New Generation Intel CPU's 'Tiger Lake' Processors

Agree with you, but consider that it's the first in the world.

 

The next will be saying a phone is a pc

 

See:
https://www.tomshardware.com/news/amd-zen-3-ryzen-5000-announcement-19-percent-ipc-1080p-gaming-lead

The tagline is "More 7nm nails in Intel's 14nm coffin."

Yet Zen3 is 5% ahead of last June's Intel 10900K efforts on a 14nm node.

As always, the news of Intel's imminent demise is not surprisingly exaggerated.

 

Basically what Apple said when the first iPhone launched in 2007.

 

5% ahead in what? I feel like you're picking a random %. It also beats Tiger lake in single core performance at the same speed in cinebench r20 single core... I do think AMD screwed up by increasing pricing, $50 bucks makes me think about getting a i9-10850K and saving some money just to mess around with a different platform for a little bit.

 

Yeah, I was about to mention that... and the fact that if he was going for lowest scores, he might as well then claim -3% in one of the games (in order to make his argument look more convincing).

What's more (and this is something I mentioned before), Tiger Lake is not impressive at all because it had 0 IPC gains over previous generation. Intel managed to optimize the node and only made a 4c/8th chip to get the clocks much higher (about 50% higher compared to previous generation).
Essentially, Intel managed to 'outscale' AMD through clock increase only and by making a CPU with half as many cores as 4800U... incidentally, because T.L. is a 4c/8th chip, its iGP can also pack more graphics cores than 4800u (big whoop - Intel has more space to work with).

When looked within proper context (which seems to escape many people such as tilleroftheearth and tech review websites), Intel is not impressive whatsoever.... but for some reason certain tech review websites seem to give T.L. big praise regardless.

 

The 15inch screen Lenovo Flex5 is going on a Tiger Lake refresh... Most likely, it'll be listed in some markets before black friday...
https://psref.lenovo.com/Product/IdeaPad/IdeaPad_Flex_5_15ITL05

 

The launch presentation by INTEL doesn't impressed me on the CPU side of performance. Infact, i believe, Tiger Lake improvements goes onto the iGPU side of performance, to take on what Ryzen CPUs can offer.

 

The iGP on T.L. does seem interesting, but to be fair, it's uArch seems to now be comparable to enhanced Vega in Renoir at best, and the only reason it 'wins' is probably because of having more GPU cores to access (it will be interesting to see if Intel will produce an 8c/16th part at same 28w TDP and then we can compare both CPU and iGP running under similar conditions as Renoir does).

If AMD put 10-11 compute units into Vega (as opposed to maxing out at 8), it would have about 30% higher performance than what it does now... but, we still need to wait for third party reviews of Tiger Lake to see how it behaves in OEM laptops (some of which will undoubtedly allow it to use its full 28W TDP, as opposed to most OEM's constricting AMD Renoir to 15W - the only Renoir laptop with 4800u allowed to use its full 25W TDP seems to be Lenovo Yoga Slim 7... everything else appears to be limited to 15W).

 

No, nothing random here. But while I still believe that 5% is accurate (can't find that right now), I did find this for you (7%).

See:
https://www.techradar.com/news/amd-ryzen-5000

So, in any case, 7% faster with 2 extra cores for ~10% more cost. Yay! (Not). This is not domination. Not even close.

On a 7nm node vs. 14nm and comparing to what was introduced back in June 2020.

I really don't care who 'wins', but I'm not swayed by blind devotion to a brand either.

This is still AMD catching up to Intel. Except for those who run Cinebench based workloads/workflows (all day) of course.

 

once again... INTEL gets it's win with
software optimization on some benchmarking programs
​

 

What software optimizations are you talking about? This is exactly what I've been saying for the last few years; 'moar cores' mean squat.

Do you want a big block V8? AMD.

Do you want something to go from point A to B? AMD.

Do you want something that will give you productivity gains? Intel.

Do you have workloads and workflows that are more complex than mere 'scores' may indicate? Intel.

