By: anon2 (anon.delete@this.anon.com), January 4, 2022 6:26 pm
Room: Moderated Discussions
Intel naturally claims Adler Lake is faster than M1,
https://www.intc.com/news-events/press-releases/detail/1517/intel-engineers-fastest-mobile-processor-ever-with-12th-gen
More surprisingly they claim better energy efficiency at equivalent performance as peak M1
https://mobile.twitter.com/witeken/status/1478441037416828932
Let's take this with a grain of salt and assume they carefully pick just the right benchmarks to use here so the reality may not not quite be so good. But if efficiency is significantly worse when they claimed it is significantly better that seems like it would cause them a lot of problems when people actually test it. If we give Intel some credit (I'm sure people will object to that!) then let's assume efficiency is roughly the same, maybe slightly worse but not a lot.
This would be quite a surprising result to me given that M1 is on N5 and ADL is on Intel 7 (previously 10nm), which common wisdom previously said was roughly equivalent to TSMC N7. I'm of the opinion that silicon performance is the first order difference between Intel/Apple/AMD cores so it's surprising if Intel has jumped a node with the Adler Lake core. Is today's Intel 7 a significant improvement over yesterday's 10nm? Or is it still equivalent to N7?
This isn't a bash at Apple or TSMC before anybody jumps down my throat. M1 Firestorm core will probably be over 1.5 years old by the time of non-paper launch, and N3 still seems about a year ahead of Intel 4 so I expect Apple to take efficiency lead again early next year if it even has lost the lead by now. So Apple bashing I'm not interested in, was more wondering about Intel 7.
https://www.intc.com/news-events/press-releases/detail/1517/intel-engineers-fastest-mobile-processor-ever-with-12th-gen
More surprisingly they claim better energy efficiency at equivalent performance as peak M1
https://mobile.twitter.com/witeken/status/1478441037416828932
Let's take this with a grain of salt and assume they carefully pick just the right benchmarks to use here so the reality may not not quite be so good. But if efficiency is significantly worse when they claimed it is significantly better that seems like it would cause them a lot of problems when people actually test it. If we give Intel some credit (I'm sure people will object to that!) then let's assume efficiency is roughly the same, maybe slightly worse but not a lot.
This would be quite a surprising result to me given that M1 is on N5 and ADL is on Intel 7 (previously 10nm), which common wisdom previously said was roughly equivalent to TSMC N7. I'm of the opinion that silicon performance is the first order difference between Intel/Apple/AMD cores so it's surprising if Intel has jumped a node with the Adler Lake core. Is today's Intel 7 a significant improvement over yesterday's 10nm? Or is it still equivalent to N7?
This isn't a bash at Apple or TSMC before anybody jumps down my throat. M1 Firestorm core will probably be over 1.5 years old by the time of non-paper launch, and N3 still seems about a year ahead of Intel 4 so I expect Apple to take efficiency lead again early next year if it even has lost the lead by now. So Apple bashing I'm not interested in, was more wondering about Intel 7.
Topic | Posted By | Date |
---|---|---|
How good is Intel 7? | anon2 | 2022/01/04 06:26 PM |
How good is Intel 7? | G | 2022/01/04 07:42 PM |
How good is Intel 7? | nope | 2022/01/05 01:27 PM |
How good is Intel 7? | Brendan | 2022/01/04 07:54 PM |
How good is Intel 7? | anon2 | 2022/01/04 08:13 PM |
How good is Intel 7? | Brendan | 2022/01/04 10:44 PM |
How good is Intel 7? | anon2 | 2022/01/05 04:52 AM |
How good is Intel 7? | Brendan | 2022/01/05 09:26 PM |
How good is Intel 7? | --- | 2022/01/05 09:42 PM |
HyperThreading and Apple M1 | Brett | 2022/01/06 12:22 PM |
HyperThreading and Apple M1 | Aar | 2022/01/06 12:46 PM |
HyperThreading and Apple M1 | anonymou5 | 2022/01/06 02:24 PM |
HyperThreading and Apple M1 | --- | 2022/01/06 04:00 PM |
HyperThreading and Apple M1 | Brett | 2022/01/08 11:23 AM |
HyperThreading and Apple M1 | Freddie | 2022/01/08 01:43 PM |
you have misunderstood what working set means | Heikki Kultala | 2022/01/08 03:08 PM |
you have misunderstood what working set means | anon2 | 2022/01/08 05:43 PM |
