Discussion Intel current and future Lakes & Rapids thread

[uzzi38]

I find it amazing that no one is talking about the cooling efficiency of the reference Intel system. Surely, that counts too?

Look literally 3 posts up.

 

[ondma]

35W Zen is close in performance to 45W TGL and you're unable to see a big efficiency difference there?! You do understand we're talking about a ~30% delta in power?


It has already been mentioned in this thread that the 65W figures are NOT representative for TGL performance on the Intel reference system, since the cooling could not sustain continuous 65W package power.


We're not discussing performance/TDP as defined by both companies for their desktop lines, we're discussing performance/package power consumption expressed in Watts. Take a look again at the CPU Power Scaling graph from Hardware Unboxed, the X axis is clearly defined as such. Both Intel and AMD products have sensor data that accurately reflect package power consumption in real time, accurate enough for this type of comparison anyway.

Yea, but if you take the absolute difference, 10 watts, compared to the power consumption of even a midrange gaming laptop gpu (say 100 watts) it is not a big percentage difference in total power consumption. I just think the hysteria of the "power hungry" Tiger Lake is a bit overblown. Now for Rocket Lake, it isnt.

 

[lobz]

Yea, but if you take the absolute difference, 10 watts, compared to the power consumption of even a midrange gaming laptop gpu (say 100 watts) it is not a big percentage difference in total power consumption. I just think the hysteria of the "power hungry" Tiger Lake is a bit overblown. Now for Rocket Lake, it isnt.

Then the same logic applies for 45W vs 55W, then the same for 55W vs 65W, then the same for 65W vs 75W, and since the total system consumption would always be even higher and higher, the 10W would count even less and less percentage wise. The more you buy, the more you save?

I think that the difference between 35W and 45W is not groundbreaking, but still significant.

 

[dmens]

Normalize the power consumption with performance and you will realize just how awful Intel is doing.

 

[LightningZ71]

Intel® Pentium® Processor 6805 (4M Cache, up to 3.00 GHz) Product Specifications

Intel® Pentium® Processor 6805 (4M Cache, up to 3.00 GHz) quick reference guide including specifications, features, pricing, compatibility, design documentation, ordering codes, spec codes and more.

ark.intel.com

Don't think it ever got any publicity, but Intel did end up releasing an Ice Lake Pentium in Q4 2020. I wonder what the actual cost difference is between it and the Tiger Lake Pentium that was also released in Q4.

Looks like the Pentium gold 7000 series is significantly better than the 6000, especially in the iGPU with both much higher clocks and more EUs. In an interesting segmentation move, the Pentium gold 7000 doesn't show AVX-512 support, but seems to have all the other tiger lake add ons. It doesn't seem like much of a penalty from the i3 line if you don't need 512...

 

[Gideon]

Look at the cache stacking AMD does, 15% gaming uplift overall:

Computex 2021: AMD's Keynote, a Live Blog (10pm ET)

www.anandtech.com

So much for Intel retaking the gaming crown

 

[Asterox]

Look at the cache stacking AMD does, 15% gaming uplift overall:

Computex 2021: AMD's Keynote, a Live Blog (10pm ET)

www.anandtech.com

So much for Intel retaking the gaming crown

If even Zen 3+(in AM4 or AM5 socket) will use 3D V-CACHE, well Alder Lake gaming advantage will vanish like morning fog.

 

[maddie]

If even Zen 3+(in AM4 or AM5 socket) will use 3D V-CACHE, well Alder Lake gaming advantage will vanish like morning fog.

Didn't she say "coming later this year" and Zen4 next year?

 

[eek2121]

Didn't she say "coming later this year" and Zen4 next year?

Correct. It sounds like Warhol ships with it. Intel is going to have to rethink life.

 

[Asterox]

Didn't she say "coming later this year" and Zen4 next year?

She said, we will be ready to start a production of high end products with 3D chiplets by the end of this year.

 

[maddie]

She said, we will be ready to start a production of high end products with 3D chiplets by the end of this year.

