How exactly did Intel make a 15W version of the 7920HQ?

Usually, lowering TDP by 66% would require a die shrink or at least a new architecture, but the 8650U is neither...

 

By throttling it.

 

Lower voltage & lower power limits.

 

There's always some variations in between CPUs so they are likelier to take the better ones for those rather than for the HQ parts.

 

Max Spec Marketing chalks another victim methinks

TDP and base clock indicate the behaviour of the CPU under sustained workload. While it *can* turbo to similar 4ghz+ speeds, the 15W CPU will quickly drop off in a direct comparison with both CPUs under sustained load to maintain its lower power limits. Turbo only adds frequency beyond the base clock when the power and thermal limits allow it.

The user experience may be indistinguishable under very light use, single webpage, opening a text only document, but try and run an intensive game or encode a video for example they will be chalk and cheese.

Intel has done nothing special, just creating different performance profiles for insertion into different products - the U CPUs usually go into tablets and small/light/thin laptops vs the high end HQs into bigger laptops and, if you buy the wrong brand, high end gaming laptops

Any CPU has this characteristic, for example my 3940xm running a 32M Pi (multicore 100% workload) in TSBench locked at various speeds comparing power consumption, voltage and performance (time to complete a fixed workload)
1.0GHz - 14.5W 0.93V - 32.6 sec
1.5GHz - 18.5W 0.94V - 21.6 sec
2.0GHz - 22W 0.95V - 14.8 sec
2.5Ghz - 26W 0.96V - 12.1 sec
3.0GHz - 32W 1.03V - 9.8 sec
3.5GHz - 51W 1.21V - 8.0 sec
4.0GHz - 61W 1.29V - 7.4 sec
4.5GHz - 88W 1.4V - 6.6 sec

Intel could take this same exact piece of silicon, call it a 15W 1GHz base clock part or a 4.5GHz base clock 88W part, or anywhere in between.

 

Read some reviews regarding the i7 8550u, yes it throttles but it is way faster than most people think

https://www.techspot.com/review/1500-intel-8th-gen-core-quad-core-ultrabooks/

 

It is "briefly" way faster, then it times out on the Turbo and ceases "melt-down".

Benchmarks that favor Turbo Timeout's - running only as long as the timeout's are operating - can make these new 15w CPU's look competitive with 45w / 91w / 95w CPU's, and for some applications they may be.

But if you have long running things, the thermal headroom and power headroom will disappear before the job is over, and power throttling + thermal throttling will make them fall apart.

It's an illusion, don't fall for it

 

Did you happen to read the review, yes they throttle, but they get up to speed and temperature after 1 benchmark run. So what you are saying is that all the other reviews showing this with a massive performance increase over the dual cores is wrong? Or the fact it is as fast as an i5 7300HQ which is 45W TDP. Maybe you should read my thread with non scientific benchmarks comparing the 2 laptops in my Sig. For intents and purposes. The 15W i7 8550u beats out my i7 3840QM,

http://forum.notebookreview.com/thr...ing-behavior-on-dell-inspiron-13-7000.809595/

Its no illusion and I haven't fallen for anything i've just managed to replace my M4700 with an Inspiron 13 and didn't lose any performance like most people proclaim you do.

 

You are answering what I already said. The short run benchmarks show great performance, but when the timer runs out and it needs to power throttle before it thermal throttles, it's game over - it turns from a beautiful princess into a dust bin.

Hey, believe what you want to believe, if you don't look at it too hard, or test for long runs, you'll remain happy, I'd hate to take that away from you.

For everyone else, watch yourselves it's just a 15w CPU, it can only pull the rabbit from out of the hat for a short time.

5 Minutes on Tech: Intel 8th Gen CPU - Worth Upgrading?

The new 8th gen are a nice upgrade within the 15w realm, but 45w/95w CPU's have nothing to worry about

 

Yeah, yeah...

Keep believing what you want to be true, but believing isn't making it so.

That video is laughable. She's comparing an i7-7700HQ to an i5-8250u, uhm... okay.

See:
https://ark.intel.com/compare/97185,124969,124967,122589,124968


See:
https://www.cpubenchmark.net/compare.php?cmp[]=3042&cmp[]=3064&cmp[]=2906


If we try a little harder to compare apples to apples, we see that at almost double the TDP (assuming we're using a chassis with TDP configured to 25W for the i7-8550U) the 7th gen HQ model is about 11% faster in multicore tasks. Not a good trade off for a mobile platform, imo.

