Processors - Page 106

It looks like Samsung is continuing to out-innovate everyone else, announcing that Samsung and even TSMC are producing chips using the new FoWLP yield technology. What the hell is FoWLP yield technology you ask?

FoWLP, or Fan-Out Wafter Level Packaging platform, allows smartphone OEMs to reduce the thickness of their devices by a great degree, which has multiple benefits. FoWLP also doesn't need a PCB, and also increases the efficiency of chips by 30%, as well as reducing the thickness of smartphones by at least 0.3mm.

Apple is reportedly using TSMC as its main supplier of 10nm FinFET processors, and Samsung wants a cut of that business - which is where FoWLP comes into play. Samsung might end up winking just right at Apple, especially if TSMC can't keep up with Apple's high demand. With smartphones hitting the whole no bezel, super-thin, more battery wishlist - Samsung's impressive new FoWLP technology is going to come in handy big time over the coming years.

We've been hearing so much about the Zen architecture from AMD, but this new information is rather exciting. Our friends at Fudzilla have exclusive information on an upcoming Zen CPU with 32 cores, and 64 threads, with a huge 64MB of L3 cache.

The chip has a codename of Naples, powered by the Zen architecture, with each CPU core featuring its own dedicated 512KB cache. A cluster of Zen cores share 8MB of L3 cache, pumping the total L3 shared cache of 64MB. There will be a 16-core variant as well, with both of the CPUs being made on the 14nm FinFET process. Naples has 8 independent memory channels, with up to 128 PCIe 3.0 lanes. This makes AMD ready for the super-fast NVMO memory controllers and drives, with Naples set to support 32 SATA and NVMe drives.

Naples also has support for a huge 16 x 10GbE network, with the controller integrated, with Naples being slotted into a SP3 LGA server socket. We should expect the server-bound Zen products to have 35W to 180W TDPs, with dual, quad, 16, and 32-core server variants of Zen, with a release window of late 2016, or sometime in 2017.

Computex 2016 - Intel exec Navin Shenoy confirmed that Intel's 14nm 7th-gen Kaby Lake Core CPU will enter production later this quarter, with an eventual release in 2016.

"I'm excited to tell you that two new products will be coming from Intel later this year: Apollo Lake for the value and entry-level PC, 2-in-1 and tablet, and the 7th Generation Core, formally codenamed Kaby Lake, will be coming later this year. We have over 400 designs coming to market on the Seventh Gen Core, and you can expect lots of innovations from our OEM partners bringing this product to market," said Intel's General Manager of Intel's Client Computing Group Navin Shenoy said during Intel's Computex 2016 keynote.

With the seventh-generation Kaby Lake chip series, Intel has broken its tick-tock cadence for the first time in the last decade, opting instead for the new "process, architecture, optimization" model. Kaby Lake--and the new low-power, entry-level Apollo Lake--will use the same 14nm process node as Skylake and Broadwell-E, but will feature significant microarchitecture tweaks and optimizations. Intel also announced Kaby Lake will natively support USB 3.1 and Thunderbolt 3, with higher-end chips maxing out at 95W TDP.

Computex 2016 - After revealing the $199 VR-ready Radeon RX 480 and the Summit Ridge Zen silicon at Computex, AMD revealed its latest 7th-generation Bristol Ridge and Stony Ridge APUs with massive improvements in CPU and GPU performance and power efficiency.

AMD's new 7th-generation APU lineup is a medley of processors separated in performance tiers across two product families: Bristol Ridge for higher-end gaming and performance, and Stony Ridge for everyday and entry-level use. The APUs range from the lower-end 2.8GHz A6-9210 to the performance-grade 3.7GHz 7th-gen FX 9830P. AMD boasts that Bristol Ridge's premiere FX chips are 56% faster than previous gens, with the entry-level A6 series 51% faster than past generations. The Bristol Ridge APUs are also beating Intel's mobile chips: According to AMD, the new FX-9800P is 50% faster than Intel's i7 6500U processor.

