As rumored earlier, AMD has now officially announced two new FX-Series CPUs based on the Piledriver CPU architecture, the FX-9590 and the FX-9370. Both announced CPUs are eight-core SKUs and a part of AMD’s Vishera series with 32nm Piledriver CPU architecture and AMD managed to bring the maximum Turbo clock speed of the FX-9590 to 5.0GHz.
According to AMD, these new CPUs, including the 5.0GHz FX-9590, are an emphatic performance statement aimed at those gamers seeking ultra-high resolutions wiht AMD Eyefinity technology. Both are unlocked for easier overclocking, feature AMD Turbo Core 3.0 technology and, although not officially confirmed, should have the same 220W TDP.
Unfortunately, AMD did not announce the precise specs but since they are pretty much the FX-8350 on steroids, we are looking at 4x2MB of L2 and 8MB of L3 cache and 1866MHz DDR3 memory support. AMD only announced the Turbo clocks for both parts set at 5.0GHz for the FX-9590 and 4.7GHz for the FX-9370 but if they share that much details with the FX-8350, you can expect the base clock to be set at 4.8GHz for the FX-9590 and 4.5GHz for the FX-9370.
AMD also did not shed any light on actual retail availability or the price of the new FX-Series CPUs as these will be initially available only for system integrators.
We will surely keep an eye out for them and hopefully we will see a benchmark or two soon.
ARM’s Cortex A50 series is arguably the firm’s most important design to date, as it uses the ARMv8 architecture that has 64-bit support. The firm announced that LG will be the lead partner for the A50 series of chips that includes the Cortex A57 and its next generation Mali GPUs.
According to ARM its agreement with LG will mean the Korean electronics giant will be able to offer a wider range of system on chip (SoC) parts and will allow LG to make use of the GPU computing capabilities that will be featured in ARM’s next generation Mali GPUs.
Bo-ik Sohn, SVP of LG Electronics said, “The close interaction between the CPU and GPU becomes increasingly important as solutions transition between the 32-bit and 64-bit world. Being able to implement Full Profile GPU Compute in conjunction with ARM Big Little processing configurations will be a key driver to the overall performance and capabilities of our devices.”
ARM’s Cortex A50 chips are expected to power tablets and servers initially, with 64-bit support helping chip designers address more than 4GB of RAM without having to use Physical Address Extensions (PAE). ARM’s licensees have been working on Cortex A50 based parts for some time, with the first parts expected to hit the market later this year.
Whether LG’s position as a lead partner for the Cortex A50 series chips means the firm plans to challenge traditional rival Samsung in the server business is not clear at this point, but it does signal that 64-bit consumer electronics can be expected from the firm in the future.
If you have any old AMD shares lying around you might like to sell them as fast as you can, according to the bean counters at Goldman Sachs.
Despite the fact that the company is doing rather well, and its share price is has gone up rapidly over recent months, Goldman Sach analysts claim that the writing is on the wall for AMD. It thinks that AMD shares will be worth just $2.50 soon. The stock’s 50-day moving average is currently $2.98.
The company said that while AMD could clean up in the gaming market even if you take those figures into account the stock is trading at 22 times its 2014 CY EPS estimate. In other words the company’s core PC business is still shagged and still will generate 45 per cent of the company’s 2013 revenue.
“We therefore believe this recent move in the stock is just the latest in a long history of unsustainable rallies, and we are downgrading the stock to Sell. We believe the current multiple is unjustified for any company with such significant exposure to the secularly declining PC market,” the firm’s analyst wrote.
Analysts at Sanford C. Bernstein think that the share price will settle on $2.00 and FBR Capital Markets thinks $3.00. In other words if you want to know what is really happening at AMD you might as well ask the cat, than any Wall Street expert.
ARM’s notable success in smartphones and tablets can obscure the fact that most of the chips using its designs are microcontrollers for using the input of sensors. The firm has announced collaboration with Logmein to push its Mbed project with developers that sign up to the Xively Cloud service.
