• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Intel Core Ultra 7 270K Plus Review - Intel's Fastest Gaming CPU

Unreleased Core Ultra 9 290K Plus shows up again
1774391348703.png


https://videocardz.com/225586/intel-canceled-core-ultra-9-290k-plus-still-showing-up-in-benchmarks
 
AMD: It was us. We were looking at doing a new naming convention. 290K had a nice ring to it. Lawyers were skeptical. But when we redid "AI", that is Artificial Intelligence which we shortened to AI, and nobody like it. We decided to just skip the Artificial altogether and go with, just "Intelligence", or "Intel" for short. So the Intel 290K concept was born. Lawyers had a fit. So we had to go back to our original naming and the 9850X3D, etc.
 
9900X down to $315 on Amazon right now. I believe that is an all-time low aside from price errors and the like.
 
Pretty cool. Seems like a halfway decent part and it's driving other prices down, too. That's win win.
 
If RAM / NVME's was affordable, I would consider a new Intel build. Besides, Summer is coming. Time to kick the heaters off till next winter.
The 270K has potential it seems. Hopefully Intel can kick out some more gems in the future.
 
Seems decent. For gaming the X3D CPUs still seem better overall but these are quite close and generally do better in non-gaming tasks, so it is a better multi use option. Good to see that. AMD seemed to be slowing down in generational improvements so some good alternatives from Intel will probably make AMD more competitive again. 7-8 years ago we were happy when AMD came out with something decent after years of Intel lagging in progress.
 
erek said:
that’s unfortunate
Click to expand...
Well, I really liked Intel for the longest time, but these last few gems, especially when talking power usage vs performance, have been.... Questionable.


I may look into a 270K, but it won't really rival anything I don't own already.
It would be a build, just to build it really.
 
  • Like
Reactions: erek
like this
Dang Intel. The 250K P matches AMD R7 non-x3d in gaming and easily beats it in most other tasks.

The 270K P makes the entire R9 lineup irrelevant, save the x3d.

Really, outside spaces like this, most would sacrifice the gaming advantages for the price and cpu advantages of Intel new parts.

AMD needs a response to keep their non X3d parts attractive, ignoring the 'dead platform' arguments.

Why not an 'X2D' refresh at current non-x3d prices?
 
I imagine that Intel is mostly doing this for the PR at this point. Who the hell cares that much about what a budget CPU is priced at? The rest of the system will be so expensive that the concept of a budget PC doesn't exist. Also, 300$ to stay on a dead end platform doesn't exactly sound that enticing.

With current ram prices and other shit, the desktop space is dead.

https://pcpartpicker.com/trends/price/memory/

Even DDR4 is a shitshow.
 
It's interesting, but sadly not tempting unless you already have DDR5, especially considering a dead end platform. I think I'll stick with my i9-10940x and quad channel DDR4 running at 3600MHz CL16 considering the cost of DDR5, especially outside of the USA. With an all core boost of 4.8GHz and a tweak to cache speeds the 10940x still holds it's own and it'll easily handle 5GHz all core should I want for more.

Not a chance I'm buying hardware in this market if I can help it.
 
Last edited:
Mazzspeed said:
It's interesting, but sadly not tempting unless you already have DDR5, especially considering a dead end platform. I think I'll stick with my i9-10940x and quad channel DDR4 running at 3600MHz CL16 considering the cost of DDR5, especially outside of the USA. With an all core boost of 4.8GHz and a tweak to cache speeds the 10940x still holds it's own and it'll easily handle 5GHz all core should I want for more.

Not a chance I'm buying hardware in this market if I can help it.
Click to expand...
i agree with you as you can see my gaming rig in my sig... but my son, has the same z690-a ddr4 mobo that I have but an i3-12100f cpu and rtx 3060 ti to which he is struggling at some games. Trying to decied if just cpu or gpu or both to fix it for him.

He said he was going to get an amd combo.. then i said.. ok you need memory (ddr5) and cpu heatsink.. he was like.. oh.. then i said oh your psu needs to be changed... then he realized it is almost a new pc lol
 
For what it’s worth my brother is still running a z690 d4 APro board 32g ram and my old 3060ti. We tossed in a used 12900k for $250 last year and he plays everything. GPU limits of course but doable.