The biggest trade off for the still not comparable i7-8650u top 8th gen model right now is double the RAM to 64GB total and 4 extra PCI lanes for the HQ. While my workflows demand the most RAM I can stuff into my platforms; battery life is equally important for a mobile system too, for me.

What is interesting is that the u model is equal to the hq model in single threaded 'responsiveness'. That is the icing on the cake for me.

Will the 8th gen processor throttle in a 15W capable chassis? Uh, yeah. That's because they're able to work at up to 25W TDP in an appropriately designed system.

With same 14nm tech size, this is a phenomenal increase in performance at the lower TDP. Ignore at your own risk.

 

Isn't it nice to see someone that gets it that the performance is a big increase. The 8250u and 8550u have a 600mhz difference in turbo speeds in single and dual core work loads that is enough of a difference to be noticeable. Also the benchmarks I ran had my cpu throttling during the cinebench work loads but it was the same speed as the i7 3840qm so how is that a negative.....

 

You keep saying that I'm saying the CPU improvement is a negative, I'm not.

I said the benchmark results are misleading because they measure the short duration that the 15w CPU can pretend is a big CPU - benchmarking higher than 45w CPU's. That's the misleading part, I didn't say it was negative, I said it wasn't representative of what the CPU is capable of in long durations.

The same can be said of 45w CPU's in comparison to full 95w desktop CPU's, they have a short duration where they can perform as full performance desktop CPU's.

 

Could you Test 1024M Bench from http://downloads.putme.net/wPrime210.zip

 

I just get an error when it loads up running Furmark CPU burner, the cpu speed never drops below 2.4ghz and fluctuates between 2.4ghz and 2.6ghz

 

Unfortunately, the 45w CPUs can't "perform as full performance desktop CPU's" for any duration of time. My 6820HQ for instance can hold 3.2 Ghz max quad core turbo indefinitely while running Prime95, but a 6700K's base clock starts off at 4 Ghz. Single core performance is a rather similar story - 3.6 Ghz vs 4.2 Ghz.

Turbo Boost goes a long way towards bridging the gap between 45w mobile and 95w desktop, but even when cherry picking the most favorable scenarios for mobile, desktop will still win.

 

And, the K CPU's can OC 4 core multipliers even higher, and with good cooling and power they can hold those higher than stock Turbo speeds indefinately.

The trick of short duration Turbo long enough to get a good score in a benchmark for 15w and 45w CPU's is just that, a trick.

The newer 8th gen 15w CPU's are an improvement, but the high cost and against their true nature of being Ultra Low Voltage CPU's for use in long battery life applications, it makes them a whole different solution than a 45w CPU or 95w CPU.

Trying to bridge the performance gap the ULV CPU's lose a portion of their appeal by losing battery life as an expense of higher performance. Adding a discrete CPU active on battery increases that battery loss.

You can't get free performance and not loose battery duration, and adding battery mass increases the carry weight and size of the laptop, all going against the purpose of a ULV tablet / 2 in 1 laptop.

At some point improving performance past the capacity of the battery to deliver it will force it onto AC for full performance, again going against the design purpose for a ULV tablet / 2 in 1 laptop.

Pick one form factor, an ultra light tablet / 2 in 1, or a heavier high performance laptop - which performs best on AC power, you really can't have both in the same device, not until compute performance can be delivered on battery in the form factor of a light carry device.

I suggest getting both form factors, one really can't supplant the other

 

You're still missing the point I see.

There is no trick. Vs. what we had before; this is an improvement (as you have already conceded).

Don't worry; this is just the first wave of 8th gen mobile chips. Higher performance ones in better chassis' will come along.

As I've stated elsewhere; desktop platforms mop up any mobile platform I've ever used. Always.

When desktop and mobile parts are both below the TDP limits of all the components used within the notebook offerings; then they'll be 'equivalent'. Until then, mobile platforms are many generations behind current desktop offerings when actual (and especially sustained) productivity is required.

There is no 'high cost' of getting the 8th gen processors vs. anything older. If anything; they're effectively cheaper today - all things considered.

 

And, you are using the 100%/100% Processor min/max settings of the High Performance power plan?

BTW, are there other items in XTU that are tunable for that CPU, besides voltage offset for undervolting?