ARM’s Mbed project aims to bring a standard workflow to hardware design in order to help more firms to make better use of the microcontroller technology that already exists. Simon Ford, director of Online Tools at ARM told The INQUIRER that the MBed project is intended to help hardware designers turn microcontrollers into final products.
Logmein and ARM worked on the Xively cloud based rapid prototyping service to offer hardware developers a way to speed up and lower the cost of the development lifecycle. Those developers who sign up for the service will also get a Xively Jumpstart Kit that includes an ARM Mbed prototype module to get started.
Ford said, “You’re trying to build a product, the intelligence you want embedded is critical but it isn’t the only problem you have. If you are trying to make a product, you have a whole raft a problems. [...] We are expanding the Mbed project to look at how do you have an industrial grade platform that is open, free to use and that removes barriers for someone that has this idea to proving a concept all the way to production.”
While ARM and Logmein promote the service as a way to build the much hyped internet of things, it can be used to develop any hardware that makes use of ARM’s extensive range of microcontrollers. With Logmein’s Xively cloud service, the firms are hoping to enable developers to cut the costs associated with hardware design, enabling smaller firms to get into the market.
AMD has said the memory architecture in its heterogeneous system architecture (HSA) will move management of CPU and GPU memory coherency from the developer’s hands down to the hardware.
While AMD has been churning out accelerated processing units (APUs) for the best part of two years now, the firm’s HSA is the technology that will really enable developers to make use of the GPU. The firm revealed some details of the memory architecture that will form one of the key parts of HSA and said that data coherency will be handled by the hardware rather than software developers.
AMD’s HSA chips, the first of which will be Kaveri, will allow both the CPU and GPU to access system memory directly. The firm said that this will eliminate the need to copy data to the GPU, an operation that adds significant latency and can wipe out any gains in performance from GPU parallel processing.
According to AMD, the memory architecture that it calls HUMA – heterogeneous unified memory access, a play on unified memory access – will handle concurrency between the CPU and GPU at the silicon level. AMD corporate fellow Phil Rogers said that developers should not have to worry about whether the CPU or GPU is accessing a particular memory address, and similarly he claimed that operating system vendors prefer that memory concurrency be handled at the silicon level.
Rogers also talked up the ability of the GPU to take page faults and that HUMA will allow GPUs to use memory pointers, in the same way that CPUs dereference pointers to access memory. He said that the CPU will be able to pass a memory pointer to the GPU, in the same way that a programmer may pass a pointer between threads running on a CPU.
AMD has said that its first HSA-compliant chip codenamed Kaveri will tip up later this year. While AMD’s decision to give GPUs access to DDR3 memory will mean lower bandwidth than GPGPU accelerators that make use of GDDR5 memory, the ability to address hundreds of gigabytes of RAM will interest a great many developers. AMD hopes that they will pick up the Kaveri chip to see just what is possible.
ARM posted market-beating first-quarter financial results, thanks to strong demand for its chip designs. The company forecast that annual revenue would be in-line with market expectations.
ARM’s first quarter revenue rose 28 per cent to $170.3 million from $132.5 million a year earlier. Analysts had expected a 20 per cent rise in revenue to $158.8 million. The company made an adjusted pretax profit rose 44 per cent to $89.4 million from $61.9 million a year earlier. Analysts were expecting a 25 per cent jump to $77.6 million.
Chief Executive Warren East said the company has “delivered another quarter of strong revenue and earnings growth, driven by robust licensing and record royalty revenue.” ARMs royalty revenues again outpaced the wider semiconductor industry, riven by market share gains in key end markets including digital TVs and microcontrollers, he said. ARM also continues to benefit from the growth in smartphones and tablets.
This year ARM said it had made an encouraging start with more leading companies choosing to sign up to ARM technology. More than 22 processor licenses were signed in the first quarter ended March 31 from smartphones, mobile computing, digital television and other technology.
In the quarter there were more than 2.6 billion ARM-based chips were shipped, up 35 per cent from a year earlier.
