News Intel 10nm Roadmap Leaked: Alder Lake May Land This Year

JayNor

Prominent
May 31, 2019
177
30
610
0
One of the upgrades described for Gracemont is that it adds vector support. There were Atom cores in Xeon Phi which contained AVX512 operations, including the VNNI operations.

The hyperthreaded server cores share avx512 units between threads while this version chip would not share avx512 units.

There are issues with sharing avx512 units between threads that result in programming recommendations to disable hyperthreading for hpc processing, so this is potentially an even bigger win for hpc. Double the avx512 execution units and not having to disable hyperthreading.

Sounds like a good trade-off for a core that is probably less than half the size of the big cores. ( about one-fourth the size in Lakefield).
 

PCWarrior

Distinguished
May 20, 2013
102
29
18,610
0
One thing that is often overlooked during the 14nm vs 10nm frequency comparison is that the frequencies we know now about the 10nm cpus are only those of the 15W U series processors. It is a mistake to project anything from them. There are three things to note.
(i) U series processors’ frequencies have always been considerably lower than desktop ones. Typically around 500MHz lower (+/-200MHz) with the difference often exceeding 1GHz when you mildly OC your desktop cpu (‘stock volts’, air cooling).
(ii) Historically whenever a new generation was introduced, the single core speed of the mainstream flagship in the mobile platform was between 3.0GHz-3.5GHz. If the Icelake U cpus came just after the (already mature) 14nm+ Kabylake U cpus we would actually be talking about the incredible jump in single core frequency, going from 3.5GHz of the i7 7500U to the 3.9GHz of the i7 1065G7. We would be talking about how great the 10nm process is compared to the 14nm one.
(iii) Just because Intel spent more time on the 14nm keep optimising for frequency, it doesn’t mean there is something wrong on the cycle of the 10nm process in terms of the achievable frequency. Frequency improvements will be achieved as the process matures further, the same way it happened with the 22nm and the various iterations of the 14nm processes in the past. I doubt that the yield problems we have heard about the 10nm process are frequency-related problems. Intel can probably already produce Icelake desktop cpus with pretty decent clocks (like 4.5Ghz or so). Instead the yield problems seem to have been “misprinting” problems. In other words, if they went ahead to produce larger dies of Icelake cpus (like 8core or 10core ones) they would be throwing a lot of them away because they are straight-out faulty, not because they can’t hit a certain high frequency.
(i) The IPC gains of Willow Cove compared to Skylake should be massive - like around 30%. A 10% loss on frequency is therefore more than compensated.
 

Hardware Geek

BANNED
Nov 11, 2019
3
1
15
0
One thing that is often overlooked during the 14nm vs 10nm frequency comparison is that the frequencies we know now about the 10nm cpus are only those of the 15W U series processors. It is a mistake to project anything from them. There are three things to note.
(i) U series processors’ frequencies have always been considerably lower than desktop ones. Typically around 500MHz lower (+/-200MHz) with the difference often exceeding 1GHz when you mildly OC your desktop cpu (‘stock volts’, air cooling).
(ii) Historically whenever a new generation was introduced, the single core speed of the mainstream flagship in the mobile platform was between 3.0GHz-3.5GHz. If the Icelake U cpus came just after the (already mature) 14nm+ Kabylake U cpus we would actually be talking about the incredible jump in single core frequency, going from 3.5GHz of the i7 7500U to the 3.9GHz of the i7 1065G7. We would be talking about how great the 10nm process is compared to the 14nm one.
(iii) Just because Intel spent more time on the 14nm keep optimising for frequency, it doesn’t mean there is something wrong on the cycle of the 10nm process in terms of the achievable frequency. Frequency improvements will be achieved as the process matures further, the same way it happened with the 22nm and the various iterations of the 14nm processes in the past. I doubt that the yield problems we have heard about the 10nm process are frequency-related problems. Intel can probably already produce Icelake desktop cpus with pretty decent clocks (like 4.5Ghz or so). Instead the yield problems seem to have been “misprinting” problems. In other words, if they went ahead to produce larger dies of Icelake cpus (like 8core or 10core ones) they would be throwing a lot of them away because they are straight-out faulty, not because they can’t hit a certain high frequency.
(i) The IPC gains of Willow Cove compared to Skylake should be massive - like around 30%. A 10% loss on frequency is therefore more than compensated.
The problem being the average consumer doesn't have a clue. They simply see higher frequencies and assume it's better. I had a friend call me because he wants a computer upgrade and he seriously couldn't comprehend that a new ryzen could possibly be faster than an older gen core processor. "But it's 200mhz faster so it has to be better". The second I mentioned IPC I could already tell he didn't have a clue how that could possibly offer better performance. I'd assume most people on here understand the nuances better than most typical consumers.
 