Disregarding the inane Cinebench 'scores', four cores on an older node (or two?) is superior to double the cores on a nominally cutting-edge node. For productivity. This is not news to me. But many are surprised when their AMD super-amped-up workstations don't deliver as they thought they would.

And by super-amped-up, I am talking about the great marketing AMD does (rah AMD). They should spend that $$$$$$$ on delivering an obviously better experience instead.

 

@tilleroftheearth you forgot one important point:

Do you care about security? AMD.

 

I do care about security. Enough that the CPU vendor I use isn't part of that equation.

 

https://www.lenovo.com/sg/en/laptop...261&options_3=G0A3140CWW&options_4=GXD0U47643

@ Approximately USD$1100, this 11th Gen new model from Lenovo is aggressively(White Chassis, 16GB RAM, 512GB SSD) out to get attentions from AMD buyers!

 

Again, you can claim that RUGS are everything and Intel's hand selected benchmarks are what rules the roost, and say no one edits videos on their laptops, but that doesn't make it a fact. AMD had some huge increases in single core performance and the tomshardware link you posted shows that it's now faster in Adobe premiere, v-ray, solid works, and GCC compile times, all lightly threaded things that Intel once had minor victories in. When it comes to gaming, Intel has touted that it's the best and now that AMD is better it seems like Intel fanboys say it's not a big deal. It's the usual if it works for my argument then it's valid, but if not then...

I'm no fanboy and will run whatever is faster, and right now for the workloads I do (not opening word documents and timing it) AMD is the superior product for me. I'm glad Intel is catching up with their 10nm process (which in density terms is about equal to AMD's 7nm, and their 14nm+++++ is about equal to a typical 10nm). Competition is good, but blind loyalty is not.

 

And yet, once again, I did not make any such claims.

 

Core i7 vs. Ryzen 4000: Which mobile CPU is fastest in Photoshop, Premiere and Lightroom pcworld.com | 8Jan, 2020

In this round of the Intel vs. AMD battle, we test Intel's Tiger Lake Core i7 and AMD's Ryzen 4000 mobile CPUs in Adobe's Creative Cloud suite, using the Procyon and Pugetbench benchmarks.

 

I'm quite happy with my i7 1185G7 in my Stealth 15m I wanted fast single thread performance and I got it





derek bottoms

 

Nikon, NEFS, Intel. Killer combo. Still. Now, proven for the Intel naysayers, with pretty graphs and all.

Now, some of those AMD fanboys can see for themselves my ' similar' workloads and why Ryzen isn't an option, still.

Thanks for the link brother @Papusan. Wishing you a grand weekend.

 

The benchmarks don't tell us anything new - 11th Gen Intel is better core for core, but for those that need it, the 8 core 4800H will always come tops. Cinebench and compile times are the two benchmarks that most directly correlate with the tasks I perform at work, although I am probably (almost certainly) in the minority here.

Eagerly waiting to splash my (company's) money on whoever wins out between 8-core rocket lake and the next gen Ryzen mobile.

 

Yeah in single core right now with the old generation Ryzen, Intel has the lead. Things will change with the 5000 series in the mobile market. I'm glad that Intel and AMD are trading blows like the good old days. You still don't seem to understand that AMD being a key player is a good thing. If they weren't, Intel would be still on 14nm silicon.

Also, I'm glad Intel finally developed an i-GPU that is worth using. This will force AMD to stop resting on their laurels in that department as well. All good things (just like AMD giving NVIDIA a run for their money in the dGPU market).

 

Technically, Intel IS on 14nm silicon still (for the most part).

 

I had totally considered a Ryzen CPU in a laptop like the G14 but since I don't do heavy multithreaded applications I stuck with intel, as my tests have shown and verified what other reviews have shown, lightly threaded workloads are better done on an 11th Gen system

 

@custom90gt I've always said that AMD joining the game again after more than a decade was an incredible thing. That doesn't mean I would buy an AMD system 'just because'. Like I've said for years now when AMD makes it happen, it will.

