Intel’s ‘Tick-Tock’ Seemingly Dead, Becomes ‘Process-Architecture-Optimization’by Ian Cutress on March 22, 2016 6:45 PM EST
As reported at The Motley Fool, Intel’s latest 10-K / annual report filing would seem to suggest that the ‘Tick-Tock’ strategy of introducing a new lithographic process note in one product cycle (a ‘tick’) and then an upgraded microarchitecture the next product cycle (a ‘tock’) is going to fall by the wayside for the next two lithographic nodes at a minimum, to be replaced with a three element cycle known as ‘Process-Architecture-Optimization’.
Intel’s Tick-Tock strategy has been the bedrock of their microprocessor dominance of the last decade. Throughout the tenure, every other year Intel would upgrade their fabrication plants to be able to produce processors with a smaller feature set, improving die area, power consumption, and slight optimizations of the microarchitecture, and in the years between the upgrades would launch a new set of processors based on a wholly new (sometimes paradigm shifting) microarchitecture for large performance upgrades. However, due to the difficulty of implementing a ‘tick’, the ever decreasing process node size and complexity therein, as reported previously with 14nm and the introduction of Kaby Lake, Intel’s latest filing would suggest that 10nm will follow a similar pattern as 14nm by introducing a third stage to the cadence.
From Intel's report: As part of our R&D efforts, we plan to introduce a new Intel Core microarchitecture for desktops, notebooks (including Ultrabook devices and 2 in 1 systems), and Intel Xeon processors on a regular cadence. We expect to lengthen the amount of time we will utilize our 14nm and our next generation 10nm process technologies, further optimizing our products and process technologies while meeting the yearly market cadence for product introductions.
While the new PAO or ‘Process-Architecture-Optimization’ model is a direct result of the complexity of developing and implementing new lithographic nodes (Intel has even entered into a new five-year agreement with ASML to develop new extreme-ultra-violet lithographic techniques), but also with new process nodes typically comes a time where yields have to become high enough to remain financially viable in the long term. It has been well documented that the complexity of Intel’s 14nm node using the latest generation FinFET technology took longer than expected to reach maturation point compared to 22nm. As a result, product launches were stretched out and within a three-year cycle Intel was starting to produce only two new generations of products.
Intel’s current fabs in Ireland, Arizona, and Oregon are currently producing wafers on the 14nm node, with Israel joining Arizona and Oregon on the 22nm node. Intel also has agreements in place for third-party companies (such as Rockchip) to manufacture Intel’s parts for certain regional activity. As well as looking forward to 10nm, Intel’s filing also states projects in the work to move from 300mm wafers to 450mm wafers, reducing cost, although does not put a time frame on it.
The manufacturing lead Intel has had over the past few years over rivals such as Samsung, TSMC and Global Foundries, has put them in a commanding position in both home computing and enterprise. One could argue that by elongating the next two process nodes, Intel might lose ground on their advantage, especially as other companies start hitting their stride. However, the research gap is still there - Intel introduced 14nm back in August 2014, and has since released parts upwards of 400mm2, whereas Samsung 14nm / TSMC 16nm had to wait until the launch of the iPhone to see 100mm2 parts on the shelves, with Global Foundries still to launch their 14nm parts into products. While this relates to density, both power and performance are still considered to be on Intel’s side, especially for larger dies.
Intel's Current Process Over Time
On the product side of things, Intel’s strategy of keeping the same microarchitecture for two generations allows its business customers to guarantee the lifetime of the halo platform, and maintain consistency with CPU sockets in both consumer and enterprise. Moving to a three stage cycle has thrown some uncertainty on this, depending on how much ‘optimization’ will go into the PAO stage: whether it will be microarchitectural, better voltage and thermal qualities, or if it will be graphics focused, or even if it will keep the same socket/chipset. This has a knock on effect with Intel’s motherboard partners, who have used the updated socket and chipset strategy every two generations as a spike in revenue.
EUV Lithography Makes Good Progress, Still Not Ready for Prime Time
Tick Tock on the Rocks: Intel Delays 10nm, adds 3rd Gen 14nm
The Intel Skylake Mobile and Desktop Launch with Microarchitecture Analysis
Source: Intel 10-K (via The Motley Fool)
Post Your CommentPlease log in or sign up to comment.
View All Comments
bernstein - Tuesday, March 22, 2016 - linkwell this basically stretches product cycles even longer... sandy bridge is still going strong... desktop as well as notebook.
nandnandnand - Tuesday, March 22, 2016 - linkThese process shrinks are more important for notebooks than desktops, because power consumption is a much bigger factor.
This is more good news for AMD. More time to show the world what Zen can do and possibly get Zen+ or a future chip near enough to Intel to beat it in more areas.
DigitalFreak - Tuesday, March 22, 2016 - linkAMD would need a miracle for that to happen. Even if they got one, Intel has more than enough resources to quickly turn the tide back in their favor. AMD only exists at this point because Intel would have antitrust issues to deal with if they didn't.
