There is just too little Intel can do with their 14nm parts to counteract Renoir 4000 series. At least on the 15W TDP, they have nothing like the 4800U, that is 8 cores and 16 threads.
For a full three seconds, then catastrophic thermal runaway, then a melted hole in the laptop where the CPU used to sit.
On a more serious note, I'd guess 4.9. 5+ sounds too high even for short-term single-core turbo on a 6c 15W SKU. I'd guess this is the high-end i7 of this generation like the i7-8650U and 8655U to the "mainstream" 8550U and 8555U in the KBL-R and WL generations, where the former two were nearly only seen in business notebooks (likely due to vPRO support).
I'd be very curious to see the PL2 rating on these CPU's. I imagine they're 15W at something like 1.8-2.2GHz, but at an all core turbo north of 4GHz they probably pull about 130W or more.
4GHz with 6 core is around 65W even on desktop chips so you'd have to push it to get to 130+ watts. That's desktop territory but interesting to see what AMD manage at 7nm/15W.
Actually 6 cores at 3.7GHz is 95W. That is easy to know since that is the 9600K's TDP which has a 3.7GHZ base clock and Intel states TDP off of base clock only.
That doesn't mean that it can't turbo higher and stay within TDP, just that that is the guaranteed spec that even the worst binned chip of that SKU will maintain within TDP. Most Intel chips outside of the very highest end ones turbo even when strictly limited to TDP, just not by much. Base clocks are also set to differentiate products, after all. 65W chips are a more reliable metric of power draw from TDP as they have less headroom to begin with, and the i7-9700 does 3GHz across 8 cores at 65W. The i5-9600 does just 3.1 at 65W, but it's unlikely it is a bin that much worse than the 9700 (to need 33% more power per core than the i7 for a similar clock speed), so the base clock could likely be higher if that didn't have the potential to cannibalize i7 sales.
If the chassis is cold after a long idle period or if shortly after power on, I think it could sustain high frequencies as it already does. But yeah, after a short while it will fall like a rock. This are the limitation of 14nm...and that is why AMD can do what it does so easily, thanks to TSMC magic.
You will see in a future notebookcheck review of 4800U that it can sustain its turbo quite well. It scores the same as 9880H, which is freaking amazing for a 15W CPU.
heh, so it is based on 14nm, the way intel has named their cpus is very confusing as to which is what. still, no 10nm based products can be found where i am though.
Hi scineram, That has not happen recently but for many years. In general, is it because mobile CPUs tweaked for running at low power? So is those tweaks what make them expensive?
While I understand that Intel is still trying to hold the fort with whatever they have, pushing clockspeed is no longer sustainable. I am sure with a boost clock of 4.7Ghz and above will push the power requirement over 150W. We just have to look at the current i5 9600K to get an idea of the power requirement. Its 14nm, so I don't think they can miraculously improve efficiency. They have already squeezed out as much as they can on 14nm.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
22 Comments
Back to Article
yeeeeman - Friday, March 13, 2020 - link
Let me guess, 5Ghz+ boost frequency?yeeeeman - Friday, March 13, 2020 - link
There is just too little Intel can do with their 14nm parts to counteract Renoir 4000 series. At least on the 15W TDP, they have nothing like the 4800U, that is 8 cores and 16 threads.PeachNCream - Friday, March 13, 2020 - link
11.3GHz, but only for a couple of microseconds. After that it reverts to its base clock of 233MHz until you power cycle the system.Valantar - Friday, March 13, 2020 - link
For a full three seconds, then catastrophic thermal runaway, then a melted hole in the laptop where the CPU used to sit.On a more serious note, I'd guess 4.9. 5+ sounds too high even for short-term single-core turbo on a 6c 15W SKU. I'd guess this is the high-end i7 of this generation like the i7-8650U and 8655U to the "mainstream" 8550U and 8555U in the KBL-R and WL generations, where the former two were nearly only seen in business notebooks (likely due to vPRO support).