After more than three years now, it hasn't happened yet on the mobile front. Don't have too much faith that the 5000 Series will do anything for my workloads either. At least, not when I already have solutions in place today.

I have just never accepted that 'moar' cores' are automatically 'better' for 'moar' people. Marketing brainwashing at its best.

And I'll repeat it here again: I don't buy 'shiny' new tech just because it is new. I buy it because it gives me a productivity increase in some measurable way for my workflows and that must be better than what I am currently comparing it against.

If Intel keeps producing 14nm 'ancient' tech for the next decade with the performance/productivity jumps we're seeing today, why wouldn't I buy it?

And once more, I'll thank AMD for waking up the sleeping giant for us. Really, I can't thank them enough! (But not enough to buy their systems, yet).

 

Seems there are only 2 people in this thread that haven't drank the AMD Koolaid, myself and tilleroftheearth.
I'll give some background and subjective analysis of my computing use and experience using the 11th gen i7 vs previous gens. I generally stick to and post actual data so I'm not always comfortable doing subjective analysis

My use case is not that heavy using a computer, some light gaming productivity software and vehicle tuning, nothing that my surface go couldn't do (with the exception of gaming)
The only use case that has changed in the last few years was using Beam.NG for creating and modelling crash scenarios (Car vs train) and just playing that game since it's a bunch of fun.

I initially went from a

Dell M4700
i7 3840QM
16GB DDR3 Ram
Msata SSD

To

Dell Inspiron 7000
i7 8550u
16GB DDR4 2400
Samsung PM961 PCIe 3.0 4X SSD

To

MSI Stealth 15m
i7 1185G7
16GB DDR4 3200
1TB Kingston PCIe 3.0 4X SSD (Even though the system is PCIe 4.0)
RTX 2060 Max-Q

From dell to dell system I lost zero performance gained battery life and single thread performance was much greater due to the higher turbo clocks, based on my work load the CPU was spending lots of time in high turbo boost states because I wasn't stressing all the cores all the time. The PCIe SSD also played a role in how "snappy" things felt, basically a huge upgrade overall compared to the M4700 I was using.
In the interim I built a cheap desktop so I could use Beam.ng

i5 9400 2.9ghz 6 Core all core turbo of 3.9ghz
16GB DDR4 2666
500GB Sabrent Rocket 4.0 500GB PCIe 4.0 (Running in a 3.0 system)
GTX 1660 Super

The desktop didn't really feel any faster than my Dell Inspiron laptop, but I could use Beam.ng better with a proper GPU. When I compare the Stealth 15M to the desktop, again overall performance "felt" faster, the laptop boots up faster loads application faster and even in heavy cpu usage scenarios the i7 1185G7 is faster than the i5 9400 in every metric


If I compare my Stealth 15m to the laptop (best comparison) its faster in every category. Now you could say that isn't fair because the CPU TDP's are different between the 2 laptops, and you would be correct. The i7 8550u in the Dell would ramp up to 3.6ghz all core turbo(44w) for about 20 seconds when all cores were loaded then settle in at 2.4ghz at 15w if all cores keep getting stressed. On light loads the CPU clocked as high as 4ghz. Basically the dell was quick and snappy and fit my use case entirely but didn't have a GPU which I now required for Beam.ng. The stealth has power limits but it never reaches the PL2 limit it and thermal throttles when all cores are stressed for a long period of time but reaches its maximum speed 4.3ghz all core and power spiking to 57W and does so for quite a while before it then settles in to a PL1 limit around 44-48w and maintains an all core turbo of 4.1ghz give or take 50mhz. The thing is I don't ever have sustained all core loads most of the time, even though Beam.Ng is very multithreaded. If you spawn more cars than you have cores then things start to slow down. In lightly threaded loads the cpu spends most of its time around 4.5-4.8ghz. Compared to the dell EVERYTHING about OS response and application loading and shutting down is quicker on the MSI. Now you could argue it's the new SSD and PCIe 4.0 but it's not because its not a Gen 4.0 SSD and it's actually quite a terrible SSD compared to the PM961 in the Dell laptop IF you go buy the benchmark numbers alone. So why does the stealth feel so much faster in general computing tasks and loading games ect? It's probably because of the CPU's ability to max out its single core turbo speed all the time it needs to load anything.

