Apple's A14 chipset at 5nm is now shipping, in an Ipad no less.

This means early next year the 5nm phones come out, and later next year AMD 5nm chiplets will be built im mass, something that will continue ongoing for YEARS and YEARS until 3nm is ready.

And here is our buddy Gary, to explain what little is known at this time.

https://www.youtube.com/watch?v=F_IsBD6E3O4

https://www.anandtech.com/show/16084/intel-tiger-lake-review-deep-dive-core-1...

I mentioned that Anandtech had promised a THROUGH review of all Tiger Lake functions.   Over the weeks since then new web pages of results have been released off this same original review, totalling 19 web pages of Anandtech testing results.   Exhaustive testing was done on each of the BS Intel Marketing claims and each one was dissected separately and completely.   (look for the drop down list of test pages to get to this information)

After literally a month of detailed analysis work, Anandtech has reached a conclusion.

by Dr. Ian Cutress & Andrei Frumusanu, Anandtech Magazine   condensed

Conclusion: Is Intel Smothering AMD in Sardine Oil?

The new Tiger Lake still falls down against the competition when we start discussing raw throughput tests. Intel was keen to promote professional workflows with Tiger Lake, or gaming workflows such as streaming, particularly when measured at 28 W or higher boost levels rather than at 15 W industry standard as is used by AMD.

AND WE REPEAT:  The new Tiger Lake still falls down against the competition when we start discussing raw throughput tests. Intel was keen to promote professional workflows with Tiger Lake, or gaming workflows such as streaming,  but ONLY when Tiger Lake was run at 28 W rather than at the 15 W industry standard as was used consistently by AMD.

 Do you get the point?  ALL of the "critical" Intel testing  ??? wins ???  were done at an intentional 13 watt + over normal voltage & current draw levels ....... (and you could just as easily say "at double the industry standard amperage draw" and be within 2 amps of totally correct)

Despite this we can easily see that the 15 W Renoir options with eight cores can blow past Tiger Lake in a like-for-like scenario in our rendering tests and our scalable workloads. The only times Intel scores a win is due to accelerator support (AVX-512, DP4a, DL Boost). On top of that, Renoir laptops in the market are likely to be in a cheaper price bracket than what Intel seems to be targeting.

Only if Intel can convince software developers across the globe to jump on board with using its accelerators in all existing software only then will the customers get full benefit from Intel’s tweeked metrics. The holy grail of this impossible trick may be when it comes to Intel's OneAPI, enabling programmers to target different aspects of Intel’s eco-system under the same toolset. However OneAPI is only just entering v1.0, and any software project base building like that requires a few years to get off the ground.  And right now OneAPI only affects Intel's own software packages.

We didn’t have too much time to go into the performance of the new Xe-LP graphics, although it was clear to see that the 28 W mode does get a good performance lift over the 15 W mode, perhaps indicating that DG1 (the discrete graphics coming later) is worth looking out for. Against AMD’s best 15 W mobile processor and integrated graphics, our results perhaps at the lower resolutions were skewed towards AMD, but the higher resolutions were mostly wins for Intel - it seemed to vary a lot depending on the game engine.

by Dr. Ian Cutress & Andrei Frumusanu, Anandtech Magazine
As a concept, Tiger Lake’s marketing frustrates me. Not offering apples-to-apples data points and claiming that TDP isn’t worth defining as a singular point is demonstrating the lengths that Intel believes it has to go to in order to redefine its market and to obfuscate all direct AMD comparisons.



===================================================



There was a time and a place where Intel apparently felt the need to share everything, as much as possible, with us as the computer press. During this period Intel actually led us in a group think tank discussion to sculpt the story of where we envisaged the market was going, and Intel OEMs/builders/customers were on hand to add their comments about the viewpoints of the customer base from their perspective.

It then let us, as the press, filter back with comments, critiques, and suggestions. This new "cooperative" twist from Intel’s client division, one that’s actually been progressing along this quagmire path for over a month now, will only serve to confuse its passionate customer base, its enthusiasts, and perhaps even the financial analysts.

It is clear that by this "request for help" Intel has attempted to co-opt the major review houses as a group to HELP INTEL CRAFT THEIR AI BASED BS "PROCESSOR PERFORMANCE STORY" in a way that the major review houses would then support.  

This fraudulent conspiratorial action, by itself, is grounds for a BRAND NEW very large EU fine for Intel, in my opinion.  

IMHO, Intel was in essence attempting to build a conspiracy to commit fraud .......

