From its early days, the main focus for OTOY was disruption of the visual effects industry (FX). Split between quick-but-unreliable rasterizing and slow-but-correct ray-tracing, both computer games and movies suffer from the same fate when it comes to rendering physically and perspective-correct worlds. OTOY is working on changing that through their Octane Renderer and Brigade, mixed RT/ROP engine. However, OTOY did not want to stop at creating a physically correct graphics engine. During our visit at the company’s HQ in 2013, OTOY was working on initial deployment of Kepler-based graphics cards, with an initial target of deploying 3,000 GPUs for their cloud rendering business. Under the
In a way, the 2016 GPU Technology Conference represented a ‘coming of age’ for Nvidia, where the company finally established their first proprietary standard that gained immediate industry traction with none other than IBM. The NVLink interconnect, and the Mezzanine connector represent the first custom interfaces (outside of BGA packaging for their silicon) Nvidia designed. Given that IBM’s OpenPOWER conference is taking place at the same time as GTC, we searched for more details about the Mezzanine connector and the NVLink itself, and stumbled on quite an interesting amount of details. First and foremost, ever Pascal (Tesla P100) and Volta (Tesla V100) product that utilizes the NVLink
At the 2016 GPU Technology Conference, Nvidia finally unveiled the Pascal GPU architecture. Perhaps the most interesting aspect of the GPU aren’t the capabilities the Pascal architecture brings, but rather the first non-Intel driven high-end bandwidth interface since AMD launched HyperTransport in 2001. NVLink standard launched in 2014, when IBM announced its tie up with Nvidia to bring the high-speed interconnect to the market. The goal of NVLink is to remove its future GPU architectures from the dependencies of PCI Express, and achieve maximum bandwidth. If NVLink was replaced with 100% PCIe lanes, the design simply would not be as efficient in terms of lines needed, and would
Recently, we posted an in-depth story detailing the birth of Radeon Pro Duo. As we all probably know, AMD created two designs for the dual Fiji GPU. In the end, AMD brought just a single design to market and a server grade one to boot. The company just introduced the FirePro S9300 X2, sharing the same server-grade design as the Radeon Pro Duo, with the main difference being that the FirePro model was clocked down in order to allow for optimal operation using server-grade air cooling. Given that S9300 X2 is rated at 46 GLOPS/Watt and operates within 300 Watt thermal limit (302W to be
As the time for GPU Technology Conference approaches, Nvidia is starting to unveil products that will be hotly discussed in sessions. One such example is Quadro M6000 24GB. This graphics card comes as an answer to customer demands for more on-board memory, satisfying the demands across the board. For example Deadpool’s opening sequence was visualized on the Quadro M6000 12GB, using the same assets in the final render. Expanding on that, 24 GB of on-board memory enables the content creators to take one step further, as seen in the upcoming Angry Birds movie. “At Sony Pictures Imageworks, we regularly push the limits of our ability to display
Seven months ago, Steve Luczo (CEO) stated that “there is no one that is using SSDs for storage.” This statement still haunts Seagate Technologies, LLC to this date, even though it was out of context. “I mean, maybe at the margin for replacing boot drives… I mean, maybe one or three per cent of the hierarchy is SSDs for storage. Most of our flash product is actually not hanging off the storage bus, it is fast memory.” As it usually goes, the dismissive statements come at the time when you know that you have a good thing coming down the line, but you need to hold the fort.