MSI's implementation of 28w TDP in their EVO and non EVO 11th CPU's is done the way more manufactures should do it, that is let the CPU stretch it's legs and hit max turbo almost all the time.

I almost want to go out and buy a Ryzen laptop just to test it out for my use cases to see how it would perform but like I mentioned before even though Beam.NG is very threaded, my use case isn't spawning a ton of cars more just making up complex crash simulations with no more than 4 vehicles. Single thread performance is what I wanted and there was not a laptop out there for my use case that fit the bill better than the 11th i7 CPU. Does that mean it works for everyone else, no it doesn't. But comparing an 8th gen core i7 CPU to an 11th Gen i7, there have been large gains made and forgetting about numbers the gains are in tangible user experiences.

TLDR, you can crap as much as you want on intel but they did something right with Tiger Lake whether people like to admit it or not.

 

Very well said! More cores doesn't mean much when you don't or can't take advantage of them, or don't even need them. They just become a waste of money. Regular consumers wouldn't benefit from more than a quad core. Gamers generally won't benefit from more than an octacore. The majority of people don't have such highly parallel loads that it makes sense to get anything above an octacore. For most people, single core performance matters the most because a lot of their tasks are serial by nature.

Games are like this too. No matter how well a game can scale with an increasing number of CPU cores, there is a point at which single core performance becomes a bottleneck because of the render and main threads. Until all software is written in such a way that only multicore performance matters, single core performance will be a bottleneck. This is impossible though because some problems are serial by nature, unless someone finds a way around that.

Having too many cores can cause a performance regression for programs that can't take advantage of them. That's another thing to think about when choosing a CPU. You have to look at your use case to determine how many cores you really need.

This is another thing a lot of people seem to glaze over. The lithography node of the CPU doesn't matter if its performance is better than ones produced on a smaller node.

 

Except for the fact that when you start pushing the cores frequencies, you get lousy efficiency (and will be limited in scaling).
Intel's power consumption goes to nearly a double for matching Zen's performance in single core and still trailing behind in multi-core performance.

Also, people like me do both content creation and game on laptops... for which a high core count as afforded by AMD is definitely useful.

Intel already lost on the litography and as a result efficiency (and performance) to AMD.
If they continue to cling to 14nm, then sure, if they continue putting new uArchs on that same node, they will experience performance gains in some aspects, but will be severely limited in scaling up the processor performance due to manuf. process limitations (which will also put a severe limit on efficiency).

Software also doesn't sit still.
Devs are now more conscious of AMD and are starting to optimize more for Zen cpu's and their instruction sets... and I'm sorry, but relatively minor MS Office gains (which I don't even notice because I already use SSD's and those are very specific set of circumstances that are an outlier rather than the rule) are (to me at least) NOTHING compared to content creation and gaming performance benefits I get from an 8 core CPU.

 

Nobody thinks that Intel is going to stay on the 14nm node process. I was being exceedingly absurd to make a point.

The point being that Intel is far from being out of this game. On the contrary, (see the post by @Papusan for proof), it is still leading even where it 'shouldn't'.

 

So you start pushing the frequencies and get lousy efficiency? So my i7 1185 is faster all around vs an i7 7700k taking less power but its not efficient? What do you think the single core power consumption is on tiger lake?

 

if you truly want to root for the underdog then amd and intel should considered as one entity and RISC-V is the emerging competitor in need of your money to create a paradigm shift in the monopoly of architecture and instructions which has dominated computing for decades!

 

It's not all about the lithography, that's true. However until AMD caught up to and surpassed Intel, there were no real gains since Sky Lake. Everything was minor bumps in performance and no real increase in efficiency. I've owned multiples of every generation of Intel since the core 2 duo and only a hand full of them were memorable when it comes to performance increases.

