• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Intel Core i7-12700 Geekbenched, Matches Ryzen 7 5800X

Then it means you have no idea of how your chip works based on bios values. Once again, it is simple in the statement I made originally. If you set PL1 and PL2 limits (and especially PL1) to specific values - it’ll run exactly this. That’s why my cpu never goes beyond 145w (pl1 and pl2 in bios) and can sustain its all-core boost without downclocking, which is easy to monitor via hwinfo for example. Pretty straightforward.
Read what you just wrote... my 125w CPU doesn't go beyond 145w after tweaking BIOS settings.

So:
1. with knowing how the CPU behaves,
2. with knowing how to tweak specific power settings in bios, and overriding MOST bioses in Z boards which default to unlimited power limits
3. while also using enthusiast software to monitor,

Then your '125W' CPU runs no more than 145W as per the spec.

Not complicated at all. It's an awesome CPU - I'm not knocking it, but please stop defending that marketing department and their departures from reality.
 
Haha, that dude is arguing for Intel and power consumption restraint? :roll:
 
Totally agree. Those should be compared to equivalent (the upcoming Zen 4) instead of the "old" Zen 3 if you want to have a fair comparison.
Problem is that both parts are not out yet.
 
Totally agree. Those should be compared to equivalent (the upcoming Zen 4) instead of the "old" Zen 3 if you want to have a fair comparison.
Zen4 gonna come out in a year, how is that fair. It's more like Raptor lake vs zen4 and Alder lake vs zen3d
 
New sockets are normal when there is a memory technology change (here DDR5 from DDR4). They are not normal when the memory doesn't even change...
What are you talking about? Socket FM1 to FM2 to FM2+, AND socket AM3 and AM3+ were all DDR3. AM2 and Am2+ were DDR2. socket 7, super socket 7, slot 1, and socket a were all SD RAM. 754, 939, and 940 were all DDR.

Why do you act like having a single socket for a memory generation is normal? Motherboards and CPUs last a LONG time nowadays, you dont need to upgrade every year, or every 5 years. And frankly, when you have to rely on public outrage to get firmware support for your old motherboard to support the new CPUs, (and still not get proper support to run 5000 series on x370/b350) who cares? Or you end up with intel's LGA 775, and CPU compatibility was based not on socket but chipset, and was needlessly confusing.
 
@btarunr I wanna know if this is a PR direct from Intel. I'm sure its safe to assume so since there is no source indicated.

As with all of Intel PRs, when its not really a PR, but to everyone else that appears to not be looking, it really is... I can say that Intel has a history of fudging numbers to candy coat the chocolate bar. The only way I'll believe these benchmarks is to have a more reliable source review the CPU, and not a engineering sample.
 
Last edited:
But it's Intel 65W.... so it's probably something like 150W then to 125W then to 65W at 90C.
IF this is an almost mature CPU, e.g. qualification sample, the base clock indicates that it’s indeed restricted to Intel’s official PL values.
 
As some people reports, ADL do not have AVX512 on desktop and that mean have a bit slower total score than RKL.
That's incorrect. Alder Lake does in fact feature AVX-512.
Don't judge any product until the final firmware and BIOSes are out.

But the integer score is up by 10%. Far from the 19% promised but still.
19% higher IPC doesn't mean 19% higher score in every benchmark.
But Geekbench is a useless benchmark anyways. Hopefully you don't buy a computer to run Geekbench. ;)
 
So you tell me, how are you not shilling for AMD?
Because I don't have any emotion in this. I have seen this go back and forth already. I don't care. The people who make the best hardware for my dollar at the time I build gets my money. I have been back and forth with Intel more than a few times now. You know why I wont buy an AMD card? Because its not worth my dollar. Especially in these days. And besides, I am still pissed at them for killing ATi. But I cant be pissed at them for kicking Intel in the bag.
 
IF this is an almost mature CPU, e.g. qualification sample, the base clock indicates that it’s indeed restricted to Intel’s official PL values.

However any benchmark testing may happen entirely within the CPU's PL1 high power setting and long term use will exit that higher limit and thus have lower performance for long term 65W loads.

I have an i7 9700 non-K 8 core (nominally 65W) and it sucks down over 130W in sustained loads at 4.5GHz max boost when I unlock the power restriction in XTU. If I restrict it to 65W it rummages around in the mid 3.x GHz range depending on the workload with obviously lower overall performance.
 
Last edited:
This a stupid discussion. AMD tried to lock out CPU's from it's latest motherboard for literally no reason, because they backtracked after people complaint. Just like AMD started to increase their prices over Intel.

People thinking Intel is evil and AMD is so much better are just being plain stupid. They will both srew you if they can, if they are in the top, and "help you", be on your side if they are down. Stop being shils for a company that only cares about your money
Amen brother!
It's all about the cha-ching!

