Laptop i7-12700H and 3080 Ti benchmarks leak ahead of official announcement

[Daniel Sims]

Why it matters: More benchmarks for new Intel and Nvidia hardware showed up on Geekbench this week. This time, it's Intel's and Nvidia's as-of-yet unreleased i7-12700H and laptop RTX 3080 Ti, respectively. The leaks, in multiple entries, offer scant clues on the CPU and GPU's performance against their neighbors.

The entries emerged on Geekbench on Monday, after which BenchLeaks picked them up. They all appear to be for a new entry in HP's Omen gaming laptop series which offers an early peek at the Alder Lake i7-12700H processor and Nvidia RTX 3080 Ti GPU.

All of the entries show the i7 having a 2.45GHz base frequency and a boost frequency of around 4.1GHz, with 14 cores, 20 threads, and 24MB of L3 cache. Its Geekbench 5 scores are 1328 for single-core, and 10517 for multi-core. GPU information, however, is a bit less consistent.

One entry clearly names the laptop 3080 Ti with no other information outside of performance in the Vulkan graphics API, with an overall score of 90114. Another shows an anonymous "Nvidia Graphics Device" with an OpenCL API score of 127392, 16GB of VRAM, a max frequency of 1395MHz, and 58 compute units. Compute units are only used in AMD graphics cards, so this may actually mean 58 of Nvidia's streaming multiprocessors, of which the laptop 3080 has only 48.

Wccftech's comparisons actually put the laptop 3080 Ti a bit below the laptop 3080, but this could simply be because the drivers for the 3080 Ti aren't ready yet. The i7-12700H in a multi-core comparison is able to outdo the 11th-generation i9-11980HK.

A leaked benchmark late last month put the i9-12900HK above Apple's latest M1 Pro and M1 Max.

Permalink to story.

https://www.techspot.com/news/92152-laptop-i7-12700h-3080-ti-benchmarks-leak-ahead.html

 

[Sausagemeat]

So if these Intel parts do consume a lot of power then in a laptop they are probably not the best choice. Although looking at gaming laptops with 3080 class GPUs you probably aren’t going to be away from a power outlet much.

I find the comparisons to the M1 odd. They don’t run the same software or operating system, one is for content creators and the other is for gamers. And one uses barely any power whilst the other has a reputation for drinking power. I must say the M1 is just far more impressive from a technical standpoint. But no one considering this gaming laptop will be interested in a MacBook.

 

[Theinsanegamer]

Alder lake will make much mroe sense on laptops then desktops. And if the desktop GPU situation never changes, perhaps I'll switch to a laptop with the RTX 4000 series (or whenever we get 5nm GPUs).

So if these Intel parts do consume a lot of power then in a laptop they are probably not the best choice. Although looking at gaming laptops with 3080 class GPUs you probably aren’t going to be away from a power outlet much.

I find the comparisons to the M1 odd. They don’t run the same software or operating system, one is for content creators and the other is for gamers. And one uses barely any power whilst the other has a reputation for drinking power. I must say the M1 is just far more impressive from a technical standpoint. But no one considering this gaming laptop will be interested in a MacBook.

I mean the M1 isnt that impressive. Sure it beats the 1185g7 and 4800u in cpu tests, but it also drinks 25 watts to do that (on 5nm no less), while the x86 chips are running at 15-18 watts (on TSMC 7nm and intel 10nm respectively).

The reason the M1 gets compared is because the peanut gallery puts out the "well well well the M1 is faster" argument in every laptop CPU review, even though, as you stated, it isnt compatible with a lot of software and is really designed more for productivity workloads and not gaming.

 

[Sausagemeat]

Alder lake will make much mroe sense on laptops then desktops. And if the desktop GPU situation never changes, perhaps I'll switch to a laptop with the RTX 4000 series (or whenever we get 5nm GPUs).

I mean the M1 isnt that impressive. Sure it beats the 1185g7 and 4800u in cpu tests, but it also drinks 25 watts to do that (on 5nm no less), while the x86 chips are running at 15-18 watts (on TSMC 7nm and intel 10nm respectively).