AMD's Zen 3 has >20% increase in IPC, it will help close the gap that Intel's first real performance increase has created (on their 10nm superfin process). Again, all good things.

 

I don't buy theoretical performance increases, I buy for actual, real-world productivity increases. I like the science behind them and enjoy reading the deep dives that AnandTech writes up. But those mean little to nothing if, in my workspace, those technical benefits evaporate in the heat of my production workflows.

I don't buy every generation Intel has offered. But the ones I've bought, they were genuine leaps to productivity, efficiency, and reliability over what they replaced.

For me to buy a computer on specifications is like buying a sports car because it's the 'fastest', for someone, somewhere, and usually for one lone and inconsequential 'test'.

Just like all tall people aren't all great basketball players; computers, workflows, and workloads are much more nuanced than what Intel, AMD, Samsung, or the fruity company may try to sell us year after year.

None of their words matter. A computer is a tool that we have the luxury to custom fit to us as much as we want. Buy the tool that is most effective in your workflows/workloads by testing in your workloads/workflows.

And leave the pretty marketing and their tunes with the great hooks, for the masses that think 'image' is the goal, rather than 'substance'.

You don't buy a Viper because it will smoke a Porsche on the straightaway, rather, you buy the Porsche because you want to drive wherever the road leads. Fast, confident, reliable, and allows you to always be in control... Can't say the same about AMD, er, I mean a Viper.

To put a fine point on it for me; AMD doesn't just need to close the gap in performance. They need to make their platforms bombproof too in a production environment and from top to bottom (hardware, drivers, compatibility, reliability). See you in 2027 to discuss that point again (I'm giving them a decade to prove it to me, three years has been like 3 minutes for the progress they've made up to now).

 

Lets see a Ryzen Mobile in the same test like this showing superior efficiency and the power consumption is almost double?????

 

You talk often about your production workflows but never expand on it. I'd love to see what you do, and any actual data you can back up your big gains with previous generations of CPUs on.

 

I'd love to see a sample data as well. If its AVX512 based workloads AMD CPUs have no answer for it at the moment.

 

See:
http://forum.notebookreview.com/thr...-lake-processors.831587/page-17#post-11069494

And that's all I have to say about that box of chocolates.

 

I think Viper vs Porche example is a bit extreme.. They both have had their share of issues. For Intel ranging from the Atom C2XXX clock failure issues to SkyLake QC issues and lack of thread isolation on HT vs AMDs SMT in regards to some of the security exploits as some notable ones.

Intel could afford to face them with out much harm due to the fact that there were no comparatively performing AMD products till the past three or so years even with mitigations and patches enabled to fix flaws/security issues. That luxury is fast disappearing.

@Vasudev, I wish the instructions comprising AVX512 weren’t so fragmented across product lines.

Both a former Intel engineer and Linus Torvalds have given their own negative views on AVX512. Pidnoel also dishes on problems with the way Intel is doing things right now and the trajectory.

Francois Pidnoel a former Intel Principal engineer also gives his thoughts on it half way down:
https://www.google.com/amp/s/www.pc...it-former-principal-engineer-unloads.amp.html

Intel has its own response to them:
https://www.google.com/amp/s/www.pc...s-who-wish-it-to-die-a-painful-death.amp.html

As for the future as I’ve stated before in terms of the overall chip market I feel Intel’s old days of near total domination/golden era are gone. I can see more companies join Amazon and Microsoft to go vertical with their own ARM implements for certain use cases and it will likely grow from there along with more general implements by the likes of Ampere. x86 obviously won’t die out any time soon just that there’s are real threat from ARM. One of the reasons why Intel invested in a RISC-V company’s called Si-Five along with Samsung and others.

Seems one of their investors wants more changes as well:
https://www.investopedia.com/activist-wants-intel-to-explore-separation-5093828

Some of those node issues are due to problems at the way Intel’s Technology Manufacturing Group was run and rigidity in the way of doing things/goals rather than a more pragmatic approach. They’ve started rectifying the issues. I guess we will have to wait and see if it was timely enough.