AMD traditionally had better midrange deals compared to Intel and NVIDA, and most of machines I've had been just that. And it's not because they were the "good" guys but because midrange is as high they could aim for.
If I were to build a new machine right now it would be Intel/Nvidia. I would go with i7 10700k it's $250 at Microcenter and $20 off when bundled with mobo.
And for GPU if for the same money I would pick RTX simple because they are evenly matched but with better raytracing perf.
Also to sum it up everybody AMD, Intel and NVIDA they all are milking the "shortages" they all have record earnings.
If they learn anything from the scalpers that we are all sheep and ready to pay ridiculous prices.
 
The first gen (Alder Lake) of this BIG little hoohah is definitely not the time to jump into Intel's new processor direction. It'll be full of teething and scheduler problems much worse than Ryzen 1st gen, especially as they're throwing nascent DDR5 memory into the mix.

2nd gen, or better yet, 3rd gen of BIG little should be there, in terms of potential.
 
"Matches" is the keyword, meaning they are still catching up. With supposed 125W TDP it has no technical advantages at all (we're yet to see any tangible benefits of Big-Little in desktops), so the only way to get back into the game is to lowball the price.
If it's a Non-K model then it's going to have 65 W TDP.
 
lol it is 2021 and ppl still defending Intel TDP trick numbers

The following is a chart from PCworld
Measuring system power consumption of 5900x (105W TDP) in blue vs 11900k (125W TDP) in purple

Now look at that chart and tells us how "125W" draws as twice as power than "105W"

Non-K CPUs are no exception.
Your "65W" CPU will conveniently draw 200W when it wants to.
It is Intel, their PL2 number is always twice as their so-called "TDP" number


cinebench_r20_runs-100882508-orig.jpg
 
Low quality post by Why_Me
Low quality post by Tardian
Stay on topic.
The topic is: "Intel Core i7-12700 Geekbenched, Matches Ryzen 7 5800X"
Stop the off topic posting, trolling, etc.
 
I have had several non-K Intel CPUs with 65W TDP. After the boost period expires they have stayed quite nicely at 65W. For non-K models motherboards seem to be more limited in how much they boost the PL2 and Tau, in my experience mostly the spec-ish 1.35x and 56s.
Why? That depends entirely on the eventual price. 5900X is $550 CPU (and 5950X is an $800 CPU). Intel desktop range has been somewhat cheaper so far. Official prices have been $500 for 9900K, 10900K and $550 for 11900K. And everyone quite unanimously agrees that i9 is overpriced.

And for specific compared CPUs - 5800X is a $450 CPU. i7 11700/10700/9700 are $330.
It is true that post Intel's specified boost duration, it will drop back to 65W. However, it depends on the duration of the benchmark. For short burst, it may appear that it is very fast because it is within boost duration. At 65W, taking the Rocket Lake as an example, it runs at around 3.5 Ghz, which will hurt performance.

I would somewhat agree on the second point about the dependency on eventual price, but I generally don't think Intel will sell something they deem competitive at a bargain price. In my opinion, Comet Lake and to some extend Rocket Lake are being sold at a lower price to stay competitive. I got a 10700K because it was so much cheaper than a 5800X @ MSRP. And this is unusually cheap from Intel considering that my current 10700K is almost as cheap as the i5 Skylake I last owned from Intel. From a product positioning standpoint, I still feel that an Alder Lake i7 is potentially going up against a Ryzen 9 5900X because it is clear that they are pushing 8 performance + 4 efficient cores to make up that 12 cores, which is consistent with the rumoured i9 with 8 performance + 8 efficient cores to make up that 16 cores.
 
Whats the clock speeds of this 12700? 5GHz Single im guessing? Multi I have no idea
 
woah, woah, woah??? a next gen equalling a previous gen! (again...) who thought ...

oh well ... this is going to be interesting later ... now it's a bit meh, to see the obvious happening ...

will be hilarious if that locked one also match the price of the 5800X ... pffffahahah :laugh:
 
Whats the clock speeds of this 12700? 5GHz Single im guessing? Multi I have no idea
4789 MHz accordind to screenshot.

And now another Geekbench benchmark emerges, this time with 12900K. The bell is ringing, round 2 begins.
 
like it's any secret that Intel is using a lot more power because they are using a much higher clock speed to try and compete with Intel. This is another absurd discussion.

TDP is a metric that does not reflect max boost frequency, neither in Intel or AMD, Intel boosts higher. Drama intensifies
 
4789 MHz accordind to screenshot.

And now another Geekbench benchmark emerges, this time with 12900K. The bell is ringing, round 2 begins.

Really? oh dear that doesnt look good then since they where claiming a lot bigger IPC over Zen 3....and casts not a good sign for the 12900K then and this is in geekbench lol
 
Low quality post by Tardian
like it's any secret that Intel is using a lot more power because they are using a much higher clock speed to try and compete with Intel. This is another absurd discussion.

TDP is a metric that does not reflect max boost frequency, neither in Intel or AMD, Intel boosts higher. Drama intensifies
I did message you, but you haven't edited your first sentence. That is what is absurd! I think if you substitute one of the Intel for AMD it might be clearer. I tried to be nice.
 
Back
Top