ARM announced the availability of POP IP for its upcoming ARMv8 based Cortex A53 and Cortex A58 chips using TSMC’s 28nm process node.
ARM’s POP IP products are designed to help smaller ARM vendors design and manufacture chips based on its various architectures using specific process nodes that ARM and the foundry, in this case TSMC, has worked on developing. ARM announced POP IP products for its upcoming 64bit ARMv8 architecture using TSMC’s 28nm high performance mobile process node and put the firm’s upcoming 16nm FinFET process node on its POP IP roadmap.
ARM stressed the ability for its Cortex A53 and Cortex A58 chips to be used independently or in its Big Little architecture. The firm said its POP IP products will enable dual-core and quad-core implementations of both chips.
John Heinlein, VP of marketing for the Physical IP Division at ARM said, “ARM is currently enabling its lead partners on Cortex A57 and Cortex A53 processor implementations with POP IP and physical IP platforms, and we are committed to support our partners on the leading process nodes.
“Only ARM can offer a complete roadmap of POP IP core-hardening acceleration solutions so deeply integrated and tightly aligned with ARM processor development activities now and into the future.”
ARM said its 16nm POP IP products for both Cortex A53 and Cortex A58 chips will be released in the fourth quarter of 2013, which gives an indication of when TSMC will have its next generation process node ready. TSMC’s various 28nm process nodes have been very popular and capacity for its 16nm FinFET process node is likely to prove equally popular with high-profile chip vendors like AMD, Nvidia and Qualcomm.
ARM has said that vendors are looking to standardize on both one chip architecture and a single operating system such as Linux across their product lines.
With many of ARM’s licensees preparing to make a big splash in the server market, the firm claimed its architecture is the only one that scales from smartphones all the way up to servers. Lakshmi Mandyam, ARM director of Server Systems and Ecosystems told The INQUIRER that the ability to stick with one chip vendor and run the same operating system throughout its product stack is something “people find very interesting”.
ARM expects most of its servers to end up powering open source software stacks, which strongly suggests Linux as the underlying operating system, especially since FreeBSD’s ARM port is seemingly in a state of flux. According to Mandyam, companies are increasingly interested in concentrating on a single processor and software stack, saying that vertical integration – the term given to keeping everything in-house – is once again becoming fashionable.
Mandyam said, “In the industry there is a trend back towards vertical integration where if you look [back] 20 years [they] did their own ASIC, they did everything themselves and then after that it migrated to outsourced [then] to merchant silicon, but I think people are seeing a benefit for integrating again.”
Mandyam used Chinese consumer and enterprise IT vendor Huawei as an example of how ARM chips are deployed from its smartphones through to its network infrastructure equipment. She said, “Huawei uses the ARM architecture for handsets, set-top boxes, routers, base stations and they have talked publicly about their server intentions.”
According to Mandyam, ARM is the only vendor that can allow companies such as Huawei to stick to one architecture from smartphones to servers.
“Definitely [customers] see the advantages of standardising across one architecture from low to high and ARM’s the only architecture that can scale down to the low, and allow people to do integration. The integration curve is the other curve ARM can go down and the other guys can’t.”
However Mandyam said the advantages of ARM weren’t limited to just being able to deploy a single vendor’s chips but also include the ability to deploy Linux throughout the product range. She said firms are working with Linaro’s Enterprise Working Group in order to have some influence on the direction of enterprise Linux deployment because they see it as a key component to a cost-effective server.
Mandyam said, “If you look at who is participating in Linaro’s enterprise group, you have end users like Facebook that are participating as well, because they see the value of getting involved early and it really shortens the time to market. If you think about the Linux kernel, it’s all standardised – based on the ARM architecture. Everyone, instead of trying to reinvent the wheel at different times, you have everyone cooperating.”
Despite Linaro’s Enterprise Working Group containing many firms that compete against each other for business, Mandyam claims that Linux still allows them to “maintain differentiation” that is needed to get a competitive advantage.