Drkrieger01 - Friday, March 13, 2020 - link
I'd be very curious to see the PL2 rating on these CPU's. I imagine they're 15W at something like 1.8-2.2GHz, but at an all core turbo north of 4GHz they probably pull about 130W or more.smilingcrow - Friday, March 13, 2020 - link
4GHz with 6 core is around 65W even on desktop chips so you'd have to push it to get to 130+ watts.That's desktop territory but interesting to see what AMD manage at 7nm/15W.
schujj07 - Friday, March 13, 2020 - link
Actually 6 cores at 3.7GHz is 95W. That is easy to know since that is the 9600K's TDP which has a 3.7GHZ base clock and Intel states TDP off of base clock only.Valantar - Friday, March 13, 2020 - link
That doesn't mean that it can't turbo higher and stay within TDP, just that that is the guaranteed spec that even the worst binned chip of that SKU will maintain within TDP. Most Intel chips outside of the very highest end ones turbo even when strictly limited to TDP, just not by much. Base clocks are also set to differentiate products, after all. 65W chips are a more reliable metric of power draw from TDP as they have less headroom to begin with, and the i7-9700 does 3GHz across 8 cores at 65W. The i5-9600 does just 3.1 at 65W, but it's unlikely it is a bin that much worse than the 9700 (to need 33% more power per core than the i7 for a similar clock speed), so the base clock could likely be higher if that didn't have the potential to cannibalize i7 sales.yeeeeman - Friday, March 13, 2020 - link
If the chassis is cold after a long idle period or if shortly after power on, I think it could sustain high frequencies as it already does. But yeah, after a short while it will fall like a rock. This are the limitation of 14nm...and that is why AMD can do what it does so easily, thanks to TSMC magic.yeeeeman - Friday, March 13, 2020 - link
You will see in a future notebookcheck review of 4800U that it can sustain its turbo quite well. It scores the same as 9880H, which is freaking amazing for a 15W CPU.yeeeeman - Friday, March 13, 2020 - link
Links to the review: https://www.reddit.com/r/Amd/comments/fhq98i/renoi...Beany2013 - Friday, March 13, 2020 - link
Blimey, those are some surprisingly impressive metrics. The long-term multicore cinebench stuff especially.Valantar - Friday, March 13, 2020 - link
Have to love an AT comment that links to a Reddit post about a leak that was originally posted on the AT forums.JayNor - Friday, March 13, 2020 - link
another article pointed out that the only difference is added vPro, base on its leak in a ThinkPad add.https://www.notebookcheck.net/Intel-Core-i7-10810U...
Qasar - Friday, March 13, 2020 - link
heh, i STILL cant buy ANY 10nm based intel notebooks here, and they are releasing yet another one ???Valantar - Saturday, March 14, 2020 - link
Comet Lake is 14nm.Qasar - Saturday, March 14, 2020 - link
heh, so it is based on 14nm, the way intel has named their cpus is very confusing as to which is what.still, no 10nm based products can be found where i am though.
vozmem - Saturday, March 14, 2020 - link
Why are Intel mobile CPUs so expensive?scineram - Saturday, March 14, 2020 - link
Supply and demand.vozmem - Sunday, March 15, 2020 - link
Hi scineram,That has not happen recently but for many years.
In general, is it because mobile CPUs tweaked for running at low power? So is those tweaks what make them expensive?
watzupken - Monday, March 16, 2020 - link
While I understand that Intel is still trying to hold the fort with whatever they have, pushing clockspeed is no longer sustainable. I am sure with a boost clock of 4.7Ghz and above will push the power requirement over 150W. We just have to look at the current i5 9600K to get an idea of the power requirement. Its 14nm, so I don't think they can miraculously improve efficiency. They have already squeezed out as much as they can on 14nm.xprojected - Tuesday, March 17, 2020 - link
FWIW, Intel posted an update to the PCN, and removed the i7-10810U from the list.