Intel Core i9-14900KF appears on Geekbench, nearly reaching 6GHz

[Daniel Sims]

The big picture: Intel will likely unveil its 14th-generation Core processors next week before shipping them next month. Most of the news concerning them hasn't been impressive, with some leaks indicating that the lineup's higher-end models may show respectable improvements over their predecessors.

Known leaker OneRaichu acquired an Intel Core i9-14900KF and recently posted two Geekbench reports with the upcoming 24-core CPU. One other result from an unknown user also appeared, showing similar numbers. The processor's highest single-core and multi-core scores on Geekbench 6 were 3347 and 23051 respectively. The figures suggest a roughly 10 to 12 percent performance uplift over the CPU's direct predecessor, the 13900KF.

All three tests occurred on an ASRock Z790 Taichi motherboard with DDR5-7000 memory without specifying the power plan used, so we still don't have a complete picture of the new generation's gains. The non-KF i9-14900K improved modestly over its 13th-generation counterpart in a Geekbench test earlier this month using 4800 RAM on a balanced power plan.

One of the most interesting statistics from the 4900KF appears in OneRaichu's Geekbench 5 test, where the maximum CPU frequency reached 5,953MHz, tantalizingly close to the 6GHz mark. The processor features eight performance cores and 16 efficiency cores.

Intel expects to officially reveal the new series of CPUs at its annual Innovation event on September 19 and 20. Reports suggest the company will launch six processors on October 17 – K and KF models ranging from i5 to i9. Non-K variants with limited wattage envelopes will likely appear close to CES 2024 in January.

A significant leak from MSI confirmed that the 14th generation is called Raptor Lake Refresh, indicating a minor upgrade from the 13th-gen Raptor Lake. The leak estimated an average three percent gen-on-gen performance improvement, with the i7-4700K as the outlier at 17 percent. Despite the modest change, leaked store prices indicate a 15 percent price increase over the previous generation.

The best situation in which to purchase a Raptor Lake Refresh CPU is likely as a replacement for an Alder Lake. The new lineup's compatibility with Intel's current LGA 1700 socket allows owners of 12th and 13th-gen processors to upgrade without buying a new motherboard.

Users looking for a more substantial boost should likely wait and build a new PC when Intel releases the 15th-generation processors sometime in late 2024. They will require new motherboards with the LGA 1851 socket and DDR5 RAM.

Permalink to story.

https://www.techspot.com/news/100158-intel-core-i9-14900kf-appears-geekbench-nearly-reaching.html

 

[yRaz]

I guess the GHZ wars are heating up again. I remember we were going to hit 10ghz by 2010. That brings me back, the 939 days with the FX-60. Oh, and Opteron overclocking. Good times.

Well lets see if we can hit 10ghz on silicon before photonics become a thing. We're already more than half way there.

 

[loki1944]

These new cpus are pretty boring; pretty clear the theemal wall has been met. 3.2Ghz on more cores than the average user cares about really isn't very impressive. Even my old Bloomfield i7 980X from 2010 can do 4.5Ghz on all 6 cores, my 2008 Q9650 and i7 920 sit locked @3.8Ghz on air; even my 2006 QX6700 can do 3.2Ghz on air. I understand IPCimprovements matter too; but from a clock speed perspective things are not evolving much; especially when you consider all core max clock vs a couple of 'favored' (such a sissy term) cores. Truth is that for the average user they wouldn't notice if you stuck them with a 2008 CPU; for gaming once you hit ivy bridge same deal with a good OC; extremely playable with good performanc; even Bloomfield is "good enough" aside from some artificial DRM hits from Ubisoft titles like AC 2017 onward.

 

[Peter Farkas]

Nearly 6Ghz while still on 10nm? hmm...

 

[takaozo]

6Ghz whatever, but still +350W?

 

[Bobbydpue]

6Ghz whatever, but still +350W?

That's not likely, it's probably not being overclocked so it's power wouldn't be that high.

 

[Nandiman]

Why is Geekbench always the first at "leaks"? It's one of the most unreliable benchmarks out there. I highly doubt the "leaker" didn't have a couple of minutes to download, install and do one Cinebench run, and then "leak" those results.

 

[PEnnn]

A leak, again.

When is a leak not really a leak? EVERY SINGLE TIME!

 

[Lew Zealand]

That's not likely, it's probably not being overclocked so it's power wouldn't be that high.

What? The 13900K didn't need any overclock to reach 350W, it did that at stock clocks. It's unlikely the 14900K will magically be lower power while being a touch faster on the same process node.

 

[The Hardcard]

Why is Geekbench always the first at "leaks"? It's one of the most unreliable benchmarks out there. I highly doubt the "leaker" didn't have a couple of minutes to download, install and do one Cinebench run, and then "leak" those results.

May I please have an example of Geekbench unreliability. There is data that shows it demonstrating relative performance of CPUs that show up in real world activities. It is very limited, but I have not seen a single example of a processor or core performing outside its Geekbench position on a broad array of computing tasks.

 

[Gastec]

6Ghz whatever, but still +350W?

That's not likely, it's probably not being overclocked so it's power wouldn't be that high.

Nah, it's going to be 300 watts.

 

[mammoth30]

I'm a bit disappointed that in single thread it is barely 35% faster than my i9-11900K. For a CPU that is 3 generations newer is lame.

Nvidia was doing this water drip method for performance improvement back in the days of the GTX790. Now they bring at least 30-40% boost for flagship products each generation.

You could do better Intel....

 

[Lew Zealand]

I'm a bit disappointed that in single thread it is barely 35% faster than my i9-11900K. For a CPU that is 3 generations newer is lame.

Nvidia was doing this water drip method for performance improvement back in the days of the GTX790. Now they bring at least 30-40% boost for flagship products each generation.

You could do better Intel....

What are the reasons for Nvidia's increase in performance? 40-60% higher power requirements last generation and much better process node this generation. Intel has neither of those advantages so it's the trickle. And it was the same set of problems for Nvidia when they were at the trickle back in the day. They're just in different parts of the tech cycle right now.