 

I haven't unervolted or tweaked anything on my inspiron just use it on the
"high performance" profile while plugged in and "power saver" when on battery with the cpu limited to 1.25ghz. I find the power saver setting with limiting the cpu to 1.25ghz to work just fine for my uses.

No one was ever comparing these 15w cpu's to the overclocking geared K series CPU's.

There is no trick with benchmarks with turbo boost. Generally usability when not pegging the cpu at 100% on all cores results in a 3.7ghz - 4ghz clock speed.

So i've got a system that is for all intents and purposes a 3.7-4ghz dual core system and a quadcore system for the more intense work loads.

If you want to look at the whole system performance, the Passmark results on my Inspiron vs my precision are close, with the inspiron taking the lead.

What does that mean? It means that I moved from a laptop with a 45W quad to a laptop with a 15W quad and didn't lose performance, and i've proved this with benchmarks time and time again.

 

It's a 5+ years old 45w CPU design against a brand new just released, first serious ULV CPU performance gain in generations, ULV 15w CPU. Intended for tablets / 2in1's.

And, the higher the performance, the less the battery life, so there's a trade off for higher performance, higher power draw under load.

Try it against a laptop 7700k or a 7700HQ, then again against an 8th generation quad core desktop / laptop CPU when they finally arrive.

BTW, the namesake CPU of this thread, the Intel i7-7920HQ, isn't a 15w part it's a tunable 45w part that can be TDP reduced (Configurable TDP-down) to 35w based on application needs.

 

I don't need to try against in i7 7700K or an 8th gen desktop CPU because they don't put those in laptop that anyone buys thinking about having a slight idea of portability.

The performance progression from the Ivy Bridge Era Quads to the Skylake i7 7700HQ isn't really that much performance at all, really more power improvements than raw performance which is why i still say its relevant that a 15W QC can outperform some 45W ones.

The i5 7300hq is only a 4 thread (not 8 thread like the 8XXXu series) but it still gets beat by the 8th gen mobile cpu's, once again 45w vs 15w, the performance is there.

As for power draw, my M4700 does 130W full load, my Inspiron does 45W, yet the inspiron is just as fast, once again i'll take the inspiron over a heavier more power hungry CPU and not lose any performance.

 

You are fitting the data to your argument, selectivly picking what defines your premise, which btw isn't relevant to anyone but you, the owner of the 5+ year old 45w CPU laptop.

Your new ULV laptop is performing as good as your 5+ year old laptop with a 45w CPU. They are not comparable laptops, they are completely different, just as you also pointed about about contemporary 45w laptops.

Stop spinning, if you are going to compare a 5+ year old 45w CPU to your new ULV CPU, then it's perfectly valid to compare it to current 45w CPU's, right?

 

Even compared to current 45W mobile processors, Silvr6' argument is still solid.

Your's? Not so much.

 

And Passmark say the opposite... Strange!! Still, would be fine if you could test both processors in wPrime 2.10 (Run 1024M). Apple vs. Apples

 

Those are averages, single thread performance is better on the 8550u

Look at the last 5 baselines to get a sense for what they score and the sample size of the 8550u is quite small right now.

My inspiron detailed CPU passmark info


upload image websites

My M4700 detailed CPU passmark info


upload image websites

 

I gotta say I am surprised and amused at the thought that ULV CPU's based devices need to compete on performance. The service ULV CPU light carry devices provide isn't based on performance, it's based non-intrusive carry and use, and on battery life.

I wouldn't want my full sized laptop to be burdened with a large battery or undersized components as it would impede on thermal and power handling, compromising the full performance I want in a large laptop.

Neither would I want a high power CPU / GPU in an thin and light tablet, 2 in 1, or light carry laptop. High performance in such a device goes against the design goals of a light carry device with long battery life.

It's really not necessary to compete with a ULV laptop against a full sized laptop of the same generation CPU / GPU, or for a full sized laptop to compete against a unlimited power and cooling of a full sized deskside / desktop computer.

They all play different roles, and play by different rules.

Stop trying to mush them all into 1 device, we are nowhere near being able to do that

 

We're more than near; we're already mostly there.

The shades aren't dark with this one; they're made of wood.