It is clear that not all of the review houses were receptive to Intel's little scheme.

So, Anandtech for one wasn't drinking any of the Intel kool-aid .......



RECOMMENDATION:    Skip the Intel 10nm generation COMPLETELY, it is based on a bogus foundation of rigged over wattage hyper-clocking lies, marketing exaggerations and outright AI trickery.  

Tiger Lake acts much like a FRAUD in motion ......


.....  and Intel will eventually figure out that pissing off the entire computer reviewer community by asking them to lie isn't the best move they have ever made  ......

https://www.xda-developers.com/us-government-antitrust-charges-google/

Attorney General William Barr has issues with Monopolies, and he quite competently has determined that Google has become one (yet again).



Google has already went through this recently with the EU, twice, and the implemented EU remedies will provide Google's main defense in this USA Monopoly trial.

The harsh fact is that once you give everybody a free choice, they will still pick Google 94% of the time because it works so much better than the alternatives.   No joke, if you don't provide a path to load Google, users will find a way to SIDE LOAD it because they bloody well want it .......   I just went through all the steps of side loading Google on my Amazon Fire devices, so I know what of I speak.

This anti-trust stuff always winds up always being a money play with the EU anti-trust actions, Google tunes a few things, pays a few 10's of Millions in fines and removes some restrictions intended to keep fragmentation at bay, then later on gets politely requested by the EU to put the same restrictions back in place later when Gooble services all starts to unravel in the EU due to the EU fiddling with things they did not really understand.

Google provides the checks and balances on Microsoft and Apple, and same thing can be said backwards for Google.   There is no real enforced monopoly here, simply strong personal preferences showing itself all over the place again.

Barr may decide to bust Google up into separate companies ---- except it already is structured as separate companies and dicing it up further into divisions=companies is sorta meaningless to a communications company like Google.  He can do that though.  He can force them to compete with each other too, like they already do for the available resources inside Google.

Google as a product is still better, and is still prefered by most people.

 Why the female Hulk images?

Going over to the flip side of  things, the Intel side.

Intel has been very successful with their BS Marketing campaign, their pet vendors have flooded the world with AI based and over voltage & wattage hyper-clocking based claims of equivalency (and even outright claims of Tiger Lake superiority which are even more dubious).

Intel wants everybody to consider their AI tricks as part and parcel of processor performance, instead of a bias implemented set of software tricks (i.e. intentional cheating).

For every honest review put out that focuses on actual processor performance, it quickly gets shouted down and muddled by a host of "other" reviews.   Shady reviews and small shady review houses seem to abound right now .......

Customers now seem to depend more on Amazon Customer Rankings than real computer magazine reviews right now since they know they cannot trust traditional computer reviewers any more, but Amazon Customer Rankings have always responded more to the very last sets of rankings received than to the historical mass of rankings on a given processor so that particular trick isn't working right either.

Some people now seem to make a living by putting bogus reviews up on Amazon now-a-days, it has now gotten so bad that Amazon now lists whether or not the reviewer had ever even bought the product in question (right up top in the review header).

Based upon a thick greasy churning sea of lies, Tiger Lake is doing quite well.

ONCE AGAIN, to regain the position lost by intentional BS Intel PR, AMD is going to have to put out a whole new series of clearly dominating processors and get laptop builders to build them into units in large quantities very quickly to keep on taking market share away from the Intel PR machine.

This will be hard, as Intel is working on the suppliers themselves again, trying to get them to only use Intel chipsets (illegally again).

This sort of exclusivity deals were made illegal in the EU a year ago, but are coming back again because Intel feels them to be worth the risk to pursue them
as no one in the EU is really enforcing the last set of decrees and actions against Intel.  

Covid 19 is taking up all their attentions .....

Intel has become emboldened by their new PR tactics, the BIG Artificial Intelligence LIE is working well for Intel right now and the BIG Artificial Intelligence LIE really is intel's only path forward at this time.

https://liliputing.com/2020/10/intel-rocket-lake-s-chips-will-bring-intel-xe-...

Intel Rocket Lake-S chips will bring Intel Xe integrated graphics to desktops in 2021

Despite the fact that Intel has moved to 10nm for its laptop processors, Rocket Lake-S chips will be manufactured using a 14nm process. But Intel says Rocket Lake’s “Cypress Cove” architecture combines features found in its recent 10nm chips, including Intel Xe graphics (which debuted in 11th-gen Tiger Lake processors) and technology from 10th-gen Ice Lake’s Willow Cove CPU cores that have been backported to 14nm.