I’d look for a used 12700k or kf for your son’s pc and call it a day for now. Maybe psu and gpu to top it off at some point.

On topic I’m tempted to try a 270k box since I have ram but my current systems don’t struggle with anything I do. I’d have to sell stuff or end up with even more computers lying around!
 
TeleFragger said:
i agree with you as you can see my gaming rig in my sig... but my son, has the same z690-a ddr4 mobo that I have but an i3-12100f cpu and rtx 3060 ti to which he is struggling at some games. Trying to decied if just cpu or gpu or both to fix it for him.

He said he was going to get an amd combo.. then i said.. ok you need memory (ddr5) and cpu heatsink.. he was like.. oh.. then i said oh your psu needs to be changed... then he realized it is almost a new pc lol
Click to expand...

12100 is a quad core. Lots of games would benefit from 2 more P-Cores. 12600K or KF, would be a great boost for that system.
You can see examples here. Call of Duty, Cyberpunk, and Hogwartz choke on 4 cores.

View: https://youtu.be/qIWZOYa2qHo?si=5M2pVf-hN8qqiUpr


There's an i5-13500 an Ebay right now used, for $160. It has the updated cache from Raptor Lake. And helps maintain performance a little better with DDR4, compared to a 12600K.
https://www.ebay.com/itm/2272737894...+5TXFuQ2b+8RpYQ==|tkp:Bk9SR5j3rqqlZw&LH_BIN=1

However, 12600K can be overclocked and that might be fun to do with your son. And they are around $145 used, right now.
ignore the 13400 and 14400. They cost about the same as a 12600K, but are slower. Despite the higher generation number of 13 and 14, they are still based on 12th Gen Alder Lake silicon. and are essentially slower 12600K.
The plain 12600 is actually a straight 6 core. Intel's last CPU with 6 P-cores, only.
 
Last edited:
As an eBay Associate, HardForum may earn from qualifying purchases.
with stock not instantly sold out (if true)

that’s a bad sign for demand


i barely got my 9800X3D on launch day
 
erek said:
with stock not instantly sold out (if true)

that’s a bad sign for demand


i barely got my 9800X3D on launch day
Click to expand...
Hopefully it'll be down to msrp in a few weeks. I've got a 265k that needs replacing.
 

"Intel Core Ultra 200S Plus Series Processors Available on iBUYPOWER Gaming PCs​

Press Release by Nomad76 Today, 16:29 Discuss (0 Comments)
The leading system integrator iBUYPOWER is happy to share they will be integrating the new Intel Core Ultra 200S Plus series processors into its gaming PC lineup. Any new computer using an Intel Core Ultra 200S Plus series processor is covered by iBUYPOWER's industry‑leading three‑year labor and two‑year parts warranty, which the leading national computer and consumer electronics retailer Micro Center honors nationwide for convenient in‑store diagnosis and repair.

Compatible with Intel 800 Series chipset-based motherboards, the Intel Core Ultra 200S Plus series processors deliver up to 15% faster geomean gaming performance -- versus existing Core Ultra Series 2 desktop CPUs -- and up to 103% more multithreaded performance compared to similar CPUs in its class. With the new platform utilizing all-new software that simplifies initial set-up and performance acquisition, the Intel Core Ultra 200S Plus processors are ideal for everyday gamers and creators that crave max value with real performance."
 
1.1.2.3.5... said:
I hate e-cores
Click to expand...
Yes, but, in all fairness, lop sided X3D causes issues as well from a "how to schedule" point of view. And, AMD has already started their "c" cores concept, while they aren't true "sub par" cores, they don't have the cache.... making it sub par (if you will). Anyway, I know that for a typical end user, they aren't going to care, but if you're into hypervisors or other things where uniformity could matter a lot and there's greater risk of creating a bad scheduling scenario, the whole cores are different trend can be a real headache (things will run differently every time).

To Intel's apparent problem going above 8 cores (that is, in a competitive way).... it was a good answer for them. And, it has made them competitive. But, as we know, AMD is about to leap frog again, so.... I'm looking forward to Intel's future responses (I don't think e,f,g cores, etc. are... the way).
 