As for AMD downside of having to rely on TSMC is that aside from TSMC being hosed by orders from multiple clients they’re actually thinking of cutting orders from AMD to divert production for other clients. I mean even though AMD might possibly be the largest TSMC client next year it’s still a worry as they already can’t meet the demand. Intel is actually looking to offload some production to TSMC as well to add to that pressure. All this aside from the fact that Apple cutting out Intel for ARM means that production load also mostly falls on TSMC.

 

Should we compare the Bagger 293 Bucket Wheel Excavator to a Ford pickup truck with a snowplow then? When the 'workload' is clearing a few SMB's parking lots as fast as possible of the white fluffy stuff for less than half the year?

There are kids with exponentially more firepower in their desktops than what was used for the entire space-race back in its glory days.

Even if the Bagger was cheaper than the Ford, it doesn't make it the tool to use for most consumers.

And those kids gaming with AMD powered desktops aren't winning more either. But AMD is. And so am I. When I don't buy/use AMD.

Problems are to be worked around (by all parties); showstoppers are something else. Intel has problems. Look below for some of the showstoppers (beyond compatibility, stability, reliability) that AMD grants my 'similar' workflows. I don't have to be a fanboy to use AMD products in my workflows, I'd have to be an idiot. Rather, I just choose to use the tools that work.

 

I’m not sure how your comment is relevant to my statement about both having had their fair share of issues. You seemed to be making a very generalized statement on reliability with the car analogy rather than being specific to your use case.

And honestly on the consumer side the average joe probably wouldn’t notice the difference in regular use between most equivalent platforms from either company...

 

So you have nothing to say at all without any evidence. Thanks, good talk.

 

Ahh, the 'evidence', is in the posts after the quoted one.

 

Again, you say you have a workflow that benefits from Intel, but in all of your posts you've never said what that is... It's cool if you have no real proof, just say it.

 

Need to read, my friend, need to read.

 

This is the same thing you do EVERY time someone asks you anything about your workflow. Here is what you said:

All I get from this is maybe you shoot a Nikon and take NEF images? You don't mention what program you use, if you edit them, if you use this for your business, if you need real cpu power. You provide 0 information and expect me to somehow read into it. If it's as simple as you run Lightroom, need a laptop, and don't actually need the grunt of a desktop than sure Intel is a great option.

I could do something similar. For my workload AMD is 137.582% faster, just don't ask me what that workload is.

Interesting to see AMD at the top of all of these charts:
https://www.pugetsystems.com/labs/articles/AMD-Ryzen-5000-Series-CPU-Review-Roundup-1962/

 

Do you ever get sick of repeating yourself? I know I'm sick of repeating myself.

Search if you really need to know more, it’s all here. I’ve been here a while too. There’s nothing left for me to say when talking to a wall.

The interesting thing about that link is all the old tech and desktop usage from half a year ago, and 'scores' that are given for ‘average content creation performance’. That’s not how I work. I do actual work, and not some random average ‘score’.

I write pretty plainly. Contrary to trendy belief, Intel’s focus, dependability, and stability are what make it the most productive platform, and not just for me either. My workflows that require snappiness, reliability, and compatibility, each time I sit down at my workstations, just also happen to be what most consumers want too. Most. Not all.

The point of the article that @Papusan linked is proof of that on the latest platforms, and is really hinting where Intel is going with their focus (i.e. where they should be going).

When 4 cores outperform 8 cores (like I’ve been saying for years already), for most used workflows/workloads, it is just proof that ‘moar’ cores are just marketing. It was great marketing 4 years ago. It is looking a little desperate to me today.

As a business professional, I choose the tools best suited for my needs. On this mobile forum here, Intel’s platforms have been untouched where actual productivity is required, for most. For the few that do need high core/thread mobile platforms, they’re available for them too. But I can’t trade bragging rights for productivity, compatibility, and dependability.