At the Embedded World in Nuremberg, Germany, AMD announced the third generation of Embedded G-Series and LX SoCs, hoping to provide customers a broadened portfolio of performance options. The latest offerings expand ability to scale x86 platforms, starting with the entry-level AMD Embedded G-Series LX SoC, which is pin compatible to the previous generation of SoC devices. Besides low-power processors, AMD also announced two higher performing SoCs, codenamed “Prairie Falcon” and “Brown Falcon,” which introduce pin compatibility between G-Series processors and high-performance AMD Embedded R-Series SoCs. The new products expand upon AMD’s low power capabilities, bringing scalable performance, power, and price across the CPU, GPU, multimedia, and I/O controller hardware, helping to lower
It looks like 2016 is turning into a year of anticipation and redemption for AMD, not just to its consumers, but also to customers which purchased millions of dollars of AMD hardware in the past, and then felt left out. We all saw Oak Ridge National Laboratories, one of first Opteron adopters – ditching a decade old AMD collaboration for IBM+NVIDIA team up. Luckily for all involved, AMD seems to have finally “get their s*** together” and started a sales campaign which might be the most successful since Henri Richard led the sales team taking over more than 50% market share from Intel (albeit only in 4P and 8P
Several years after NVIDIA launched its GRID vGPU architecture, AMD is entering the graphics virtualization market with its FirePro S7150 and S7150 x2 graphics processors. This is probably the last product AMD is launching utilizing its Tonga GPU architecture, also known as ‘Volcanic Islands’ or ‘GCN 1.2’ family of products. FirePro S7150 packs a single Tonga GPU and 8GB of ECC GDDR5 memory, while as the name says, FirePro S7150 x2 (why the small ‘x’?) is a dual-GPU with two Tonga GPUs and 16GB of ECC GDDR5 memory. Hardware capabilities are quite impressive – the products in question can support up to 16 users per GPU. By taking a GPGPU
Buying a compute device – be that a personal computer, workstation or a server, typically comes with a certain set of features which does not change. All workstation vendors for example, offer 1-4 expansion cards, typically a GPU and computational or storage cards. Those features are limited by what motherboard vendors can offer, and ATX/eATX/SSI standards can offer up to X amount of cards. BOXX Technologies decided to challenge that by launching APEXX 5, a custom designed system which can host up to seven expansion cards. For example, a Quadro GPU for sync/display output and five Tesla boards for compute. If you go with Dual-GPU boards, such as Tesla
Even though Solid State Drive (SSD) technology is only now reaching significant levels of popularity, capturing more and more market share, the fact of the matter is that SSD as such is not a new technology. In fact, if memory serves me right – the very first SSD came out in 1990 in the shape of NEC 5.25″ SCSI drive (anyone remembers SCSI cables?). However, this drive was actually based on RAM-chips with a built-in battery. The dawn of modern SSD’s came with the arrival of NAND Flash memory, and in 2005 Samsung made a milestone strategic decision to pronounce Flash memory and SSD as ‘the future’,
AMD’s path to this moment was a difficult one. The company spent $334 million to acquire SeaMicro, an investment which was latter written off completely, and ultimately lead to shutting down the micro-server division. At the same time, 2012 saw the announcement of ARM-based Opteron and its arrival on the product roadmap (and a continuous push back), with ARM-based AMD product being available only as a single processor, 1U half-rack development kit from a third party vendor. This changes today, with the introduction of AMD Opteron A1100 Series processors, based on a silicon codenamed ‘Seattle’ – which may or may not have to do anything with two major customers from that area. In order to bring
The Workstation Performance Characterization Group inside SPEC (Standard Performance Evaluation Corporation) launched a new workstation benchmark at the Supercomputing Conference (SC’15) held this week in Austin, TX. New SPECwpc V2.0 benchmark marks a new chapter for SPECwpc body, which creates professional benchmarks utilized by commercial entities and research institutes in order to find the ultimate hardware performance on a set of industry-standard benchmarks. Most of upgrades were focusing on changes going on with the PC architecture. According to SPECwpc, these are the upgrades inside the benchmark suite: Improved storage workloads that better reflect performance of NAND Flash devices for high-capacity data storage. Better scalability measurement for