 

I'm not sure why it matters all that much anyhow. When having the highest performance possible is the goal, nobody in their right mind buys a thin and light notebook. When having something thin and light is important, nobody in their right mind purchases a DTR or performance-oriented notebook. Using the wrong tool for the job almost always ends in failing to achieve what you set out to achieve in the beginning. We expect components like CPUs and GPUs to get faster over time. So, an 8th generation consumer level CPU performing on par with a 3rd generation consumer CPU while using less power probably ought to be taken for granted. If it didn't do that we would have something extra (and legitimate) to complain about.

 

Lets compare the 5yr 45w cpu to a newer 45W CPU



it is fair to compare against a 5yr old CPU because the 45W cpu's from today haven't really gotten any faster

 

IPC can make some difference, but one thing remains a constant. When core and thread count is equal, clock speed matters more than anything else. As the old saying goes, "Clock is King."

 

Yeah, I know. As you didn't want test with Wprime... Could you test CINEBENCH_R11.5 and CINEBENCH_R15 with your i7-8650U ? Below is results from +5 years old Mobile Core i7 (same as what you self have) you can compare with. Don't bother with the clocks, as the older +5 years *antiquated* mainstream Mobile Core i7 can be fixed. New ain't always better as we know...

http://hwbot.org/submission/3025220_dekoter_cinebench___r15_core_i7_3840qm_717_cb


http://hwbot.org/submission/3025220_dekoter_cinebench___r15_core_i7_3840qm_717_cb



Edit. If the Cinebench results below is about correct for your i7-8650U, there is no need for the suggested testing. Numbers as you can see... Found it on the web.

 

There is and has long been "something extra (and legitimate) to complain about" - a 3rd generation consumer level CPU (3840QM) will absolutely smash a 7th generation consumer level CPU (7500U). If anything, not having something to complain about per your conditions is new to the 8th generation, not something to be taken for granted.

 

Oh, I agree with you on that. Let's hope the days of lousy performance are over for everyone now. Not having a basis to complain about lower performance on a much newer product is a good thing. Nobody should complain about more performance. There are other things more worthy of complaint (like BGA and lower TDP). It should be expected and taken for granted because of the huge 5-generation gap. The 7500U was just lame and there is no excuse for the fact that it couldn't hold its own against a locked antique mobile CPU like the 3840QM.

 

You talk about 7500U. Even 8650U can't compete. Yeah, it can in single threaded workflow, but can't hold its own vs. antique 3840QM in multi threaded load. I expext its's a reason Intel pushed hard for high single core clocks for their ultraportable processors and not for all 4 cores.

 

Is the 3840QM based on Ivy Bridge?

Because the 7700HQ is slightly slower than the 7820HK because of less L3 (?) cache, but I don't know if this applies to the 7820HQ, when it is forced to run at the exact same clockspeed as the 7700HQ and the QM.

Seems IPC is only slightly better.

 

Yeah, Ivy Bridge. 7820/7920Hq should score exactly as 7820hk (same Cashe) with same clocks.

Edit. IPC clocks for clocks with 3.5GHz latest gen Core i7. Kaby even slightly lower than Skylake.

 

Nowadays it's all about power efficiency, which is nice in a way but when it's too prioritized the overall performance will suffer...


But then again, "power users" can probably be counted by fingers

 

If you happened to look at my bench thread you can see I ran Cinebench R15 Basically identical

http://forum.notebookreview.com/thr...n-dell-inspiron-13-7000.809595/#post-10612645

Here is a video of the run



I'm not making anything up with my benchmarks, the scores don't lie, these aren't suicide runs from a cold boot on the inspiron,

 

Yeah, Normally shouldnt the benchmark scores lie. But I don't know what Dell doing with their laptops. This below is from Hewlet-Packard with a +6 years old lady Core i7 2860QM who as well is in front vs. your i7-8550U CBR15 - 540cb score. Sandy Bridge is even a year older than Ivy Bridge.

http://hwbot.org/submission/3605072_deadthings_cinebench___r15_core_i7_2860qm_543_cb

 

3 Points? you are going to argue over 3 points? That isn't something i would think would ever be noticeable, but isn't than fun a 5 year old 45W QC beat my 15W Quadcore by 3 whole points i'll try not to loose any sleep over this. What difference does it make whether or not we know what dell is "doing" with their laptops. I took 2 laptops and ran the same tests on them, didn't handicap the older system at all, the M4700 didn't throttle at all. As a matter of fact it will do 4 cores loaded at 3.5ghz all day. I still can't see what you are arguing about. My point all along is that you can get near 45W quadcore performance from the new 15W chips and i'm not the only person with benchmarks to prove it.