Frenetic Pony - Tuesday, March 22, 2016 - linkHuh? Resources have nothing to do with it. Intel's been shovelling money at Moore's law and still isn't ahead, and the current CPU architecture for silicon and modern apps is unifying towards a common optimization. For all sakes and appearances Zen looks like Core looks like Apple's architecture at low levels more and more and more.
We're getting into diminishing returns on all fronts as far as CPU architecture goes, and same goes with the advancement of Moore's law, which has benefited wide architecture like GPUs far more than CPUs as time has gone on. Until there's a major change, most likely from silicon to some other material, and both the process and architecture need to change dramatically, Intel's vast resources aren't actually going to do much for it except soak up the damage as profit margins drop.
Michael Bay - Wednesday, March 23, 2016 - linkIf Zen looks like Core down low, you can kiss that modern level of IPC goodbye. ^_^
Seriously though, of course designs will converge on something that works best, and here`s where Intel`s Big Money are a great help, since they can pursue multiple research avenues simultaneously in hopes of at least one winning the day. AMD doesn`t have such luxury.
ImSpartacus - Tuesday, March 22, 2016 - linkYeah, it's frustrating that amd is so pitiful, but I don't see them seriously their situation any time soon. Still, I look forward to seeing zen the next year or two if only because it'll be something new and potentially interesting.
BurntMyBacon - Wednesday, March 23, 2016 - link@DigitalFreak: "AMD would need a miracle for that to happen."
It is looking very much like AMD will have access to a 14nm node before Intel leaves it, so ...
@DigitalFreak: "Even if they got one, Intel has more than enough resources to quickly turn the tide back in their favor."
I don't think many people are suggesting that Intel won't reach the next node well before AMD has access to it. That said, wouldn't it be nice to see every other or every third product release be at node parity. Intel would still be a step or two ahead given time to optimize for the new process, but AMD has historically introduced new architectures and new process nodes concurrently, so they could close the gap a bit during these overlap periods. In any case, they would be in a much better position competitively, than they are currently.
Consider the 65nm Phenom (Agena) architecture. The initial release was underwhelming and didn't really compete with its 45nm Core2 (Penryn) competition. It was widely believed that the Phenom (Agena) was hindered by its small L3 cache size. When the 45nm Phenom II was released, one of the major upgrades was the L3 cache. Interestingly, the Phenom II (Deneb) was a worthy competitor to the Core2 (Penryn) with the Phenom II 955 and the Core2 Quad Q9550 posting results as close as any AMD / Intel comparison since the pre-Pentium days. The problem is, it was now competing against Core ixxx (Nahalem/Westmere). If AMD had access to 45nm in time to compete with Core2 (Penryn), Intel would likely have still been ahead given the lack of time to optimize other parts of the architecture. However, just getting an appropriate cache in would have allowed the Phenom to compete much better. They would have been able to put more money into developing the current architecture and they may not even have felt the need to make a highly risky and radically alteration of the architecture into the Bulldozer (and derivatives) line. We may then have avoided the whole Sandy Bridge vs Bulldozer comparison that resulted in little competition from the AMD side forcing no more than incremental upgrades from Sandy Bridge - Ivy Bridge - Haswell - Broadwell - Skylake. Yes there have been notable improvements, but the fact that there are so many people with an i5-2500K that still don't feel the need to upgrade should tell you something. Meanwhile, AMD's "High-end" Piledriver is still on the same process node as Westmere (32nm). Their more mobile oriented Excavator is only a halfnode ahead of that (28nm) and not even close to node competitive with Intel's 22nm tri-gate process, much less their 14nm tri-gate process.
webdoctors - Tuesday, March 22, 2016 - linkYa I'm still rocking my Sandbridge desktop at home and it still hasn't shown any signs of being out of date. DDR4, USB3.1, and probably some other minor things I'm missing out on, but I still have all the main things like PCIE 3.0 and virtualization for my VMs. Its really hard to convince anyone with a 5 year old desktop to upgrade right now...
DigitalFreak - Tuesday, March 22, 2016 - linkI'm in the same boat with my Ivy Bridge i7-3770k. It's more than enough for what I need, but the lack of newer features on the motherboard are an issue. Only 2 SATA 3 ports, etc. If I could run the 3770k in an Z170 motherboard, I'd do it in a heartbeat.
StevoLincolnite - Tuesday, March 22, 2016 - linkI have Sandy-Bridge-E. 3930K... Can still out-bench the 5930k once I throw overclocking into the mix. (The 3930K will overclock higher. 5ghz under water.)
And will push past the 5960X at stock...
Just don't see the point upgrading this 5 year old machine other than to save electricity, Intel has given me no reason to do so.
Ironically... Prices have skyrocketed since I built my 3930K. Which was about $400 AUD, today it's replacement the 5930K is $870 for the CPU's alone. Ouch.