At the recently held 2015 HotChips conference, Avinash Sodani (KNL Chief Architect, Senior Principal Engineer, Intel) gave a speech how Intel plans to expand the Xeon Phi product lineup from a server-only, PCIe card concept into three different packages, which would appeal to the workstation and server customers in different fields. On SC’15 Conference, which takes place in Austin, TX – Intel finally confirmed the strategy and is coming out with a workstation product that will feature a fully-enabled Knights Landing (KNL) Many-Core processor. In the first half of 2016, the company will ship Intel-built, Intel-branded workstation powered by self-booting Xeon Phi processor. The processor will be able to boot standard
As the current Hewlett Packard management is proceeding with the company split into two divisions, a move started-and-stopped under the previous management, more jobs are getting cut as the company wants to ‘maximize the shareholder return’ and reduce glut which is somewhat inevitable in an organization that currently employs more than quarter of million people (September 2015: 302,000 employees worldwide). The company management lead by Margaret Cushing ‘Meg’ Whitman claims the latest round of cuts will reduce in a one-time charge of 2.7 billion USD, and produce annual savings in the same range. Two billion dollars are to be saved in the Enterprise division, while the ‘PCP’ division should save 700
Couple of months ago, we exclusively reported that the U.S. government blocked Intel from selling its products to Chinese supercomputer firms such as Inspur, responsible for building the Tianhe-2 supercomputer. Originally, Tianhe-2 was planned to expand all the way to almost 100,000 Xeon processors and Xeon Phi co-processors, targeting to beat the 100 PFLOPS barrier. Initial deployment included 4,096 Chinese Galaxy FT-1500 processors (Chinese SPARC) and 16,000 processor nodes with two 12-core Xeon E2692 processors and three Xeon Phi 31S1P co-processors for a grand total of 3.12 million cores. Full installation of Tianhe-2 was scheduled to feature 48,000 processor nodes, or 9.93 million cores. However, those
Given the slow disintegration or refocusing of its competitors, we were not all too surprised when Intel started to slow down its famous “Tick-Tock” manufacturing cadence. Originally introduced in 2006 with the “Conroe”, Core 2 Duo processors, Tick-Tock was mixed between a new microarchitecture and current manufacturing process (Tock), and a new process, die-shrink processor with some architectural optimizations (Tick). Further separation in Tick-Tock was a cadence between mainstream parts (desktop, mobile) and high-end parts (gaming, workstations) – Tick or Tock would always debut as mainstream parts, followed by high-end 6-12 months after. High end would typically mean Intel Xeon and Core i7 ‘X’ line-up.
If we remove the borders on this planet, the number of Facebook netizens would place it as the largest country in the world. With over 1.44 billion users, it is larger than China (1.39 billion) or India (1.25 billion). As the Facebook platform matured over time, the company added more and more features which are akin to living in a country. You have ability to communicate with your neighbors through all (digital) means – text, voice, video; attend sport events, watch movies (currently relying on link sharing) and voice out your opinions. The next step of maturing the platform was somewhat easy to predict, but hard
Taiwan IT industry, for a long time the leading among all Asian nations in terms of hardware manufacturing, has had its ups and downs over the decades. Does anyone still remember FIC (First International Computer), once the world’s leading mainboard maker? Or Elitegroup as the close second, while today’s leaders, Gigabyte, ASUS and MSI were far behind? Not many do… The gigantic Formosa Plastics consortium, which owned FIC, continued down the road with VIA Technologies and HTC, showing their love for three lettered acronyms. While these two still survive, the twisted sense of marketing and PR in these companies – one of the stones around their
While the world’s attention to electric powered planes is being focused around quite a unique project, Solar Impulse and its Round The World project, Airbus is quietly preparing to make electric dreams a reality. While Solar Impulse RTW project is now hitting delays as the weather conditions are not allowing for the plane to travel from China to Hawaii, words come from Airbus that they will build a final assembly line for their Airbus E-Fan aircraft in Pau, France. This puts the manufacturing plan in close proximity to Daher, a design firm known for its disruptive innovations in the fields of surveillance, aviation and aerospace.