I don't think Zen4 will be ready to start production late this year. Do you? If not, what's left?

Also' they have apparently redesigned the Zen3 chiplet to allow connections with the cache die. Not a trivial matter.

 

[HurleyBird]

Also' they have apparently redesigned the Zen3 chiplet to allow connections with the cache die. Not a trivial matter.

Sure about that?

 

[randomhero]

Sure about that?

Exactly.

 

[majord]

someone's confused

 

[misuspita]

Sure about that?

She shown a 5900x on stage with the new cache. I think she was not kidding.

 

[Det0x]

Sure about that?

Hmm could the B2 stepping be in preparation for this ?

AMD confirms B2 stepping for Ryzen 5000 series brings no functionality or performance improvements - VideoCardz.com

AMD Ryzen 5000 “Vermeer” on B2 stepping is only to improve production and availability Yesterday it was reported that AMD is developing a new stepping of its Vermeer processors. Meanwhile, the Polish website Benchmark.pl has reached out to AMD for a comment on this matter. The site received an...

videocardz.com

 

[coercitiv]

Somebody at Intel is having a bad Q2 2021, and it's only the 1st of June.

 

[IntelUser2000]

Correct. It sounds like Warhol ships with it. Intel is going to have to rethink life.

Wow, if that's Warhol then it's a very credible answer to Alder Lake.

 

[Asterox]

One question for Intel, where is/when it arrives any shorter meaningful Alder Lake preview?

Page 416 - Discussion - Intel current and future Lakes & Rapids thread

Page 416 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

AMD showed very strong future cards, or what can we get with 3D V-CACHE even on AM4 Zen3 CPU-s.

Hard times for Intel, in situation where green CPU combination can set a heavy knockout at Alder Lake in new Intel 1700 socket.

 

[Zucker2k]

Look literally 3 posts up.

What has that got to do with the ACTUAL design/cooling capacity of the Intel reference system? Are we going to pretend boosting algos are no longer tied to thermals?

 

[uzzi38]

What has that got to do with the ACTUAL design/cooling capacity of the Intel reference system? Are we going to pretend boosting algos are no longer tied to thermals?

I did say look 3 posts up, right?

It has already been mentioned in this thread that the 65W figures are NOT representative for TGL performance on the Intel reference system, since the cooling could not sustain continuous 65W package power.

We already know perfectly well the reference numbers of Intel's laptop are not indicative of 65W sustained performance. Thankfully, HUB's chart does not use Intel's reference platform, and as you can see there is still a sizable difference in efficiency.

 

[coercitiv]

What has that got to do with the ACTUAL design/cooling capacity of the Intel reference system? Are we going to pretend boosting algos are no longer tied to thermals?

Sigh, let's talk about TGL reference system cooling. We have detailed info on this from Tom's Hardware:

To check stability over a longer duration, we ran Cinbench R23 for 20 runs. The cooling, which was exceptionally loud during all of the tests (and sometimes while the sample system was doing absolutely nothing) kept it stable.

It started at a high of 11,846.31 while largely settling in the 11,600 range. During the Cinebench stress test, the CPU ran at an average of 3.5 GHz and an average temperature of 85.77 degrees Celsius (186.39 degrees Fahrenheit). While the chart looks largely stable, the monitoring tool HWinfo reported that the CPU was being thermally throttled for the majority of the test. This is the downside of putting a high-wattage processor in a slim system, and also explains the constant fan noise.

The laptop has three fans, while even most gaming laptops stick to two larger ones. That may explain the decibels. But what's also fascinating is that the motherboard in the reference platform has been placed effectively upside down. This means that we can't see the full cooler, including the heat pipes. That would require far more disassembly.

To complicate things, this reference design is meant to represent a new "thin enthusiast" sector for Intel, which meant we couldn't see how the Core i9-11980HK will perform at its best, in a thicker laptop with more elaborate cooling.

