Leak Exposes Seven New Intel Core Ultra 200S Processors, Including T & F Series

[Nomad76]

Intel is expanding its Core Ultra 200S lineup with seven new SKUs, a recent leak by @momomo_us shows. The leak gives details on non-K, T, and F versions across the Core Ultra 5 7, and 9 families. The top-end Core Ultra 9 285 will come in 65 W and 35 W versions. These match the 285K's 24-core setup but run at lower speeds. The 285T keeps the 4-core Xe built-in graphics.

The Core Ultra 7 series adds three models: 265 265F, and 265T. All share a 30 MB L3 cache being differentiated by different clock speeds. The T version has much lower base clocks. For the Core Ultra 5 series, the leak shows two models: 225 and 225F. Both have 10 cores running at 3.3/4.9 GHz (P-cores) and 2.7/4.4 GHz (E-cores), with 2 Xe GPU cores and a 65 W TDP. Unlike their higher-end models, this tier doesn't plan to have a T version right now.



Wccftech reports that we can expect to see more locked models in the Core Ultra 3 range (235, 215, and 205), however this leak didn't include them.

SKUs | Cores | P-Core (Base/Boost) | E-Core (Base/Boost) | L3 Cache | TDP

• Core Ultra 9 285T | 24 (8-P + 16-E) | 1.4/5.4 GHz | 1.2/4.5 GHz | 36 MB | 35 W
• Core Ultra 9 285 | 24 (8-P + 16-E) | 2.5/5.6 GHz | 1.9/4.6 GHz | 36 MB | 65 W
• Core Ultra 7 265T | 20 (8-P + 12-E) | 1.5/5.3 GHz | 1.2/4.6 GHz | 30 MB | 35 W
• Core Ultra 7 265F | 20 (8-P + 12-E) | 2.4/5.3 GHz | 1.8/4.6 GHz | 30 MB | 65 W
• Core Ultra 7 265 | 20 (8-P + 12-E) | 2.4/5.3 GHz | 1.8/4.6 GHz | 30 MB | 65 W
• Core Ultra 5 225F | 10 (6-P + 4-E) | 3.3/4.9 GHz | 2.7/4.4 GHz | 20 MB | 65 W
• Core Ultra 5 225 | 10 (6-P + 4-E) | 3.3/4.9 GHz 2.7/4.4 GHz | 20 MB | 65 W

View at TechPowerUp Main Site | Source

 

[watzupken]

To be honest, if Intel is unable to meaningfully optimize their Arrow Lake K CPUs in the near future, it will negatively impact demand for these lower powered versions. Not to mention, I think there was a leak earlier that suggested that non-K SKUs will not be able to run the memory as fast as the K SKUs, which again, will negatively impact performance.

 

[Nomad76]

To be honest, if Intel is unable to meaningfully optimize their Arrow Lake K CPUs in the near future, it will negatively impact demand for these lower powered versions. Not to mention, I think there was a leak earlier that suggested that non-K SKUs will not be able to run the memory as fast as the K SKUs, which again, will negatively impact performance.

Remain to be seen as Intel says that "a set of future patches will fix them".

 

[Daven]

I guess the rest will be something like

235 - 6P x 8E
215 - 6P
205 - 4P

Hopefully no more 2P only CPUs.

 

[bug]

Save for the 225, I'm not sure what was leaked here. Intel has had K, F and T variants for ages. With the exact meaning we see here.

 

[phints]

Gaming performance so mediocre with the 265K (although the 1% lows are decent), I worry how slow a 65W 265 might be. Still interested in seeing how it does though.

 

[AhmadMZ99]

most likely the T series for OEMs, other models i expected to release later after initial launch of high end cpus.
btw when will core utlra 3 will launch for desktop cpu?

 

[JustBenching]

Gaming performance so mediocre with the 265K (although the 1% lows are decent), I worry how slow a 65W 265 might be. Still interested in seeing how it does though.

It's basically within 2% of the 7950x in gaming (720p) while consuming 77w (vs 116 for the 7950x). I don't think the 65w will impact it that much.

 

[Hecate91]

The 265K is less efficient than a 7700X in most applications according to TPU,limiting these to 65W means they'll be even slower compared to previous gen.

 

[JustBenching]

The 265K is less efficient than a 7700X in most applications according to TPU,limiting these to 65W means they'll be even slower compared to previous gen

Limiting the to 65w will make them both more efficient and faster than the 7700x you just mentioned.

I don't know where you got your numbers from but according to TPU, the 7700x consumes 135w vs 155w for the i7 in blender but the i7 finishes in half the time. The 7700x is obnoxiously inefficient compared to the i7.

 

[pressing on]

I guess the rest will be something like

235 - 6P x 8E
215 - 6P
205 - 4P

Hopefully no more 2P only CPUs.

Based on previous leaks the 235 should be a Core Ultra 5 processor with 10 cores (6-P + 4-E). P cores base of 3.4 GHz boosting to 5.0 GHz.

The 215 and 205 have been claimed to be Core Ultra 3 processors with 8 cores (4-P + 4-E). 215 said to have P cores base of 3.9 GHz boosting to 4.8 GHz.

 

[N/A]

Since the 10 cores in 225 is the equivalent of 8/16 threads, this is a 33% better offering than the 12400F, only if it's fully ported on N3 and no more than 2 tiles same as Lunar lake. Realistically, this 6+4 SKU could be a new tile variant. probably not a 8+16 tile. Just add 2 P cores to Lunar and INtel could have a winner here if the price is right.

 

[Lycanwolfen]

I really could care less, Switched to AMD now pure 16 core monster and it shows. No E core no I core just pure 16 core Performance and man does it shine. It's like Ford in the GT40 for a test drive saying I had no Idea.
AMD is king sorry intel you lost it.

 

[Prima.Vera]

Sorry, this generation is already doomed.
Instead of focusing on those garbage CPUs, Intel should focus on the next gen and bring something substantial to the table.

 

[BillyRazOr]

I really could care less, Switched to AMD now pure 16 core monster and it shows. No E core no I core just pure 16 core Performance and man does it shine. It's like Ford in the GT40 for a test drive saying I had no Idea.
AMD is king sorry intel you lost it.

Not everyone would like just focusing on performance core, also you said "AMD is king" are only when they got the best CPUs in THIS generation, but I won't 100% guaranteed of every company in future life.

 

[95Viper]

Stay on topic.
Don't be making this a "I' versus "A" thread.

 

[AleXXX666]

I guess the rest will be something like

235 - 6P x 8E
215 - 6P
205 - 4P

Hopefully no more 2P only CPUs.

hopefully no amd 8000 series nonsense with "e-core" count dominating the p-cores...