A Geekbench score of the Intel Core i7-1065G7 'Ice Lake' running on the HP Spectre x360 convertible laptop is now available and seems to indicate very strong single-core performance that can rival processors much higher TDP such as the AMD Ryzen 9 3900X. Intel's 10nm could offer a significant advantage over AMD at least in the laptop segment with even more improved IPCs.
keep in mind it is only 3.88ghz if the desktop model could achieve higher clock speed this would a bit better than a 3900x, i wonder if the cpu could oc to 5ghz or the node will be limiting it
That is a massively clickbaity and really badly written article. I was expecting more from NBC...
Intel's TDP != AMD TDP.
Intel's is long-term sustained maximum heat generation at base clock, while AMD's is absolute peak load power consumption.
The Intel CPU turboing to 3.88GHz on a single core will be drawing far over 15W. AFAIK it allows around double TDP by default on single core peak turbo for a very short while (just enough for the duration of the geekbench single core benchmark).
The 3900X will be using 105W maximum on all core full load, though with only 1 core loaded will realistically be in the 40-50W range. It can, however, sustain that load over more than a few seconds or minutes.
Even given enough cooling, Intel's CPU wouldn't come anywhere near the 3900X in single core performance running any load for more than 5 minutes. Once the peak turbo timer runs out, it will return to stable turbo or even lower, close to stock, which is likely around 2 Ghz.
I really dislike these super short benchmarks, they don't say anything about actual performance. Who's gonna play a game for only 5 minutes?
I really couldn't get along with any of the dual core ULV Intel's even for normal (non-productivity) use, they are just too freaking slow. While these quad cores have made it a little better, their sustained performance vs battery consumption (note: I didn't say TDP) is not good enough compared to an old Sandy, Ivy or Haswell 35-45W socketed quad core to warrant upgrading.
That is a massively clickbaity and really badly written article. I was expecting more from NBC...
Intel's TDP != AMD TDP.
Intel's is long-term sustained maximum heat generation at base clock, while AMD's is absolute peak load power consumption.
The Intel CPU turboing to 3.88GHz on a single core will be drawing far over 15W. AFAIK it allows around double TDP by default on single core peak turbo for a very short while (just enough for the duration of the geekbench single core benchmark).
The 3900X will be using 105W maximum on all core full load, though with only 1 core loaded will realistically be in the 40-50W range. It can, however, sustain that load over more than a few seconds or minutes.
Even given enough cooling, Intel's CPU wouldn't come anywhere near the 3900X in single core performance running any load for more than 5 minutes. Once the peak turbo timer runs out, it will return to stable turbo or even lower, close to stock, which is likely around 2 Ghz.
I really dislike these super short benchmarks, they don't say anything about actual performance. Who's gonna play a game for only 5 minutes?
I really couldn't get along with any of the dual core ULV Intel's even for normal (non-productivity) use, they are just too freaking slow. While these quad cores have made it a little better, their sustained performance vs battery consumption (note: I didn't say TDP) is not good enough compared to an old Sandy, Ivy or Haswell 35-45W socketed quad core to warrant upgrading.
My thoughts exactly. Yeah the potential for super high performance in bursts is excellent, but when it's literally limited to something like... 30 seconds for these super low TDP parts, that's falling flat on its face for when it really has to sit there, chugging away. AMD doesn't have this problem.
"Now, before drawing any conclusions, we'd also like to point out that this score is the highest we've seen for the Ice Lake 1065G7 so far. In a previous leak, we've seen the same CPU score somewhat lower (5234) in a Dell XPS 13 7390 although, there could be many factors for the same including pre-production units, unoptimized software, etc."
I'm sorry what though? Unoptimized software? Excuse you? It's freakin' INTEL. They've been the only serious choice for gamers and other CPU users in laptops and desktops since until two years ago! You mean to tell me that this new processor scores substantially lower in a LAPTOP, the environment in which the performance actually matters, because software may be unoptimized? Or is it because they can't mount a desktop cooler on it and keep this chip unrealistically cool under load with unlimited power and TDP restrictions lifted as well?
Lord, look a the 8750H for example. How many laptops are there with varying performance numbers in subsequent Cinebench runs? Initial runs are spot on, but after like... 10+ runs, unless the laptop has top-tier cooling, the scores are anywhere from 10-15% lesser than the first runs, generally. It's kinda funny. The 45w TDP thing needs to stop being used in the spectrum of "technically it's 45w TDP because..." and go back to "It's 45W TDP."
Wow! even if you factor in all the issues with understated TDP etc, it's still quite impressive. If they can get a 6/12 part with higer clocks that will be the new chip for Gaming Laptops
since 8th gen are 4C+ then yes you can say that, I love my Xiaomi with 8250U, it's indeed desktop level and currently even faster they my backup PC with i3 6100 .
From Day 1 I loved the new 8th gen CPU's with 4C/8T, I have no problem with Intel regarding Notebook CPU's performance .
*only the Vendors who put **** cooling and make i7-U Top SKU's to perform like i5 and les with thermal throttling.
Bravo you found a Linux result which everyone knows is higher than Windows. If you noticed in the link you provided the the 1065G7 are all Windows runs at 3.8Ghz while the 3900x are between 4.4-4.5Ghz running Linux. If the 1065G7 was running Clear Linux it would slaughter the 3900x single core score.
Interesting. If only intel would commit to a release date. I am most interested in single threaded performance, as the applications and games I run on my PC scale best to at most 8 threads, and most of them are single or dual threaded only.
@WannaBeOCer the question is if the Max clock on 10nm will limit the ST performance vs Zen2 even if the IPC will be better,for example in CBR15, if Intel will be limited to 4Ghz max boost then if will have same ST performance as Zen2 @~4.4Ghz.
Impressive performance in a notoriously Intel-friendly benchmark test.
Be something if they can really increase performance, decrease power (legitimately, not via trickery and PR lies), and decrease price. I doubt it.
Sent from my iPhone using Tapatalk
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Related Threads
?
?
?
?
?
Ask a question
Ask a question
Overclock.net
27.8M posts
541.2K members
Since 2004
A forum community dedicated to overclocking enthusiasts and testing the limits of computing. Come join the discussion about computing, builds, collections, displays, models, styles, scales, specifications, reviews, accessories, classifieds, and more!