  • Like
Reactions: erek
like this
1.1.2.3.5... said:
I hate e-cores
Click to expand...
cjcox said:
Yes, but, in all fairness, lop sided X3D causes issues as well from a "how to schedule" point of view. And, AMD has already started their "c" cores concept, while they aren't true "sub par" cores, they don't have the cache.... making it sub par (if you will). Anyway, I know that for a typical end user, they aren't going to care, but if you're into hypervisors or other things where uniformity could matter a lot and there's greater risk of creating a bad scheduling scenario, the whole cores are different trend can be a real headache (things will run differently every time).

To Intel's apparent problem going above 8 cores (that is, in a competitive way).... it was a good answer for them. And, it has made them competitive. But, as we know, AMD is about to leap frog again, so.... I'm looking forward to Intel's future responses (I don't think e,f,g cores, etc. are... the way).
Click to expand...
what about this, “The Binary Optimization Tool reflects the best performance possible from the CPU, but it might not reflect real-world reality, so Geekbench is calling it out.”

https://x.com/HotHardware/status/2037271806705348619
 
cjcox said:
Yes, but, in all fairness, lop sided X3D causes issues as well from a "how to schedule" point of view. And, AMD has already started their "c" cores concept, while they aren't true "sub par" cores, they don't have the cache.... making it sub par (if you will). Anyway, I know that for a typical end user, they aren't going to care, but if you're into hypervisors or other things where uniformity could matter a lot and there's greater risk of creating a bad scheduling scenario, the whole cores are different trend can be a real headache (things will run differently every time).

To Intel's apparent problem going above 8 cores (that is, in a competitive way).... it was a good answer for them. And, it has made them competitive. But, as we know, AMD is about to leap frog again, so.... I'm looking forward to Intel's future responses (I don't think e,f,g cores, etc. are... the way).
Click to expand...
Well, I do a ton of simulation that uses a lot of linear algebra and the e-core p-core instructions seem to be all over the place. I don't use any AMD cpus with 3d cache since I am not chasing the highest possible frame rates in games. Even the new Intel AVX10 won't be consistent, while AMD has effectively doubled down on AVX512 giving devs a simple path for strong SIMD.
 
erek said:
what about this, “The Binary Optimization Tool reflects the best performance possible from the CPU, but it might not reflect real-world reality, so Geekbench is calling it out.”

https://x.com/HotHardware/status/2037271806705348619
Click to expand...
Geekbench is meant to be a static codebase. So that you can see how different CPUs handle the same exact code. Its......not really useful because it doesn't mean anything. Its just a benchmark which you can choose to arbitrarily pay attention to, or not.

IBOT potentially swaps out code for versions which run better on Intel. IBOT can also help focus an app on certain cores, for better efficiency. The goal being faster exection/better IPC---with the same end output to the user. (same visuals, same behavior, etc).

In Geekbench's opinion, this means they can't say that the results are the same/comparable (because Geekbench's output to the user is a score. Not an image or a model, etc). and that goes against the point of the benchmark, to give a "score" to a CPU.

Intel used IBOT on Geekbench to show that its not only for games. It can and presumably will be applied to popular productivity apps, at a later time. Similarly, Intel could potentially work with software developers and game developers, to bake in these optimizations, so that you don't need to specifically activate APO and IBOT. AMD did that sort of thing in the past year with Photoshop. And AMD now spanks intel in most portions of Photoshop.
 
1.1.2.3.5... said:
Well, I do a ton of simulation that uses a lot of linear algebra and the e-core p-core instructions seem to be all over the place. I don't use any AMD cpus with 3d cache since I am not chasing the highest possible frame rates in games. Even the new Intel AVX10 won't be consistent, while AMD has effectively doubled down on AVX512 giving devs a simple path for strong SIMD.
Click to expand...
Yes, Intel's inability to "figure things out" right now has really hurt them. I do think right now is the beat window for making strides. We'll see. Regardless, while AMD is hot, Intel, even at the desktop CPU level, still rules overall (just not the enthusiast DIY markets). But yes, for high server Epic/TR/Xeon style computing needs on a tight budgets, AMD ftw.
 
chameleoneel said:
In Geekbench's opinion, this means they can't say that the results are the same/comparable (because Geekbench's output to the user is a score. Not an image or a model, etc). and that goes against the point of the benchmark, to give a "score" to a CPU.
Click to expand...
geekbench results are pretty much all verifiable that ibot does not change the result (many of the benchmark have validation tools that the output is correct or not) I do not think that geekbench issues, a big one will be that no apps have an ibot code path made for them yet, that make the score unrepresentative.