The reason the M1 gets compared is because the peanut gallery puts out the "well well well the M1 is faster" argument in every laptop CPU review, even though, as you stated, it isnt compatible with a lot of software and is really designed more for productivity workloads and not gaming.

I don’t agree with you on the M1, it’s far more power efficient than X86 equivalents. Your power numbers are wrong, the M1 uses significantly less power than Ryzen or Core series. For the people who would buy an M1 there is no windows hardware and software combination that comes close in terms of performance. I find it far more impressive than anything Intel or AMD have come up with for a long time personally.

Also, how does Alder lake make more sense in a laptop? It uses more power than Ryzen. Now using more power on a desktop doesn’t matter as you have the space to cool the part and you’re always plugged in. But if I’m buying a laptop I want it to have good battery life and choosing a CPU that uses more power will reduce your battery life.

Although I am basing this on the power numbers for the desktop, for laptops things always change so we will need to see how they actually perform. But currently my opinion is the less energy efficient parts will be worse in a laptop.

 

[Theinsanegamer]

I don’t agree with you on the M1, it’s far more power efficient than X86 equivalents. Your power numbers are wrong, the M1 uses significantly less power than Ryzen or Core series. For the people who would buy an M1 there is no windows hardware and software combination that comes close in terms of performance. I find it far more impressive than anything Intel or AMD have come up with for a long time personally.

My power numbers are correct, under compute workloads a m1 mac mini pulls 31 watts of total power, or 26 watts more then idle once loaded.

The 2020 Mac Mini Unleashed: Putting Apple Silicon M1 To The Test

www.anandtech.com

This is comparable to the power draw of 28 watt 1165 processors,w hich are built on 10nm superfin, and the 4800u which is built on TSMC 7nm. Apple's chip is built on 5nm TSMC.

Performance also is comparable to amd/intel laptop chips on older nodes.

Also, how does Alder lake make more sense in a laptop? It uses more power than Ryzen. Now using more power on a desktop doesn’t matter as you have the space to cool the part and you’re always plugged in. But if I’m buying a laptop I want it to have good battery life and choosing a CPU that uses more power will reduce your battery life.

Only during serious compute workloads. Look at techspot's 12900k and 12700k review. Small (5-7%) performance difference, yet the 12700k pulls 90 watts less power then the 12900k. The i5 power draw is comparable to the 5600x. Is zen 3 too power hungry for laptops? No? then likely alder lake will do just as well.

Laptop chips with lower max turbos, running on binned silicon, will likely do very well, idle and low load power draw on alderlake, along with gaming performance, looks really good compared to zen 3.

Although I am basing this on the power numbers for the desktop, for laptops things always change so we will need to see how they actually perform. But currently my opinion is the less energy efficient parts will be worse in a laptop.

For regular workloads, that's zen 3. In non compute alder lake's power draw is quite good, and as software is updated to take better advantage of the little cores, that advantage will grow. There is also DDR5, which doesnt give huge performance bumps but does run at significantly lower voltages, enough to make a 12w difference on desktops. In laptops a simple 0.1v decrease in memory voltage can show up as 30-50 minutes of additional battery life.

 

[Sausagemeat]

My power numbers are correct, under compute workloads a m1 mac mini pulls 31 watts of total power, or 26 watts more then idle once loaded.

The 2020 Mac Mini Unleashed: Putting Apple Silicon M1 To The Test

www.anandtech.com

This is comparable to the power draw of 28 watt 1165 processors,w hich are built on 10nm superfin, and the 4800u which is built on TSMC 7nm. Apple's chip is built on 5nm TSMC.

Performance also is comparable to amd/intel laptop chips on older nodes.

Only during serious compute workloads. Look at techspot's 12900k and 12700k review. Small (5-7%) performance difference, yet the 12700k pulls 90 watts less power then the 12900k. The i5 power draw is comparable to the 5600x. Is zen 3 too power hungry for laptops? No? then likely alder lake will do just as well.

Laptop chips with lower max turbos, running on binned silicon, will likely do very well, idle and low load power draw on alderlake, along with gaming performance, looks really good compared to zen 3.

