Intel Core Ultra 5 245K Review: Why So Slow for Gaming?

Adobe Photoshop 2025 graph is wrong for me, showing the 14 game average instead.

Editor’s note. Thanks. Fixed it earlier today.
 
Maybe Intel was suffering from the slow sales of 13-14th gen CPU's, due to the silicon degradation issue. So they thought they had to release SOMETHING - as a matter of urgency - that was free from that legacy, even if uncompetitive.
 
In the meantime, the US government wants AMD to buy Intel CPU division...

Why would AMD do that with the current state of Intel CPU business...?

If they are doing that, it would be for becoming a monopoly and scrapping the whole division by replacing Intel CPUs for AMD CPU in all the Client Compute Group.
 
Kind of reminds me of AMD's Zen 1... a step in the right direction, but don't buy until it's "fixed".... Took AMD a couple of generations to truly "fix" Zen... let's hope Intel works a bit faster - but wouldn't hold my breath.
It is not a step in the right direction. It is literally a Bulldozer moment.

Intel went to TSMC with their latest node for relevancy, and still they fail at it.

At the moment, Intel is in hot water and the government is already starting to ask for Intel to split their business for their homeland fabs.
 
It is not a step in the right direction. It is literally a Bulldozer moment.

Intel went to TSMC with their latest node for relevancy, and still they fail at it.

At the moment, Intel is in hot water and the government is already starting to ask for Intel to split their business for their homeland fabs.
Not the government, a few board members.
 
It finally happened. A main player in the CPU market released a new generation product with overall slower performance than the previous generation, no improvements in efficiency, and just slightly lower MSRP.

Is there someone who still doesn't believe Moore's Law is over? We'll probably only see slightly incremental improvements during the next decades, probably more focused on efficiency than performance. Next 20 years in a best case scenario, 200 years at worst.

Maybe Intel was suffering from the slow sales of 13-14th gen CPU's, due to the silicon degradation issue. So they thought they had to release SOMETHING - as a matter of urgency - that was free from that legacy, even if uncompetitive.

Yeah, the main problem is that line must go up ad infinitum. In that case, until coming up with an actual solid and tested design, it would probably have been wiser for Intel to just re-release 12th gen (or even better, ditch e-cores completely and re-release 11th gen). At MSRPs low enough to be competitive against current AMD offerings in price/performance ratio, of course.
 
A buyout from AMD is not going to happen. That would give them a monopoly on x86. Other than this site you mention, I can find nothing else on it.
Apple has a monopoly on Mx CPUs (and everything that goes along with them, including the OS and all system hardware). Why would this be any different?
 
LATENCY KILLS. lots of dies == more latency. I suspect like AMD it will get better in future iterations. but this is a CLEAR early adopter penalty
 
Apple has a monopoly on Mx CPUs (and everything that goes along with them, including the OS and all system hardware). Why would this be any different?

Because M# CPUs are based on licensed ARM ISAs and ARM licenses those to many other companies. There's good competition in the ARM ISA space, even if Apple's kicking the most butt design-wise.
 
Back in 2011, when AMD's "Bulldozer"(FX-8100) architecture proved inferior to Intel's "Sandy Bridge"(Core i7 2600k). Intel made a(BAD) decision. Profits over research and innovation!
6 generations of Intel Core processors saw incremental improvements and small architectural changes.
Then came ZEN!
Suddenly Intel decided that 4c/8t Core i7's(7700k) now should have 6c/12t(8700k).
Then they released a 8c/16t Core i9(9900k) to lesser dominance over ZEN 2.
Then came ZEN 3, 4, 5, and I think Intel will be playing catch-up for a very long time.
 
A buyout from AMD is not going to happen. That would give them a monopoly on x86. Other than this site you mention, I can find nothing else on it.
Well, someone wanted some kind of proof, and I provided it. Do I necessarily agree? Nope, not a chance!
 
Apple has a monopoly on Mx CPUs (and everything that goes along with them, including the OS and all system hardware). Why would this be any different?
Because the M series chips were developed by Apple to run an operating system they developed. The M series chips are better described as an ASIC with ARM features than it is a general purpose CPU.

Also, saying Apple has a monopoly on M series chips is inaccurate. There isn't a market for M series chips outside of Apple products. A monopoly only applies when a consumer is forced to buy from one manufacturer. You want a computer, it doesn't have to be a Mac. If macs were the only computers then, yes, Apple would have a monopoly. However, at this point, Apple products are a preference not a necessity.
 
LATENCY KILLS. lots of dies == more latency. I suspect like AMD it will get better in future iterations. but this is a CLEAR early adopter penalty
Ringbus is clocked much lower than for Raptor Lake for a start and latency is 15-20ns with memory controller on the SoC tile and not the cpu tile. Panther Lake is said to bring it back to the cpu tile.

Although Arrow Lake currently sucks for gaming, I would still take it in a heartbeat over 9700X as I do more productivity than gaming. Also I think gaming will improve a decent amount over next few months, so that isn't a concern. I'll probably get 265K though unless 9800X3D is really much better than 7800X3D for productivity, as leaks have been saying. I guess we'll know in a few days.
 
It feels like Intel focused too much on balancing power efficiency with productivity at the cost of gaming performance. The 245K might appeal to professionals who occasionally game, but for gamers who occasionally work, it’s hard to justify the price tag.
 
It feels like Intel focused too much on balancing power efficiency with productivity at the cost of gaming performance. The 245K might appeal to professionals who occasionally game, but for gamers who occasionally work, it’s hard to justify the price tag.
Most modern CPUs are not holding back a 4080 most of the time (high quality, high resolution gaming). Add to that the size of the gaming pc market and we see why the last gen CPUs are not focused on improving gaming performance. I think intel chose right to focus on improving power usage and productivity while not sacrificing gaming altogether. TBH, the power consumption in games seems too low and might signal a low utilization that might improve with software optimizations.
 
Back