An other one could be that code safety could be compromissed (with how robust the memory stack was made to resist attack vs going for speed), lot of geekbench is running a browser or displaying PDF, being a black box, it could give the exact same results but in a less secure way, which would be unrepresentative of code that choosen to be slow in exchange of more secure branching/stack decision.

If this is very similar to what Nvidia do with the just in time compilation of CUDA ptx, maybe they could do something similar in term of dev control, where you can say this need to stay like this and to avoid optimising it for speed and nvidia let you see the final optimised assembly it made to validate for yourself.
 
Last edited:
chameleoneel said:
Geekbench is meant to be a static codebase. So that you can see how different CPUs handle the same exact code. Its......not really useful because it doesn't mean anything. Its just a benchmark which you can choose to arbitrarily pay attention to, or not.

IBOT potentially swaps out code for versions which run better on Intel. IBOT can also help focus an app on certain cores, for better efficiency. The goal being faster exection/better IPC---with the same end output to the user. (same visuals, same behavior, etc).

In Geekbench's opinion, this means they can't say that the results are the same/comparable (because Geekbench's output to the user is a score. Not an image or a model, etc). and that goes against the point of the benchmark, to give a "score" to a CPU.

Intel used IBOT on Geekbench to show that its not only for games. It can and presumably will be applied to popular productivity apps, at a later time. Similarly, Intel could potentially work with software developers and game developers, to bake in these optimizations, so that you don't need to specifically activate APO and IBOT. AMD did that sort of thing in the past year with Photoshop. And AMD now spanks intel in most portions of Photoshop.
Click to expand...
LukeTbk said:
geekbench results are pretty much all verifiable that ibot does not change the result (many of the benchmark have validation tools that the output is correct or not) I do not think that geekbench issues, a big one will be that no apps have an ibot code path made for them yet, that make the score unrepresentative.

An other one could be that code safety could be compromissed (with how robust the memory stack was made to resist attack vs going for speed), lot of geekbench is running a browser or displaying PDF, being a black box, it could give the exact same results but in a less secure way, which would be unrepresentative of code that choosen to be slow in exchange of more secure branching/stack decision.

If this is very similar to what Nvidia do with the just in time compilation of CUDA ptx, maybe they could do something similar in term of dev control, where you can say this need to stay like this and to avoid optimising it for speed and nvidia let you see the final optimised assembly it made to validate for yourself.
Click to expand...
what if someone gets ibot working for 265K and turns them into 265K Plus SKUs?

what would you expect the performance difference would shrink down to between the 270K Plus?

ibot’s gotta be worth few FPS at least, its just artificially locked out of other arrow lakes

and ibot approach + modest OC could even close the gap if not entirely
 
Last edited:
erek said:
what if someone gets ibot working for 265K and turns them into 265K Plus SKUs?

what would you expect the performance difference would shrink down to between the 270K Plus?

ibot’s gotta be worth few FPS at least, its just artificially locked out of other arrow lakes

and ibot approach + modest OC could even close the gap if not entirely
Click to expand...
Yeah, we'll see if modders can figure it out.

Intel insists that Arrow Lake Plus has a newer silicon revision, with "hardware hooks" which aren't present on previous Arrow Lake. But, that may not be exactly true.

That said, there is something about AL Plus, which is working better. Even before IBOT. As the 250k is basically always better than the 265K, in gaming. Despite 2 less P-Cores and lower turbo clocks. And its often not far off in multicore, either.
 
  • Like
Reactions: erek
like this
You must log in or register to reply here.
Back
Top