For regular workloads, that's zen 3. In non compute alder lake's power draw is quite good, and as software is updated to take better advantage of the little cores, that advantage will grow. There is also DDR5, which doesnt give huge performance bumps but does run at significantly lower voltages, enough to make a 12w difference on desktops. In laptops a simple 0.1v decrease in memory voltage can show up as 30-50 minutes of additional battery life.

No your numbers are incorrect. You don’t have a clue and your links don’t back your claims up! I’ve set you to “ignore”. I really can’t be bothered to waste my time with people who have some kind of agenda to warp the truth, whatever it is.

The M1 is a lot more efficicent than X86 CPUs, they don’t even have active coolers as they use so much less power.

Goodbye dude, and give up the gaslighting.

 

[HardReset]

Only during serious compute workloads. Look at techspot's 12900k and 12700k review. Small (5-7%) performance difference, yet the 12700k pulls 90 watts less power then the 12900k. The i5 power draw is comparable to the 5600x. Is zen 3 too power hungry for laptops? No? then likely alder lake will do just as well.

Well, 5600x is TWO chip solution, AMD laptop chips are ONE chip solutions. Making statements about AMD mobile chips and using "downgraded server chip" 5600x as "Zen3 for laptops" just won't work.

 

[Carmaine]

So if these Intel parts do consume a lot of power then in a laptop they are probably not the best choice. Although looking at gaming laptops with 3080 class GPUs you probably aren’t going to be away from a power outlet much.

I find the comparisons to the M1 odd. They don’t run the same software or operating system, one is for content creators and the other is for gamers. And one uses barely any power whilst the other has a reputation for drinking power. I must say the M1 is just far more impressive from a technical standpoint. But no one considering this gaming laptop will be interested in a MacBook.

Yea, not sure why the M1 was even compared. Hardware, software are both completely different and the target audience for Apple really doesn't care tbh. They know macs aren't meant for serious gaming. They never have been.

 

[Hardware Geek]

So if these Intel parts do consume a lot of power then in a laptop they are probably not the best choice. Although looking at gaming laptops with 3080 class GPUs you probably aren’t going to be away from a power outlet much.

I find the comparisons to the M1 odd. They don’t run the same software or operating system, one is for content creators and the other is for gamers. And one uses barely any power whilst the other has a reputation for drinking power. I must say the M1 is just far more impressive from a technical standpoint. But no one considering this gaming laptop will be interested in a MacBook.

Agreed. Apple might have a fantastic first gen processor but Intel is proving they have the ability to compete in terms of performance, which will push AMD to lower prices and release even faster and better versions of its architecture. Every consumer will benefit from either lowered power consumption, higher performance, or both. Sounds like a good thing all around to me.

 

[Watzupken]

So if these Intel parts do consume a lot of power then in a laptop they are probably not the best choice. Although looking at gaming laptops with 3080 class GPUs you probably aren’t going to be away from a power outlet much.

I find the comparisons to the M1 odd. They don’t run the same software or operating system, one is for content creators and the other is for gamers. And one uses barely any power whilst the other has a reputation for drinking power. I must say the M1 is just far more impressive from a technical standpoint. But no one considering this gaming laptop will be interested in a MacBook.

It isn't always true that people eyeing a gaming laptop is going to use it for gaming, or at least not primarily. Gaming laptops come with beefy specs that can be beneficial for work. The comparison to M1 Macs are valid in a sense that if you are using a laptop for work, with little or no emphasis on gaming, then the M1 Macbook Pros can really be very competitive, and also great to bring around due to the impressive battery endurance. The M1 is outscored by Alder Lake in single threaded load, but if I were to get a machine specialised at video editing for example, I feel the M1 Macs are very attractive solution.
Also as a gaming laptop user in the past, I generally don't like gaming laptops because they tend to run very hot and loud when gaming. To add on to the insult, if you unplug in from the mains, most processor and GPU run at a severely throttled clock speed. So it is really not that portable having to lug a massive power brick or be prepared to get crappy battery life + performance.