The combination of ARM’s architecture scaling from embedded devices to servers combined with a similar capability for Linux could be the combination that allows firms to cut costs and ultimately increase profits.
GlobalFoundries has announced that it demonstrated chip stacking using through-silicon vias (TSV) on its 20nm process node.
Spun off from AMD and perhaps best known for fabbing AMD’s processors, Globalfoundries has been investigating TSVs as a method of stacking chips, much like rival chip foundry TSMC. Now the firm just announced that it has demonstrated working 20nm silicon that incorporates TSVs.
According to Globalfoundries, it inserts the TSVs after the Front End of the Line flow and before the Back End of the Line flow, meaning it can use copper for the TSV fill material. The firm claims it developed its own contact protection scheme in order to make TSVs possible during its transition from 28nm to 20nm.
AMD has already said that it is waiting for the move to 20nm, presumably with Globalfoundries, however it is likely that fabbing TSVs at 20nm will take longer. At the GPU Technology Conference Nvidia announced that it will use TSVs to stack DRAM on its upcoming Volta GPUs.
David McCann, VP of packaging R&D at Globalfoundries said, “Our industry has been talking about the promise of 3D chip stacking for years, but this development is another sign that the promise will soon be a reality.
“Our next step is to leverage Fab 8′s advanced TSV capabilities in conjunction with our OSAT [outsourced semiconductor assembly and test] partners to assemble and qualify 3D test vehicles for our open supply chain model, providing customers with the flexibility to choose their preferred back-end supply chain.”
Globalfoundries didn’t say when its 20nm TSV process will be ready for volume production, but the firm’s announcement that it has working silicon might be tempting for potential customers.
ARM has been pushing its Mali GPU architecture both as a graphics engine and as a GPGPU accelerator through its support of OpenCL. The firm said that not only could chip vendors see improved performance by offloading onto the GPU but they could also cut costs by moving tasks from dedicated silicon to the GPU.
Ian Smythe, director of marketing at ARM also said that firms could move tasks such as image stabilization over to the GPU, which is already on the system on chip (SoC), and cut back on other parts of the chip.
Smythe said, “You might be able to save some cost somewhere in the SoC by cutting out a bit of hardware that you had and run it on the GPU instead. So cost reduction and an improved capability. So maybe they will cut out some of the ISP and they will do it on the GPU because the silicon is already there, it’s power efficient, it’s a quicker way of doing [it], you get a cost reduction, and performance goes up.”
Aside from lower cost chips, Smythe’s comments suggest that chip vendors and in particular vendors such as Samsung, which also develops software for its Android smartphones, will make use of GPGPUs as a way of cutting bill of materials costs and reducing chip complexity.
Smythe added that GPGPU computing doesn’t have a single ‘killer application’ but rather could be applied to use cases such as computational photography, as seen in the HTC One and gaming. However given Smythe’s comments, perhaps the real killer application for hardware vendors will come from lower material costs.
Richland is set to replace AMD’s Virgo platform, powered by Trinity processors, and this change will happen in June 2013, most likely coinciding with Computex 2013.
AMD has just launched the first batch of Richland mobile APUs and we still have to see some notebook designs hitting the market. We wrote about mobile Richland APUs.
As of late last year Desktop Richland was always set to launch in June 2013 and the fastest of them is the A10 6800K, clocked at 4.1GHz and 4.4 with Turbo. It also features Radeon HD 8670D graphics that run at 844 MHz. This is the fastest Richland part and it comes unlocked, ready to replace the current AMD A10 5800K. In Europe, the A10 5800K currently sells for 112, while in US the same CPU sells for $129.00 (boxed).
The alpha dog A10 6800K is followed by A10 6700, A8 6600K (Unlocked) and A8 6500. AMD has a mix of 100W and 65W quad-core Richland desktop SKUs. There will be a single A6 6400K (Unlocked) SKU and the A4 6300, both dual-cores with 65W TDP.
Production ready samples were churned out in late January, while volume production is scheduled for late March 2013. The announcement was always scheduled for June 2013 and Richland last through most of 2013, until Kaveri with 28nm Steamroller comes on line.