So, keep showing the ‘scores’ to prove how wrong I am in the above and all my previous attempts to educate here. I know you know they don’t mean anything in the real world.

The ultra-mega companies you point to as AMD users are not doing ‘what most consumers do’, nor are they doing it ‘mobile’, either. They also have the manpower, inside connections, expertise and spare systems/parts and $$$$$$$$$$$ to handle all the glitches the still unproven to me (and many others; search for enlightenment if you dare) ‘moar core’ platforms can only offer to me today.

If I can get only one point across to you tonight; options are great if chosen wisely. Choosing the best platform by ‘score’, while ignoring the workflows/workloads that make up most people's days? That’s missing the point/benefits of the choices we are fortunate to have available to us today.

The only ‘benchmark’ score that matters is the one that is based on your actual, real-world, varying, and specific to you, workloads/workflows. Always.

Relying on canned ‘scores’ that isolate/highlight and effectively distort features/benefits that an article, a manufacturer, or a narrowly focused group is hellbent on shoving down your throat is a blow to my intelligence, today, in 2021.

And before you say it; I am not parroting Intel’s similar agenda either. They are parroting me (or more specifically; my experience in the 5 decades I’ve been using these once new and shiny tools called ‘computers’). More accurately, maybe they are just seeing computing from a more holistic view than their competitors.

I’ll repeat, I don’t buy everything Intel offers with blind devotion. I don’t buy platforms, period if productivity isn’t increased. That’s because I don’t buy on marketing claims, ‘scores’, new/better/more process nodes/cores/widgets, etc.

I’m part of the buying department in my company. I’m part of the IT department too. I’m not there because I read online rags to know/find out the core worth of stable/reliable/dependable and performant platforms with a ‘score’ to prove my points. No, I’m there because the points raised above and many times previously have been proven time and again. I benchmark the entire workflow/workload in its entirety, not on microscopic sections to show one side or another a 'win'.

Productivity gains are only measured in time units gained, if they’re real. This takes into account everything and anything that may contribute towards those gains while dropping the bs at the front door. Anything else is just so much ‘blah blah blah’ to not even be worth debating. My vision is clear on the computing tools I rely on every day. Is yours?

I can only hope to make you understand the points I’ve repeated above, once more. But I’m more hopeful that others see them as they are presented.

P.S. I’m not selling you anything. Everybody else is.

 

I find it a bit weird that while you argue fiercely for people to trust nothing except benchmarks on their own workloads, you tout specific software 'victories' for the Intel platform as evidence of their vast superiority - just as code compilation and other general multi-threaded workloads might be irrelevant to the masses, equally, not everyone is using (those specific) Adobe software in their daily workload.

Furthermore I would question the most recent set of benchmarks between the laptops, as it is not clear to what extent the tasks are GPU accelerated - one Intel laptop is packing a 2080 Super-MaxQ while two of the AMDs are using 1650ti/radeon integrated. Outliers in the results (for both AMD/intel) are all over the place, and the article does mention the possibility of problems switching between integrated and dedicated graphics.

 

Again no mention of your workflows, just that you know Intel is more reliable and better for these workflows that you do. And like mentioned above, benchmarks are only worthwhile when they suite your point but other times they are rubbish. I'm not going back to search your old posts, I just wanted to know if you actually had a reason. Clearly you don't, or don't have good evidence so you'll just post some random comments about it. It's fine, I don't have a dog in the fight, but wanted to know why you're so set on Intel.

 

In other news:
https://www.crn.com/news/components...tion-of-some-top-chips-to-samsung-tsmc-report

It would be weird to think about Intel not producing their own chips in house.

 

although the final decision of actualizing has only just now circulated, this was reported back in the summer and throughout fall was essentially universlly accepted as an an eventuality. tldr, industry and investors and others with interest have long known this was not only in the works but well on its way thru the pipeline.

just sign of what's to come, at least until/unless US truly commits to a homegrown foundry arms race to avoid this trend metastasizing further.