Reactions: Soaptrail

Soaptrail

Honorable
Jan 12, 2015
168
42
10,620
1
I am not holding my breathe for a second Intel refresh/launch this year, even without the supply chain issues hampered by coronavirus it was not realistic to see two launches less than 9 months apart.
 
Reactions: bit_user

watzupken

Proper
Mar 16, 2020
176
45
110
0
My gut feel is that Intel's 10nm is not going to give them an advantage in the desktop space. Looking at the rumored power draw from an earlier article here about Alder Lake S, it doesn't look that great for a new fab where we typically see lower TDP requirements. 80W and 125W TDP for an 8 "big cores" and 8 "small cores" is nothing to write home if the rumor is true.
 

PCWarrior

Distinguished
May 20, 2013
102
29
18,610
0
The problem being the average consumer doesn't have a clue. They simply see higher frequencies and assume it's better. I had a friend call me because he wants a computer upgrade and he seriously couldn't comprehend that a new ryzen could possibly be faster than an older gen core processor. "But it's 200mhz faster so it has to be better". The second I mentioned IPC I could already tell he didn't have a clue how that could possibly offer better performance. I'd assume most people on here understand the nuances better than most typical consumers.
Well IPC is actually much more nuanced. You see IPC is a per workload metric and what is quoted as an IPC increase is the average uplift across several workloads. In workload A it can be 10% faster, in workload B 20% faster and in C 30% faster, the average across all three is 20%. Consequently, it largely depends on your test suite, which tests are included in that suite and how many. It is similar to saying it is X% faster in gaming. But in fact it depends which games you compare and how many you include in your test suite. And that assuming you won’t be running into a gpu bottleneck scenario.

IPC increase is tricky for another reason too. You see, one way that manufactures use to increase IPC is by making the cores wider/larger, so that they increase parallelism within a core. This benefits applications such as Cinema4D that would also scale through parallel processing across more cores, but it does nothing for workloads that can’t scale more with parallel processing because of algorithmic dependencies or latencies. For the latter you need a more computationally efficient architecture, a more optimised mapping of the algorithms to the core logic and a topology that allows for smaller inter-core latencies. For the most part AMD’s IPC uplift stems from in-core parallelism as well as attempts to limit the originally large inter-die or inter-CCX latencies. The first is easy and the second is just fixing an originally flawed design. So no, in this context of IPC vs frequency, for a more apples to apples comparison, I wouldn’t compare Intel to AMD. I would compare Intel (newer architecture) Vs Intel (older architecture). And I would do the same for AMD.
 

bit_user

Splendid
Ambassador
One thing that is often overlooked during the 14nm vs 10nm frequency comparison is that the frequencies we know now about the 10nm cpus are only those of the 15W U series processors. It is a mistake to project anything from them.
Perhaps not, but don't forget that Intel Launched 14 nm Comet Lake for the laptop performance segment after 10 nm Ice Lake. This is basically proof that their current 10 nm node cannot deliver the clock speeds to compete at the top of their product stack.

That said, any desktop 10 nm CPU will probably use their 10 nm++ process node (with Ice Lake using their 10 nm+ version).

Instead the yield problems seem to have been “misprinting” problems. In other words, if they went ahead to produce larger dies of Icelake cpus (like 8core or 10core ones) they would be throwing a lot of them away because they are straight-out faulty, not because they can’t hit a certain high frequency.
Uh, that is yield.

And the Comet Lake mobile CPUs are also 4-core, so the yield explanation doesn't hold water for explaining why they needed to revert to 14 nm for their current mobile performance models.

(i) The IPC gains of Willow Cove compared to Skylake should be massive - like around 30%. A 10% loss on frequency is therefore more than compensated.
I'll believe it when I see it.
 

bit_user

Splendid
Ambassador
"But it's 200mhz faster so it has to be better". The second I mentioned IPC I could already tell he didn't have a clue how that could possibly offer better performance.
Though it probably won't help with someone who doesn't understand the concept of IPC, the analogy of horsepower works quite well. Just because one engine produces more torque or revs to a higher RPM doesn't necessarily make it more powerful than another.
 

bit_user

Splendid
Ambassador
Alder Lake is the codename for Intel's upcoming 10nm desktop CPUs that are rumored to adopt the hybrid architecture like Lakefield. Rumors point to a 16-core configuration comprised of eight big cores and eight small cores.
For a desktop CPU? That's just mad!

Nah, I don't buy it. I just don't see the value of Big.Little, in the desktop platform. It'd make more sense to replace the 8 little cores with 2 more big cores. You can save enough power simply by cutting back on clock speeds, or even powering off cache blocks, if needed.
 

ASK THE COMMUNITY

TRENDING THREADS