Intel Core i5 13400 Linux Performance - Raptor Lake 10 Cores / 16 Threads For $239

[phoronix]

Phoronix: Intel Core i5 13400 Linux Performance - Raptor Lake 10 Cores / 16 Threads For $239

Earlier this month at CES, Intel announced the 35 and 65 Watt Raptor Lake processors alongside the 13th Gen Core Mobile CPUs. Those new Raptor Lake desktop CPUs are now available from multiple Internet retailers and the mid-range Core i5 13400 recently arrived at Phoronix to put it through its paces under Linux.

https://www.phoronix.com/review/intel-core-i5-13400

[milkylainen]

There is one thing that has annoyed me for a while now (personal usecase).
AMD systems usually has higher idle power consumption.
Which is a real shame, since they do really well under load.
As a home user, my stuff tends to be up 24/7, mostly idling and keeping various bounces, irc-sessions and whatnot up.
The idle consumption thus becomes a significant portion of the total energy consumption.
Because fully loading it doesn't really happen often or for that long.

So I really wish they'd do a bit of polishing on the idle consumption.

[pWe00Iri3e7Z9lHOX2Qx]

The i5-13500 may end up being the most interesting Intel chip of this generation for a lot of people (and OEMs). It was supposed to be a ~$20 USD step above the i5-13400, but that nets you another 4 efficiency cores. 14 cores and 20 threads at that price isn't too shabby.

[stormcrow]

At this point, where we're at the point of 32 GB+ of system RAM, I'm becoming less concerned about raw performance than I am about data integrity and Intel's stubborn refusal to support ECC RAM on their desktop class CPUs while AMD does - so long as you pick a motherboard that does as well.

[Michael]

The i5-13500 may end up being the most interesting Intel chip of this generation for a lot of people (and OEMs). It was supposed to be a ~$20 USD step above the i5-13400, but that nets you another 4 efficiency cores. 14 cores and 20 threads at that price isn't too shabby.

For the stores where I can find the i5-13500 in stock, it's closer to $300.

[jrch2k8]

Well, giving that is basically slower than the 7600 and this Intel mobo generation dies with raptor lake AFAIK, i think AM5 do makes more sense here

[brucethemoose]

There is one thing that has annoyed me for a while now (personal usecase).
AMD systems usually has higher idle power consumption.
Which is a real shame, since they do really well under load.
As a home user, my stuff tends to be up 24/7, mostly idling and keeping various bounces, irc-sessions and whatnot up.
The idle consumption thus becomes a significant portion of the total energy consumption.
Because fully loading it doesn't really happen often or for that long.

So I really wish they'd do a bit of polishing on the idle consumption.

Its a high level design tradeoff for the multi chip thing.

But (if you dont want to buy an AMD APU) it should get better as packaging tech gets better (which both Intel and TSMC have in the pipe).

[Cooe]

This chip is NOT Raptor Lake based, regardless of the 13th Gen branding! This SKU (and everything in the 13th Gen below the i5-13600K) is an Alder Lake rebrand, nothing more, nothing less. Stop regurgitating Intel's BS lying propaganda that this chip uses Raptor Lake (aka Raptor Cove cores) please!

[otoomet]

Michael , how does the motherboard power configuration look like?

I am surprised that the power consumption is so low. overclocking.com (https://en.overclocking.com/review-i...re-i5-13500/8/) reviews i5-13500, which is basically the same chip, just with 8 E-cores instead of 4. They find that the total power consumption (at wall) is 70-75W larger than for Ryzen 7600 (non-X). That particular graph is about Cinebench 23, but I do not think that the multi-core tests here are that different in terms of pushing all the cores. Even more--what does the "maximum turbo power" of 150W even mean if the cpu only sips a half of that amount?

Is there something going on with the power measure (intel self-reported) versus at wall? Or something with the chipset? Or do I misunderstand something?

[pWe00Iri3e7Z9lHOX2Qx]

Is there something going on with the power measure (intel self-reported) versus at wall? Or something with the chipset? Or do I misunderstand something?

You are misunderstanding that the other site is running benchmarks with max / unlimited power settings and water cooling. The sustained 4.5GHz P / 3.5GHz E all core clocks they are showing are only possible when you remove the power limits. With the default 65W PL1 setting, they'd only be running at ~3GHz.