HyperThreading and Apple M1 | Kevin G | 2022/01/08 06:56 PM |
HyperThreading and Apple M1 | Jukka Larja | 2022/01/09 02:23 AM |
HyperThreading and Apple M1 | Anon | 2022/01/09 04:48 AM |
HyperThreading and Apple M1 | Jukka Larja | 2022/01/09 09:28 AM |
HyperThreading and Apple M1 | Doug S | 2022/01/09 12:55 PM |
HyperThreading and Apple M1 | Aar | 2022/01/10 02:15 AM |
HyperThreading and Apple M1 | --- | 2022/01/10 12:48 PM |
A14 -> A15 | Mark Roulo | 2022/01/10 04:29 PM |
A14 -> A15 | --- | 2022/01/10 07:13 PM |
A14 -> A15 | Rayla | 2022/01/10 08:16 PM |
A14 -> A15 | --- | 2022/01/11 12:05 AM |
A14 -> A15 | Jörn Engel | 2022/01/11 08:49 AM |
A14 -> A15 | --- | 2022/01/11 09:40 AM |
A14 -> A15 | Chester | 2022/01/11 05:40 PM |
A14 -> A15 | --- | 2022/01/11 08:25 PM |
A14 -> A15 | Jörn Engel | 2022/01/11 08:44 PM |
A14 -> A15 | Chester | 2022/01/12 12:02 AM |
A14 -> A15 | Doug S | 2022/01/12 10:25 AM |
Memory Request Tagging | Chester | 2022/01/19 10:21 AM |
A14 -> A15 | Kevin G | 2022/01/12 10:33 AM |
A14 -> A15 | Kevin G | 2022/01/12 10:35 AM |
A14 -> A15 | Jukka Larja | 2022/01/11 05:37 AM |
A14 -> A15 | Linus Torvalds | 2022/01/11 10:53 AM |
A14 -> A15 | Doug S | 2022/01/11 01:28 PM |
A14 -> A15 | Linus Torvalds | 2022/01/11 04:08 PM |
A14 -> A15 | Foyle | 2022/01/12 02:28 AM |
A14 -> A15 | Yoav | 2022/01/11 09:29 PM |
A14 -> A15 | Groo | 2022/01/12 03:59 PM |
A14 -> A15 | Jukka Larja | 2022/01/13 05:33 AM |
A14 -> A15 | Groo | 2022/01/13 12:03 PM |
A14 -> A15 | Simon Farnsworth | 2022/01/13 12:59 PM |
A14 -> A15 | Jukka Larja | 2022/01/15 01:17 AM |
A14 -> A15 | Simon Farnsworth | 2022/01/17 03:26 AM |
A14 -> A15 | Jukka Larja | 2022/01/17 08:23 AM |
Crystalwell performance was great | David Kanter | 2022/01/11 09:27 PM |
Crystalwell performance was great | never_released | 2022/01/11 10:33 PM |
Crystalwell performance was great | Robert Williams | 2022/01/12 07:29 PM |
A14 -> A15 | Aar | 2022/01/11 01:24 AM |
"hyperthreading" is intel marketing name. Please use correct echnical term (NT) | Heikki Kultala | 2022/01/08 04:49 AM |
How good is Intel 7? | anon2 | 2022/01/06 02:03 AM |
How good is Intel 7? | Brendan | 2022/01/06 04:38 AM |
How good is Intel 7? | Adrian | 2022/01/06 07:41 AM |
How good is Intel 7? | Doug S | 2022/01/06 09:31 AM |
How good is Intel 7? | anon2 | 2022/01/06 09:40 PM |
How good is Intel 7? | Brendan | 2022/01/08 06:50 AM |
How good is Intel 7? | Anon | 2022/01/08 08:22 AM |
How good is Intel 7? | Doug S | 2022/01/08 10:35 AM |
How good is Intel 7? | Anne O. Nymous | 2022/01/09 06:28 AM |
How good is Intel 7? | rwessel | 2022/01/09 10:06 AM |
How good is Intel 7? | Doug S | 2022/01/09 12:58 PM |
How good is Intel 7? | Anne O. Nymous | 2022/01/10 12:00 AM |
How good is Intel 7? | Doug S | 2022/01/10 09:38 AM |
How good is Intel 7? | Anne O. Nymous | 2022/01/11 01:02 AM |
How good are e-cores? | Per Hesselgren | 2022/01/08 08:55 AM |
How good are e-cores? | Anon | 2022/01/08 10:38 AM |
How good is Intel 7? | Doug S | 2022/01/08 10:21 AM |
How good is Intel 7? | Brendan | 2022/01/08 10:09 PM |
How good is Intel 7? | Anon | 2022/01/09 04:53 AM |
How good is Intel 7? | Doug S | 2022/01/09 02:00 PM |
How good is Intel 7? | Brendan | 2022/01/09 08:12 PM |
How good is Intel 7? | Doug S | 2022/01/10 09:46 AM |
How good is Intel 7? | Brendan | 2022/01/10 07:21 PM |
How good is Intel 7? | Doug S | 2022/01/11 08:15 AM |
How good is Intel 7? | Brendan | 2022/01/11 12:04 PM |
How good is Intel 7? | Doug S | 2022/01/11 01:19 PM |
How good is Intel 7? | Brendan | 2022/01/11 08:25 PM |
How good is Intel 7? | Doug S | 2022/01/12 10:34 AM |
How good is Intel 7? | Brendan | 2022/01/12 05:40 PM |
How good is Intel 7? | blue | 2022/01/13 09:29 AM |
How good is Intel 7? | Brett | 2022/01/08 11:56 AM |
Apple's Google Search Revenue | me | 2022/01/08 04:11 PM |
How good is Intel 7? | anon2 | 2022/01/08 05:16 PM |
How good is Intel 7? | Brendan | 2022/01/09 02:37 AM |
How good is Intel 7? | anon2 | 2022/01/09 03:09 AM |
How good is Intel 7? | Brendan | 2022/01/09 09:23 AM |
How good is Intel 7? | anon2 | 2022/01/09 06:17 PM |
Apple service profit. | blue | 2022/01/08 08:10 PM |
How good is Intel 7? | anonymou5 | 2022/01/04 08:11 PM |
How good is Intel 7? | David Kanter | 2022/01/04 10:44 PM |
No footnotes? (NT) | Anon | 2022/01/04 11:19 PM |