The lineup puts emphasis power efficiency and performance, which is exactly what users need in laptops. AMD's new 7th generation APUs leverage four Excavator cores and eight Radeon Graphics Core Next Cores, with integrated GPUs ranging from Radeon R2 to Radeon R7 graphics. The A10, A12, and FX series support performance gaming features like FreeSync and pairing up with discrete desktop video cards with Dual Graphics, whereas EyeFinity and Framerate Target Control are reserved for the top two models. The FX series also sports DirectX 12 and Vulkan with multi-adapter support.

Computex 2016 - AMD's Computex press conference saw Team Read announce its new roadmap built on Polaris FinFET technology, culminating in the new $199 VR-ready Radeon RX 480, the 7th generation APU line, and a beefy update on Zen.

It's been a bit since AMD's last Zen update, but the next-gen Summit Ridge processor was revealed on stage by Dr. Lisa Su, along with a host of new specifications and details on Zen's different configurations. "We are in the early stages, but the product looks really really good. Zen is delivering 40% more IPC than our previous generations," said AMD President and CEO Dr. Lisa Su. "This product is 8 cores, 16 threads, it is FinFET technology, and it is integrated as part of our new AM4 desktop platform. Zen is a new high-performance CPU that scales across multiple market segments, and we're working very hard on our server version of Zen, which is also going very well."

Dr. Su also discussed Zen's integrated future across multiple platforms, including high-performance APUs and embedded markets: "As we bring the high-performance CPU to life, we're also going to integrate Zen with our high-performance graphics in our next-generation APUs. So after Bristol Ridge, you will see integrated APUs with Zen and our GPU architectures. And you'll also see Zen across a number of embedded markets. The power of Zen's grounds-up design allows us to scale it across performance segments as well as low-power markets."

Computex 2016 - Intel has announced its new 10-core 3GHz i7 6950X desktop CPU at its press conference in Taipei, but the chip comes with a staggering price tag.

If you've been looking forward to Intel's celebrated Broadwell-E lineup with its insane 10-core Core i7 6950X CPU, then we have kind of a buzzkill: the processor will cost $1,732. For that premium price, you'll get one of the "most powerful chips" has ever created. With 20 threads, the Core i7 6950X Extreme Edition is fully optimized for simultaneous compute-intensive mega-tasking like 4K gaming while rendering video in the background, and Intel boasts that the i7 6950X renders 3D twice as fast as the Intel Core i7 6700K.

As we expected, Intel also announced three more CPUs in the Broadwell-E Extreme Edition lineup: the 8-core Intel Core i7 6900K ($1,089), the 6-core Core i7 6850K ($617) and Core i7-6800K ($434). Every CPU in the Extreme Edition lineup is unlocked for easy user overclocking, and the processors are all based on Intel's new 14nm Broadwell-E silicon. Check below for full specs on each CPU as well as key features.

AMD is ramping up for quite a large Computex 2016, with the big reveal of the next-gen Polaris-based video cards, and then the exciting new Zen architecture on the CPU side of things. Well, it looks like the new Zen APUs will be exciting, with a purported 50% increase in performance than the previous Kaveri-based APUs.

AMD's new processors will arrive on the AM4 socket, with the flagship SKU offering 12 compute cores, with 4 of them Compute-based, while the rest will reside on the GPU side. The flagship APU will have 4 x CPU cores at 3.7GHz, with Turbo boosting it to 4.2GHz, with 2MB of L2 cache. On the GPU side, we'll have 8 x GCN 3.0-based compute units, with 512 streaming processors and a TDP of 65W. According to the latest leaks, AMD will have a low-power version that consumes 35W with a default clock of 2.9GHz and Turbo boost of up to 3.8GHz.

The performance tier SKU will be kicked down to 10 compute cores, with 4 dedicated to the compute side, and 6 to the graphics side. These chips will arrive in two different variants, with the highest APU hitting 3.5GHz and 3.8GHz on boost, while the 35W version will have a clock speed of 2.8GHz and boost of 3.5GHz. The budget ones will have 6 compute cores, at up to 3.5GHz and 3.8GHz on turbo, with 1MB of L2 cache. The graphics side has 4 compute units, and 256 GCN 3.0-based stream processors.

AMD will reportedly keep its next-gen Zen CPUs at the higher-end of the scale when they launch later this year, with 6- and 8-core variants being the status quo.