There have been more than enough leaks dealing with Richland, AMD’s successor to the Trinity powered Virgo platform, and we even had a chance to see some leaks regarding its successor, codenamed Kaveri. As you may already know, Richland is planned to last through 2013 and it is clear that this is very important chip for AMD.
Based on the Piledriver architecture and built using 32nm technology, Richland will feature an integrated GPU that will be upgraded to Radeon HD 8000 series, a generation ahead of Trinity. As you know, there has been a lot of leaks regarding the Richland parts and the quad-core A10-6800K with Radeon HD 8670D graphics is expected to pack quite a punch. Best of all, Richland will still use the same FM2 socket.
According to our sources, the NDA will be lifted on 12th of March, 8am EST, and we are sure that we will see at least a couple of reviews as well as some additional info regarding the price and the availability date.
Following the recent tease, AMD has detailed its new TressFX technology aimed to create much more realistic hair in games. The new technology will debut in Tomb Raider 2013, making the hair of Lara Croft impressively more realistic.
According to the AMD blog that explains TressFX, it uses DirectCompute to unlock the processing capabilities of GCN architecture and is based on AMD’s previous work on Order Independent Transparency (OIT), a method that makes use of Per-Pixel Linked-List (PPLL) data structures to manage rendering complexity and memory usage. The new TressFX for Hair has been developed in collaboration between AMD and Crystal Dynamic in order to bring quite an improvement to hair rendering and physics.
The blog post additionaly explains that DirectCompute is utilized to perform real-time physics simulations of TressFX Hair and treats each strand of hair as a chain with dozens of links thus allowing realistic influence of elements like gravity, wind and movement to Lara’s hair. Each strand is also given a collision detection which ensures that strands will not pass through each other or any other solid surface.
According to what we can see from the blog post, AMD recommends the GCN-based AMD Radeon HD 7000 series as a particularly well equipped graphics card series for this type of task but did not exclude other GPUs either.
In any case, the effects on the hair certainly look impressive and we surely look forward to see what else can AMD pull off with this same TressFX technology and how open-source will it actually be.
AMD has released its Firepro R5000 graphics card that has video over IP capabilities.
AMD typically promotes its workstation class Firepro cards using CAD/CAM software, however this time the company is relying on remote viewing as the big selling point for its latest workstation graphics card. AMD’s Firepro R5000 has a GPU that uses its Graphics Core Next (GCN) architecture and Teradici PC video over IP technology to send graphics output over the network.
AMD used its Pitcarin GPU coupled to 2GB of GDDR5 memory in the Firepro R5000. However it isn’t AMD’s GPU that is the big selling point of the Firepro R5000 but rather Teradici’s Tera2240 chip that encrypts display output before sending it out on the network, while supporting up to 60fps (frames per second).
AMD’s Firepro R5000 is intended to be used in render farms, with each final image being sent over an IP network to the end host, and the firm claims that the technology can be used in education, financial and media environments.
The Firepro R5000 is a single slot graphics card that has two mini Displayport outputs that can drive two 2500×1600 displays, however it can also drive a further four remote displays at 1920×1200 resolution by sending data over its RJ45 Ethernet port.
Both AMD and Teradici talked up the low configuration overheads of the Firepro R5000.
Although it was detailed back in August last year, AMD has just now officially released its new “affordable” Vishera based FX-4130 quad-core socket AM3+ CPU.
The new CPU is part of AMD’s 4100-series and is based on Vishera core design with four Piledriver cores. It works at 3.8GHz base clock and can “turbo” up to 3.9GHz. It packs 4MB of L2 and 4MB of L3 cache and has a 125W TDP.
According to the slide over at Xbitlabs.com, the FX-4130 replaces the FX-4100 with the same US $101 price but should provide between 3 and 9 percent more performance.
As things get better with Globalfoundries and their 32nm process technology, AMD is expected to introduce new models based on cut-down versions of Vishera, according to the report.