AMD has confirmed it will be launching its next generation Zen CPU architecture in early 2017, during the company's Q3 2016 earnings call.
Chris Hemmelgarn from Barclays asked: "With Summit Ridge launching in Q1 of 2017, I guess how would you expect the channel to ramp that? Do you see it ramping pretty fully in the first couple of quarters of the year, or are you looking for more normal PC seasonality?", to which AMD CEO Lisa Su replied: "You know, I would expect that there will be a relatively good initial demand for Summit Ridge that may be you know, not quite at the seasonal patterns".
Su added: "From where we see, Summit Ridge is playing in a space in the high-end desktop that we currently aren't offering a product. So we believe we'll be competitive certainly with Core I5 as well as Core I7 and we will be launching in those areas".
Intel's next generation Kaby Lake architecture has been teased, but we haven't seen CPUs in the wild just yet - even though there are countless leaks on the new Core i7-7700K, and the latest one is the best.
The upcoming Core i7-7700K has been overclocked on a 100-series motherboard, with it hitting 6.7GHz thanks to LN2 cooling - and considering its stock 4.2/4.5GHz clocks for base and boost, respectively - this is the fastest Core i7 CPU from the Kaby Lake family.
HKEPC is behind the Core i7-7700K overclocked to 6.7GHz, where he used an ASRock Z170M OC Formula motherboard and GALAX HOF DDR4 RAM. The multiplier on the 7700K was upped to x67, while the bus speed wasn't touched. We don't know what voltage was used, but with HKEPC using LN2, we're sure it was pretty high.
Samsung is dealing with massive issues with the halt of sales of its Note 7 smartphones after a handful of them had battery-related issues, with the number meant to not just stop with the purportedly fixed Note 7, except it didn't. Now we're hearing about the future of SoCs from Samsung, as they've just announced mass production of their 10nm FinFET process.
Executive VP and Head of Foundry Business at Samsung Electronics, Jong Shik Yoon announced: "The industry's first mass production of 10nm FinFET technology demonstrates our leadership in advanced process technology. We will continue our efforts to innovate scaling technologies and provide differentiated total solutions to our customers".
How does 10nm benefit us? The new 10nm FinFET LPE process uses an advanced 3D transistor structure that includes new process technology and design compared to the previous node. With Samsung shrinking down to 10nm, we'll see a 30% increase in area efficiency, as well as a 27% higher performance ratio, and 40% lower power consumption. 10nm isn't easy, with the new node being one of the smallest in mass production, Samsung used "cutting edge techniques such as triple-patterning to allow bi-directional routing are also used to retain design and routing flexibility from prior nodes".
Samsung says that SoCs with the new 10nm process technology will be used in devices "launching early next year", adding that they're "expected to become more widely available throughout 2017".
Intel has something truly monstrous with its upcoming Knights Landing LGA 3647 socket, with Serve the Home teasing us with some pictures of the gigantic LGA 3647 socket.
The upcoming Knights Landing platform will support 6-channel DDR4 support, as well as no dual-latch system that is used on the LGA 2011 platform. Serve the Home explains that the socket and thermal designs that they have seen "do not have latching mechanisms", and instead they're seeing "solutions that secure the CPU to the heat sink".
Here we have a comparison shot of Broadwell-EP, LGA 3657 (Knights Landing), and Broadwell-DE.
STH had an Intel Xeon D (Broadwell-DE) BGA package next to the massive LGA 3674 socket, and it looks super small.
Update: It seems that the benchmarks results of the Core i7-7700K might have been overexaggerated, with WCCFTech noticing that the clock speeds seem to be "completely wrong". The performance gains of the new Kaby Lake-based Core i7-7700K are much harder to put down on paper - or, er - your LCD screen - because we don't know the exact clocks. Expect more news on the 7700K as it breaks.
Intel is very close to the release of its Kaby Lake architecture, arriving in the form of the Core i7-7700K, but first we have some leaked benchmarks that tease the improvements of Intel's new 14nm+ architecture.
The upcoming Core i7-7700K performance is shaping up well, with leaked Geekbench 4 results showing that Intel has managed to pull off some great single-threaded performance improvements over Skylake, but the multi-threaded performance isn't left out in the cold. Intel's upcoming Core i7-7700K arrives with a base clock of 4.2GHz, while boost clocks will have it driving up to around 4.5GHz. Intel has slapped a 95W TDP on the Core i7-7700K, and it still uses the LGA 1151 socket.
Performance wise, the new Kaby Lake-based Core i7-7700K had a single-core score of 6131 on Geekbench 4, while the multi-threaded score was 20,243 - this is a 42% increase over the current Skylake-based Core i7-6700K, which scores just 4300 points or so in the single-core test. The multi-core result is just as interesting, with a 20% increase in performance over the 6700K and its 16,756 score.
AMD has so many different great things going on at once, with the upcoming Zen-based Snowy Owl due for consumers early next year on the Zen architecture, but the upcoming Naples CPU is shaping up well and reportedly due for a launch in Q2 2017 for the enterprise market.
