Intel's reportedly planning to skip 10nm on desktop systems

Intel will use 14nm for desktop chips until 7nm is ready?

Intel accelerates 7nm Plans, 10nm will be a short lived node

Intel's reportedly planning to skip 10nm on desktop systems

Intel has been using its 14nm manufacturing process since the release of Broadwell in 2014. Since then Intel has refined their core architecture and manufacturing techniques to deliver Skylake, Kaby Lake, Coffee Lake, Coffee Lake Refresh and now, Comet Lake on 14nm. That's six generations of products on a single process node, an almost unheard-of feat. 

Now it has been reported that Intel's 10nm woes aren't over. Rumours are emerging that Intel has eliminated 10nm desktop processors from their product roadmap, leaving 14nm products on the table until 2022. 

So far, there are no signs of 10nm Ice Lake-S and Tiger Lake-S processors for desktop products. If the performance of today's 10nm mobile products is of any indication, it appears as if 10nm is suffering from performance issues. Intel's Ice Lake products offer lower core clock speeds than their 14nm Comet Lake counterparts, negating Ice Lake's IPC advantages and giving end-users minimal performance gains in standard desktop workloads. If the same holds true for desktop, it means that Intel may be unable to best the single-threaded performance of its 5GHz Skylake/Coffee Lake series processors on 10nm, at least in all workloads. 

HardwareLuxx has claimed that a reliable source has told them that Intel would replace 14nm processors with 7nm Meteor Lake CPUs in 2022. This makes sense given Intel's current process node roadmap, which has 7nm arriving in 2021, supposedly starting with graphics chips. 

Right now, Intel seems to be preparing the release of their Comet Lake-S series processors for early 2020, offering Skylake/Coffee Lake esque processors with up to 10 cores and 20 threads. These processors will reportedly be succeeded by Rocket Lake, which is rumoured to ship with 14nm CPU cores and a 10nm graphics chip. After Rocket Lake will come 7nm Meteor Lake chips from Intel, which will reportedly arrive in 2022.   

 Intel accelerates 7nm Plans, 10nm will be a short lived node

When moving to 7nm, Intel's slides indicate that they are aiming for 2x scaling in silicon density over 10nm and plan to use EUV (Extreme Ultraviolet) technology to accomplish this. Samsung and TSMC are already utilising EUV technology in their enhanced 7nm process nodes, though deeper utilisation is planned for their next-generation 5nm technologies. 

What's more significant is that Intel plans to enact a 4x reduction in design rules for their 7nm lithography node, which will make it much easier for the company to execute their designs on the platform. Given the complexities involved in making cutting-edge silicon, the impact of this change will make product design much easier for the engineers involved, and will likely allow Intel to be more creative with their products. 

If Intel's 10nm desktop processors have indeed been dropped, AMD will have effectively been given two years of process tech advantage over its rival. This will help allow AMD to build its brand as a technological innovator, while also allowing AMD to gain market share until Intel is able to strike back with its 7nm technology. By this stage, Intel will need to be ready to leapfrog AMD, or risk its position at the top of the x86 processor market.  
Intel accelerates 7nm Plans, 10nm will be a short lived node

Intel's 10nm server processors are reportedly still on the cards for a 2020 release, though Intel's 14nm Cooper Lake release will come within a similar timeframe, making it possible that much of Intel's next-generation server lineup will remain dominated by 14nm silicon. 

Please note that these rumours may not be true, though they make sense given Intel's lack of a detailed desktop processor roadmap. 

You can join the discussion on Intel's ongoing 10nm woes on the OC3D Forums
«Prev 1 Next»

Most Recent Comments

14-10-2019, 12:05:57

It makes sense to me that they would prioritise the 10nm silicon for servers since that is where the big bucks are. However AMD is simply going to wipe the floor with Intel for the next 3 years technically speaking.

The advantage that Intel has is that the server space is notoriously slow to adopt new tech. The recent security flaws which have hurt Intel more than AMD have however sped things up to a certain extent.

The truth is that any serious builds over the next 3 years, server or desktop, should probably be build with AMD parts. They will just simply be better in every way. However Intel have a whole load of money available to ride the storm and will seek to maintain market share. Unfortunately that will probably mean making deals with partners and will unlikely translate to a significant discount at the customer level.Quote

14-10-2019, 12:47:31

As long as there is a performance increase of at least 10% over the previous gen that's all that matters.Quote

14-10-2019, 13:00:35

They should have cut their losses long ago and put the money towards 7nm since they seemingly have had less trouble with it. Keeps getting pushed back by Intel wanting to make a profit off of 10nm. But by staying and spending with 10nm it only gets worse and worse for them... Now they are behind everyone else.Quote

14-10-2019, 13:09:29

10nm is mostly to please the shareholders at this point, iirc they scrapped three quarters of the fab space planned to be used for it a while ago and behind the scenes moved most of the focus to 7nm, in a practical sense 10nm should have been fully scrapped, but that could have p*ed off the investors so what was clearly illogical to the engineers leaking out Intel's 10nm woes since 2015 seems to have been perceived as a necessity to the boardroom fodder.

While Ice Lake/10nm probably won't be able to see any meaningful uptake in the enterprise market that's at least one area it could provide some very nice graphs and powerpoints, in those densities the limit is usually still thermal so the lower max clock speed limitations they've faced won't show through as they'll probably still be firmly thermally limited, similar situation with most mobile parts too today ofc. In performance desktop however where 100W+ coolers are a regular occurrence that 4Ghz limit would be laid bare, even a paper launch of cherry picked benchmarks would probably earn ridicule.Quote

14-10-2019, 15:41:54

Man, Intel are losing so much potential profit. That's an awful long time to rely on such a dated process.Quote

Register for the OC3D Newsletter

Subscribing to the OC3D newsletter will keep you up-to-date on the latest technology reviews, competitions and goings-on at Overclock3D. We won't share your email address with ANYONE, and we will only email you with updates on site news, reviews, and competitions and you can unsubscribe easily at any time.

Simply enter your name and email address into the box below and be sure to click on the links in the confirmation emails that will arrive in your e-mail shortly after to complete the registration.

If you run into any problems, just drop us a message on the forums.