Intel was as arrogant as ever, and decided to invent this "thin enthusiast" segment for mobile, which is just an euphemism for "thin and loud". As you said before in this thread, Intel really needs a Common Sense department to run reality checks on every idea Marketing comes up with.

That being said, the cooling solution on this thin reference system was as powerful as possible given the form factor (3 fans... sigh), and also very aggressive in behavior according to Tom's reporting. I'm pretty sure it had no problem ensuring optimal performance in the 35-45W TDP range for "CPU only" benchmarks.

 

[Zucker2k]

I did say look 3 posts up, right?



We already know perfectly well the reference numbers of Intel's laptop are not indicative of 65W sustained performance. Thankfully, HUB's chart does not use Intel's reference platform, and as you can see there is still a sizable difference in efficiency.

Let's see then:

Ryzen 5800H & 5900HX chassis:

The test system we’re using for benchmarking the R9 5900HX is the XMG Apex 17, an upcoming gaming laptop based on a Clevo chassis that’s almost identical to the Apex 17 we used previously to test both the Ryzen 7 5800H and RTX 3060 laptop GPU, making this a higher-end configuration of the same system.

11800H

The test system for today’s review is not an Intel reference platform but a production laptop from Gigabyte, the new Aero 15 OLED. This is more of a creator/productivity focused system than Gigabyte’s Aorus line-up, so the Aero features a gorgeous 15.6-inch 4K OLED panel which is amazing for content creation and viewing.

Do you discern any advantage here for any of the platforms? And in gaming, when the DGPUs kick in? Again, see the LTT video I linked for a side-by-side visual comparison, instead of being dismissive outright. I know a few of you who bring this up when the shoe's on the other foot.

 

[Zucker2k]

I'm pretty sure it had no problem ensuring optimal performance in the 35-45W TDP range for "CPU only" benchmarks.

And this is what I find puzzling. How much heat is the cooling dissipating in gaming when it can't sustain 65w handling only the CPU? Less than 65w?

 

[coercitiv]

And this is what I find puzzling. How much heat is the cooling dissipating in gaming when it can't sustain 65w handling only the CPU? Less than 65w?

Hehe, it doesn't work quite like that. Here's the big catch about heatsinks and coolers in general: the amount of heat they can displace increases linearly with the temperature delta between the heatsink and the air passing through the fins. The hotter the heasink is allowed to get, the more heat it able to dissipate.

Let's start with a familiar example: moving from Intel 8th gen CPU with paste thermal interface material, to 9th gen with solder, and last to 10th gen with lowered thickness interface. While using the same cooler with the same fan & speed, simply improving thermal transfer between the CPU die and the heatsink will result in a better performing cooler (heatsink gets hotter, more heat gets dissipated). The cooler is the same, yet heat dissipation has improved due to lower thermal resistance between CPU die and heatspreader.

Back to the mobile scenario, when attempting to prevent the CPU from throttling, the laptop cooler will use only part of it's potential. For example some heatpipes may not make (direct) contact with the CPU, but with the GPU. The same may apply to fan assemblies. On top of that, when using both GPU and CPU, the heatsink will capture more heat and potentially rise higher in temperature (in all areas). This can easily result in more heat being dissipated during combined workloads, as long as we're not talking about 65W package power for CPU + xy W package power for the GPU, since obviously adding GPU power will introduce some additional limitation to max CPU power. You may find that 45W + 45W is possible, or maybe something like 35W + 60W.

It all depends on heatpipe configuration and where the overall bottlenecks of the cooler lie: thermal resistance close to die (meaning bad paste), heatpipes, radiator, airflow, noise profile.

PS: one of my laptops had a 45W CPU, a 50W dGPU and a dual fan config with 2 shared heatpipes. The laptop was able to keep the CPU from throttling while runing Prime95 in long sessions, but temps were close to the limit (PL2 was ~56W, boost window was 1 minute or less). The same system was able to run combined gaming workloads which could easily have reached 70W+ of power use.