You can also expect much higher power consumption from these desktop chips – while Ice Lake and Tiger Lake processors tend to top out at 25 to 28 watts, Rocket Lake-S chips will likely have a TDP of up to 125 watts.

According to Intel, the new chips will offer up to 50-percent better graphics performance than an equivalent 10th-gen Comet Lake-S processor, and “double digit” performance gains in instructions per clock (IPC).

Other features include:

Up to 8 CPU cores and 16 threads
Up to 20 CPU PCIe 4.0 lanes
New overclocking features
Support for DDR4-3200 memory
Support for up to three 4K/60Hz or two 5K/60Hz displays
Intel is also promising improved AI performance thanks to new Intel Deep Learning Boost technology.


Translation:   more Intel 14nm lithography yet again, but this time using all the dirty tricks piloted by Tiger Lake including hyper clocking, extensive use of AI and over volting and over amping of all major items during benchmark runs.  

Expect very high running temperatures as both the hyper clocked CPU and the GPU will run very hot ......





But the company did give us a clue about power consumption by explaining that its performance comparisons are what you’d get if you pitted a computer with a 125W Rocket Lake-S 8-core processor against a machine with a 125W Intel Core i9-10900K 10-core Comet Lake processor.

If Intel IPC has "double didgeted" in reality, much better throughput will seem to be present with these chipsets (in general compute tasks, anyway).  
This is not seen as happening so far.

However, Intel has figured out how to muddle up all forms of benchmarking now using AI effects, so expect your benchmark results to lead you off into yet more layers of confusion.  

Intel counts on this confused AI mess now, lying voraciously in an organized manner to make sure confusion takes place and remains in place.

In an AI laced environment, Intel will claim 2x to 4x improvements in certain items (mainly graphics effects).   HOWEVER, Intel will make these as "general IPC claims" on the box, knowing Joe and Rita Six Pack really don't care all that much for the fine details when comes to computer shopping ......

Do not trust Intel to tell you the truth any longer, as the truth is simply not in Intel's best interest any longer.  

"Intel Marketing" is Intel's main progress maker now-a-days.

https://www.datacenterknowledge.com/amd/xilinx-stronger-amd-or-without-it





Xilinx’s Adaptive Platform

“Traditional CPUs are not able to provide the performance scaling required for some of these compute-intensive workloads,” Xilinx senior VP Salil Raje said at this time last year during the company’s XDF conference in San Jose, the company’s home base.

“That is why domain-specific accelerators are stepping up to take on some of these larger workloads,” he continued. “The issue is, however, that the workloads continue to change. They depend on algorithms and standards that continue to evolve. That is why adaptable devices, such as FPGAs and ACAP [Adaptive Compute Acceleration Platform] devices, have started to gain momentum as adaptable, domain-specific accelerators. They can be customized for your domain, for your application, and they can be configured on the fly.”

ACAP is a name Xilinx has given to what it considers a new class of processor – if that wording rings familiar, it’s because that’s what Nvidia said it would build together with Arm – in which programmable logic, built-in logic (in the form of “digital signal processor” or DSP, engines), Arm Cortex-A72 cores, and programmable networks reside in attached blocks on the same processor. The re-programmable, “adaptable” part are the concept’s selling point. Unlike a traditional ASIC, where an upgrade involves a truck roll, an ACAP would evolve in place. For now, ACAP is being marketed under the umbrella brand Versal.

The AMD and Xilinx deal is about the convergence of CPU, GPU, and other accelerators on a system or motherboard, Kurt Marko, principal analyst at Marko Insights, told us. “FPGAs provide the flexibility to implement a multitude of features that can augment x86 cores by offloading and accelerating particular functions.”

The Xilinx platform is comprised of well-respected programmable logic boards governed by a system-on-a-chip (SoC) called Zynq. A programmable logic board can be called an “accelerator,” if you have a system that’s geared to run applications that use the exact same software routines repeatedly, such as AI, video transcoding, or encryption. Rather than rely on a CPU to run these routines in DRAM, an FPGA card may already have those routines stored and running in highly available local memory.

All of the major Compute makers are pairing off with an AI company.   AMD has Xilinx,  ARM has Nvidia,  Intel has Altera.

All of them are looking forward 5 years to when there is really nowhere to go with shrinking the basic lithography in any meaningful fashion.

So, in the end you pick the pair that you feel you can trust the most, and then you go forward into the future with it.


        quote from Harry Potter book #3's talking shrunken head on the Wizard's bus


"Hang on tight, it is going to be a bummmpy ride ........"