrical spine also serves as an attachment point for an optional keyboard that transforms the Venue 10 7000 into a laptop. The keyboard accessory allows the tablet to be used in five different configurations: Tablet Mode (w/o keyboard), Tablet Mode (w/ keyboard), Laptop Mode, Tablet Stand Mode, and Tent Mode.