The dual- and quad-core variants of Zen won't launch at first, but AMD is expected to launch Bristol Ridge-based dual- and quad-core processors right away. AMD will reportedly be using 8-core dies in the fabrication process of Zen-based processors because it's the best way to get the most out of the production of its Summit Ridge processors.

This means AMD can not only provide a very powerful processor, but it'll cost the company less to only make a higher-end version at first, rather than multiple SKUs. This all falls onto the shoulders of the yields, which if they're not good - we might see dual- and quad-core versions of Zen quicker than normal. If the yields are good, AMD could finally find itself in a better position to fight Intel's unstoppable CPU dominance.

AMD's Wraith Cooler has been an unbridled success for them, and a way to show that the stock cooling system that comes with a boxed CPU doesn't have to be an afterthought. The Wraith cooler, however, was only available for certain processors despite it being a very good cooler. They've listened to the feedback from reviewers and gamers alike and have thoroughly expanded their in-house stock coolers to encompass more of their processor line.

The Wraith cooler itself is now included with the FX 8350 and the FX 6350 processors. Those aren't exactly the hottest processors at stock, though they do overclock well if they have the right solution connected to them. The Wraith cooler, though not the ultimate device, can still allow for raising the multiplier and the voltage by a bit for a modest overclock without melting your system or the CPU. Actually, it's pretty silent and does a great job of controlling thermals at stock and slightly beyond.

They've also developed other cooling solutions to include with their lower-end CPU and APU's as well. Those solutions aren't quite as spiffy as the Wraith, though they're far more efficient than what was included before. More fins and a better-designed fan mean lower temperatures at far more acceptable sound levels.

Intel's next-gen CPUs are right around the corner, with NCIX listing the new Core i7-6950X with a pre-order price of a wallet-busting $2349.99. The Core i7-6900K has a pre-order price of $1495, while the Core i7-6850K is $889, and the Core i7-6800K at $629.

These prices are not official and are inflated for pre-orders, with the normal pricing set to be somewhere around $1609, $1024, $602, and $422 for the Core i7-6950X, Core i7-6900K, Core i7-6850K and Core i7-6800K respectively. Why should we be excited about these new CPUs from Intel? Well, the new Broadwell-E processors will feature four variants, with the flagship being the 10-core/20-thread Core i7-6950X. The one below it will be the 8-core CPU found in the Core i7-6900K, while the other two CPUs will be 6-core variants.

Intel is expected to launch its new Broadwell-E family of processors in the coming months, with a big show at Computex 2016 in June.

Intel's Broadwell-E processors have already been listed on NCIX, available for pre-order, just ahead of the official release. The prices that are shown, however, are actually somewhat higher than what we think the actual MSRP to be once they hit the streets. If you're an early adopter, then you can now be among some of the first to receive your CPU's if you pre-order now.

The server-oriented Broadwell-EP parts have been released in full already and have shown incremental, though significant, increases in performance over their Haswell-EP counterparts. The real gain is power-efficiency, however, where the smaller node and accumulative changes have led to far better performance/watt numbers.

The new processors will be compatible with the 2011 v3 socket physically and also be compatible with Intel's X99 motherboard chipset, though the latter will require a BIOS update to be able to support the new microcode. That makes it an easy upgrade, though, for those looking for either power savings or wanting to continue to have the absolute fastest processor available. All four new processors were spotted, and their prices are after the break. The top-end 10 core i7-6950X is listed at $2349.98.

Intel has officially unveiled its new Apollo Lake platform, which is the next-generation family of Atom-based notebook SoCs. Apollo Lake uses a new x86 microarchitecture, as well as a new generation graphics core that will provide more performance.

The new Apollo Lake family is set to take on the affordable all-in-ones, mini PCs, hybrid devices, notebooks and even tablet PCs when it launches in the second half of this year. Apollo Lake is built on the new Atom-based x86 microarchitecture known as Goldmont, with a new graphics core that includes Intel's impressive 9th-generation architecture that is found in the current Skylake processors.