The upcoming 32-core processor is based on Zen CPU cores, with 32 of them pumping away, capable of executing two threads simultaneously - so we're talking about a 64-thread CPU for the enterprise market, based on Zen CPU cores. Beasty. Leaked benchmarks of AMD's upcoming monster CPU have arrived, with an engineering sample of Naples teasing some impressive performance.
AMD's upcoming Naples engineering CPU had a base clock of 1.4GHz, boosting up to 2.9GHz, while it features an absolute titanic amount of L3 cache. AMD is throwing 512MB of L3 cache, something that is a first for the CPU industry. But does this translate into monster performance? Well, benchmarks from a few weeks ago have been improved, with the Naples CPU performing 13% better in just a few weeks.
AMD is rounding out 2016 with some big teases of its next-gen APU architectures, with Zen ready to launch in 2017, we've been hearing rumbles about the next step in their APU dominance: Gray Hawk. Gray Hawk is expected to hit the 7nm node using the Zen+ architecture, and will be released on the AM4+ socket in 2019 - but we're now hearing about the APUs that will be released between now and then.
AMD will reportedly launch Snowy Owl early next year, with the APU being aimed at the server market, based on Zen CPU cores and on the AM4 socket. After that, Horned Owl will be released into the mainstream consumer market on the same 14nm FinFET node, but on the FP5 BGA / AM4 packaging, and will arrive in 2H 2017. Banded Kestrel (seriously, these are awesome codenames, AMD) will succeed Horned Owl, and will be released in the first half of 2018 for the mainstream consumer market, using the same Zen cores as Horned Owl and Snowy Owl.
Horned Owl and Banded Kestrel are part of the mainstream APU roadmap, being manufactured by GlobalFoundries and Samsung on the 14nm FinFET process, with the Horned Owl APU succeeding Raven Ridge. Banded Kestrel is a lower tier platform that will have half the specifications and power that Horned Owl will have, but both platforms will be aiming at the embedded market on the FP5 BGA socket.
AMD has been all over the next-gen CPU game, which seems to be looking to really take off with Zen launching in 2017, but the processors that will follow in the years after will be beyond impressive.
We're now hearing about CPUs that will arrive all the way out in 2019, with a 4-core/8-thread CPU on the 7nm FinFET process using Zen+ cores would use just 10W. The graphics side of these new codenamed Gray Hawk processors will be courtesy of the next-gen Navi GPU architecture due out in 2019, which was reportedly delayed from 2018 just recently.
Gray Hawk will be released after Raven Ridge, with Raven Ridge made on the 14nm process and due sometime in mid-2017. The new Gray Hawk architecture will be used in desktops, notebooks, arcade gaming systems, embedded devices, automation, retail signage to medical imaging and industrial control systems.
AMD is looking to have a gigantic year in 2019, with the release of 'Starship', which is a beast of a 48-core/96-thread CPU.
Intel's new Kaby Lake architecture is so close I can smell it, with the upcoming Core i7-7700K processor already up for pre-order on an Estonian online retailer website for what equates to $402.
The specs listed on the website state the Core i7-7700K is a quad-core processor with 8 threads at 4.2GHz, but it'll hit 4.5GHz under Turbo Boost, a higher frequency than previously reported. Intel's upcoming Core i7-7700K should rock 8MB of L3 cache, and a 95W TDP, while slotting into the LGA1151 socket.
Kaby Lake will work on some current Z170-based motherboards, while newer Z270-based boards should be launching in the near future. The new Kaby Lake architecture is made on Intel's latest 14nm+ node, which provides up to 12% more performance over the older Skylake architecture made on 14nm.
Intel will have a bunch of Kaby Lake-based CPUs, including: Core i7-7700, Core i7-7700T, Core i5-7600K, Core i5-7600, Core i5-7600T, Core i5-7500, Core i5-7500T, Core i5-7400, and the Core i5-7400T.
AMD's upcoming Zen CPUs and new AM4 socket are closer and closer to being revealed, and now we have leaked shots of the PGA (Pin Grid Array) design which looks to have 1331 pins - just 6 short of that awesome 1337 number, AMD.
The current AM3+ socket from AMD is running out of steam, with the upcoming AM4 socket delivering native PCIe 3.0 support as well as native USB 3.0 and USB 3.1 functionality. AM4 will support dual-channel DDR4 RAM with up to 4 x DDR4 DIMMs at 3200MHz. The new socket will support up to 24 PCIe lanes, depending on the chipset - bringing AMD to an equal playing field against Intel's mainstream LGA 1151 socket.
AM4, just like previous sockets from AMD, is a PGA socket with zero insertion force. The same locking mechanism is in play, with the same 40x40mm space used on AM4, just like AM3+ before it. AMD has changed the pin count up on the AM4 however, with 1331 pins over the 942 of AM3+ and even more than Intel's current LGA 1151 socket - a first for AMD.