Intel Core i9-13900K and i5-13600K Review: Raptor Lake Brings More Bite
by Gavin Bonshor on October 20, 2022 9:00 AM ESTCore-to-Core Latency
As the core count of modern CPUs is growing, we are reaching a time when the time to access each core from a different core is no longer a constant. Even before the advent of heterogeneous SoC designs, processors built on large rings or meshes can have different latencies to access the nearest core compared to the furthest core. This rings true especially in multi-socket server environments.
But modern CPUs, even desktop and consumer CPUs, can have variable access latency to get to another core. For example, in the first generation Threadripper CPUs, we had four chips on the package, each with 8 threads, and each with a different core-to-core latency depending on if it was on-die or off-die. This gets more complex with products like Lakefield, which has two different communication buses depending on which core is talking to which.
If you are a regular reader of AnandTech’s CPU reviews, you will recognize our Core-to-Core latency test. It’s a great way to show exactly how groups of cores are laid out on the silicon. This is a custom in-house test, and we know there are competing tests out there, but we feel ours is the most accurate to how quick an access between two cores can happen.
(Click on image to enlarge)
Looking at core-to-core latencies going from Alder Lake (12th Gen) to Raptor Lake (13th Gen), things look quite similar on the surface. The P-cores are listed within Windows 11 from cores 0 to 15, and in comparison to Alder Lake, latencies are much the same as what we saw when we reviewed the Core i9-12900K last year. The same comments apply here as with the Core i9-12900K, as we again see more of a bi-directional cache coherence.
Latencies between each Raptor Cove core have actually improved when compared to the Golden Cove cores on Alder Lake from 4.3/4.4 ns, down to 3.8/4.1 ns per each L1 access point.
The biggest difference is the doubling of the E-cores (Gracemont) on the Core i9-13900K, which as a consequence, adds more paths and crossovers. These paths do come with a harsher latency penalty than we saw with the Core i9-12900K, with latencies around the E-cores ranging from 48 to 54 ns within four core jumps between them; this is actually slower than it was on Alder Lake.
One possible reason for the negative latency is the 200 MHz reduction in base frequency on the Gracemont cores on Raptor Lake when compared with Alder Lake. When each E-core (Gracemont) core is communicating with each other, they travel through the L2 cache clusters via the L3 cache ring and back again, which does seem quite an inefficient way to go.
Facebook
Twitter
RSS
169 Comments
View All Comments
flyingpants265 - Thursday, October 20, 2022 - link
That doesn't matter. All that proves is TDP is a phony measurement. If the CPU draws up to 300 watts, then it's a 300 watt CPU. Replyyh125d - Friday, October 21, 2022 - link
Exactly ReplyIketh - Friday, October 21, 2022 - link
proving TDP is a phony measurement is the entire point of that post ReplyYojimbo - Friday, October 21, 2022 - link
Firstly this discussion is not confined to Intel. All the modern CPUs use turbo clocks. They all have various performance characteristics dependent on the thermal design of the product they are in.Please cite where Intel writes that. Intel only uses TDP in its technical literature these days for the very reason that consumers are confused about it. Intel uses PL1 and PL2. TDP is the MINIMUM power that one should be designing for, not the maximum. The amount of turbo clock exposed by the cooling solution is optional, but the thermal solution associated with the processor must be capable of handling the TDP. The processor will not be damaged with a cooling solution that only handles the TDP. The processor will not use its turbo clocks much and will stay at or below the TDP power except for short periods of time. On the other hand if a cooling solution cannot handle the TDP there could he bad consequences.
Again. This isn't an Intel-specific thing. TDP and turbo clocks are ubiquitous in the industry. What is also very widespread is massive misunderstanding and misinformation about the term. Perhaps Anandtech should stop using the term with respect to CPUs because it seems to me that it's a minority of readers who understand it. Reply
Meteor2 - Saturday, October 22, 2022 - link
Reviews should stop quoting TDP. Intel no longer uses it; their latest product spec pages e.g. for the i9-13900K quote Maximum Turbo Power: "The maximum sustained (>1s) power dissipation of the processor as limited by current and/or temperature controls. Instantaneous power may exceed Maximum Turbo Power for short durations (<=10ms). Note: Maximum Turbo Power is configurable by system vendor and can be system specific."Which for the i9-13100K is 253W. Reply
Meteor2 - Saturday, October 22, 2022 - link
AMD still quotes TDP (e.g. 170W for the 7950X) with no definition of TDP provided, which I would suggest IS misleading. Replyat_clucks - Monday, October 24, 2022 - link
Ah, not confined to Intel, solid argument that it's not a problem to do it but that "people are uneducated". Scale matters. When your real power consumption is 120% over the advertised one (see link below) this isn't an "everybody's doing it" but it is indeed a matter of "people are uneducated". At this time Jimbo, anyone trying to find excuses for Intel, and downplaying the shenanigans is _really_ uneducated, was born yesterday, or benefits from the lie.This doesn't mean you should stop using Intel if it does the job for you,. But only a fool or the fraudster would defend or downplay what they're doing.
https://images.anandtech.com/graphs/graph17585/130... Reply
catavalon21 - Sunday, November 20, 2022 - link
"Please cite where Intel writes that."Step right up, folks...
https://www.intel.com/content/www/us/en/support/ar... Reply
Truebilly - Friday, October 21, 2022 - link
🫳🎤 ReplyHarryVoyager - Friday, October 21, 2022 - link
Doesn't especially matter whether they are conforming to the technical definition or not as it is tells me nothing useful about the CPU in the context in which it is presented. Reply