The new Apollo Lake family includes support for dual-channel DDR4, DD3L, and LPDDR3/4 memory which allows PC makers to make devices with all sorts of form factors, thanks to the impressive memory support. We also have support for the usual SATA drives, PCIe x4 drives and eMMC 5.0 - USB Type-C support is also included, as well as various wireless technologies.

Today isn't just about NVIDIA and their GTC event, AMD too has something new going on today as well. The 7th generation of the APU, based on the Excavator core and codenamed Bristol Ridge, are shipping far ahead of schedule. They're already shipping the new 28nm APU's in an updated HP Envy x360 2-in-1 laptop.

We were able to take a look at some Geekbench scores that inadvertently showed up yesterday in their online database and those seem to be just in time to announce the launch of their new processors that'll take the place of Kaveri based systems, and be along side of Carrizo based systems that also have the same Excavator cores.

Full specifications haven't quite been given yet, but we do know that the various 28nm processors will come in two and four core configurations with either Radeon R5 or R7 class GPU's with 8-10 compute units attached and support for DDR4. They're saying that compared to Kaveri, we should see a 50% increase in general performance. Gaming performance itself should definitely experience a nice boost. These processors are a stopgap until we see the Zen-based Summit Ridge APU's and CPU's later on. We're expecting more detailed information to come during Computex 2016, with the full break-down of clock speeds and all the other innovations that come along with it.

AMD's next generation Zen is still quite a ways off from being a reality in desktops or mobile products, but it look like their more mainstream series of mobile products based off of their 28nm Excavator core, codenamed Bristol Ridge, has been spotted in Geekbenches database. The results seem to point towards a slight improvement in performance, given the reported CPU frequencies.

The sample is the AMD FX 9800P installed in a Lenovo device of some kind. This particular chip has been rumored to have 4 28nm Excavator based cores that run at a nominal 2.7GHz with 8 GCN 1.2 compute units attached. Here, though, it's recorded as running at 1.85GHz and has a multi-cor score of 5596, which is on-par with an Intel i5-6200U running at 2.4GHz. That's not too shabby if the reported clock frequencies are correct.

You can also find scores for the A10-9600P and the A12-9700P in what could possibly be upcoming Lenovo laptops or test systems on the Geekbench website. Bristol Ridge will be the mainstream Excavator based APU's available both on the new AM4 socket as well as on mobile through the FP4 interface. There's no real difference between this and Carrizo aside from naming scheme and the switch to allow DDR4 instead of strictly DDR3. Otherwise, the corresponding A[x]-8xxx Carrizo SoC is virtually the same. The power-savings from moving to 28nm should prove advantageous, however. This will be a sort of stop-gap between Zen, which is still due sometime this summer, or so we hope.

Intel has teased its next-gen Core i7-6950X processor on its own website, with the entry for the "Intel Core i7-6950X Processor Extreme Edition" listed on the latest Management Engine software on its Support website.

If Intel hasn't stuffed something up, the Core i7-6950X should feature 10 CPU cores (20 threads thanks to HT), with 25MB of L3 cache and up to 3.5GHz for the clock speed. The other CPUs teased are the Core i7-6800 and the Core i7-6900 which are six- and eight-core CPUs, respectively.

These new CPUs are based on Intel's 14nm Broadwell-E silicon, but the new LGA 2011 v3 processors are set to be compatible with Intel's current X99 Express chipset motherboards, with future BIOS upgrades. We should hopefully hear more about Intel's next-gen CPUs soon, before they start shipping in the coming months.

Intel has just unleashed its new Broadwell-EP family of processors, starting with the huge Xeon E5-2600 V4 which features a huge 22 CPU cores. Thanks to Hyper-Threading technology, we have a total of 44 threads of CPU power, which is simply insane for the prosumer market - especially those who work in video editing.

The new Intel Xeon E5-2600 V4 hasn't been completely detailed by the company, but the enthusiast part will be the Xeon E5-2699 V4 which packs a base clock speed of 2.2GHz, 55MB of cache, and a pretty tame 145W TDP. What will this 44-threaded processor set you back? A hefty $4115, which works out to $187 per CPU core. If we consider the 8-core/16-threaded Core i7-5960X costs $1059 (which works out to $132 per CPU core) then the new Xeon E5-2699 V4 isn't too badly priced at all.