 

You've been around longer than me, so you should know that NBR is the land of nitpickers and technically correct perfectionists. This is pretty much the only forum where I have to use asterisks and fine print every time I make a generalized statement, cuz otherwise someone's gonna call me out* on a technicality.

* Not every time, of course. Just most of the time. But see what I have to do?

 

Sandy Bridge was more intended as an example that newest 8th gen ULV processors don't give an advantage in load with more than 1 core. But still, this 5 years old i7-3840QM will run around 33% faster in multithread load. Say near 45W quadcore performance from the new 15W chips should be taken with a pinch of salt. 33% more ain't near. Not even close.


See also +4 year old medium powered Hotwell i7-4810mq With some love will the older 4810mq throw 7700hq down the hill in benchmarks.

 

And, that, boys and girls, is why things kind of suck in notebook land. When the bar gets lowered and only a few people even notice, they (OEM/ODM) can get away with murder. If power users were a majority they would draw back a bloody stump every time they tried to hand us a castrated piece of crap.

 

These are ULV CPU's meant for light carry and long battery life, not high performance.

Wrong thread for that argument

That's why I don't understand the need to compare the performance of the ULV CPU against a 45w CPU, it's actually showing me to avoid the top end ULV CPU's and get the lower performance one's so the design can fit more battery life.

The comparison is proving to be a negative indicator for the application, I wouldn't want to buy it

@Silvr6 - what's the battery size and duration / life like? Better or worse than the previous generation under battery load testing? Not idle testing

 

Well, sort of. Seems like they could still do better with that also. I don't want to let them off that easy. Why not have a octacore ULV? My ZTE Blade X Max has an octacore processor. They would not have to be clocked as high to keep up.

 

I've heard comments that there wasn't an advantage for more than 4 cores in phones / tablets for IOS / Android, so they backed off on 8 core, went back to 4 core, and now added 2 fast cores with 4 slow power saving cores.

The software needs to take advantage of the higher core count, just like in desktop OS's.

 

Mine seems pretty snappy, but it's just a phone and that's all I need it to be. I got it mainly because it has a large 1080p screen and was inexpensive. But, I think a ULV octacore CPU in a thin and light notebook would be a good idea. They could have 8 cores/16 threads clocked at like 2.8-3.2GHz turbo and probably be a lot better than a quad CPU. Especially so for business applications like content creation and CAD. Probably would need to be 35-45W though.

 

Yeah, but again it would draw more power, maybe requiring increased cooling taking room away for other things, like larger batteries, and increasing size / weight.

Although as software improves use of more cores on desktop's now that they have a lot more cores, that could trickle down into tablet / phone / ULV laptop designs.

The power savings stuff doesn't help for high performance hardware, but it's everything in a light carry long battery life device.

People already complain about phones / tablets not making it 24 hours, for full day use without needing a top off charge, it's really a drag when your phone loses power.

I plug my mobile devices in all the time I'm able.

 

This is the future for thin and flimsy. Intel should really change direction. Rather look towards high performance PGA / LGA chips for laptops. Aka for real laptops... With high performance.

Windows 10 on Snapdragon SoCs Coming Soon Says Microsoft – Battery Is Going to Be a Huge Selling Point


Windows 10 Mobile might not receive the development love that it rightfully deserves but it looks like there is a silver lining in all of this. Microsoft has been working at a frantic pace to make sure that its Windows 10 operating system can work fluidly on Snapdragon chipsets. According to the company, Windows 10 on Qualcomm chipsets are in the final stages of development and will be arriving soon, with a lot of perks being laid out at the same time.

 

An 8 core CPU would probably not be that ideal, although based on how intel implemented the 4 cores it might be an option. You are looking at 2.4ghz across 4 cores to fit in the 15W envelope. Add 4 more cars and you'll probably lose even more clock speed.

As for the power draw, it drawing around 17W package power at full load which is 2.6ghz.

The i7 3840QM did 2.8ghz at 28W and 1.6ghz at 15W so process improvements are noticeable.

For general use (which is why i have the inspiron) it usually spends most of its time turboing up to 4ghz in single and dual core use when web browsing watching videos ect

Battery size on this inspiron is 38W/Hr so the battery life is not the best under full load a little over 2 hours