The new Broadwell-EP powered Xeon processors can take DDR4-2400, with up to 12 DIMMS per CPU socket. If you're using registered DIMMs, you can cram in up to 385GB of RAM per CPU, using 32GB DIMMs of DDR4. To put it simply: I want one, well - probably two.

IBM's Watson and other AI systems like it are very impressive showcases of the kind of learning that a well developed deep neural network is capable of. Even Tay, the rogue Microsoft millennial AI that favors the PS4 over the Xbox One and seems to dismiss the Holocaust, is a feat of software engineering and learning that's pretty fantastic. But compared to the human mind, it still takes these machines, which rely on GPU's CPU's and at times even specialized ASICS to process such enormous amounts of data in parallel, far longer to learn even simple tasks. And it can be energy intensive, far more so than the human brain. But IBM thinks, and knows, that there's a better way.

IBM and the crew at the T.J. Watson Research Center want to use a specialized processor called the resistive processing unit, which is a marriage of CPU with non-volatile memory, that could exponentially speed up machine learning. It does this, essentially, by allowing the different parts to communicate at rate that's at least 27x faster than a traditional DNN setup. Learning involves moving forward and backward, analyzing data that's stored in memory, making that a bottleneck in this application. It could then massively increase the ability of these networks to learn, making speech recognition and similar AI functions in what could almost be near-realtime.

This type of processor is only theoretical at the moment though solving this obstacle in even an incremental fashion could bring about a sizable speed increase. The researchers even mention the ability to see an advantage of up to 30,000 times should they design and implement a device made specifically for their own DNN software. "We propose and analyze a concept of Resistive Processing Unit (RPU) devices that can simultaneously store and process weights and are potentially scalable to billions of nodes with foundry CMOS technologies. Our estimates indicate that acceleration factors close to 30,000 are achievable on a single chip with realistic power and area constraints,"

It's an end of an era: Intel has confirmed through their latest K-10 filing that its infamous 'tick-tock' process development cycle is dead.

Instead of having two processor families on each die shrink, Intel will be using three or more over the coming years. The K-10 filing states that Intel will "expect to lengthen the amount of time we will utilize our 14 [nanometer] and our next-generation 10 [nanometer] process technologies".

Intel will continue to release new products each year, but there will be a tighter control over architecture optimization, as the development of process technology slows. So... what does this mean? It confirms that Intel's next-gen 'Kaby Lake' platform will be made on 14nm. It also confirms that the release window for 10nm from Intel will be 2017 at the earliest, and 7nm - well, that's 2019-2020 or beyond now.

One of the fastest APUs from the Bristol Ridge family will be just as fast as the Xbox One, according to a new rumor from Bitsnchips.

AMD's new Bristol Ridge family will feature a powerful APU that will be quite powerful, easily taking on the consoles in providing a 1080p gaming experience, in a small package and price. AMD is expected to launch its new Bristol Ridge family at Computex, so we should expect more details in June.

As for the rumor, the flagship Bristol Ridge-based APU would feature 16 compute units that are based on the GCN 1.3 architecture. The 16 compute units would include 1024 stream processors, which is the same SP count as AMD's Radeon HD 7850. The HD 7850 launched in 2012, and was a great budget/mid-range GPU - if we see this performance in an APU, things could get very exciting for AMD.

Marvell just expanded their line of ARMADA SoC ecosystem, that are frequently used in NAS and other networking devices, to include native support for open-source software platforms like OpenWRT and openSUSE.

Before now Marvell didn't officially support any other software than that which was initially installed on their platforms. Adding support in the kernel of the various open-source OS's required a lot of time from volunteers to make it work properly. Because of that, support was always a bit precarious, and it could take quite awhile for new devices to be added to the compatibility lists.

Now, however, their 64-bit ARMv8 powered ARMADA 3700 Cortex-A53 device family and ARMADA 7K and ARMADA 8K Cortex-A72 device families are getting full-fledged support for the Linux kernel as well as U-Boot support. That means that it'll be compatible with a much wider range of OS's, anything that has ARM support baked in can run on their chips, essentially.