Home » Core 13 vs Ryzen 7000, consumption and efficiency, which CPU is better?
Technology

Core 13 vs Ryzen 7000, consumption and efficiency, which CPU is better?

Yesterday we launched an article on the Zen 4 architecture and the changes that are going to take place in its processors thanks to the presentation of AMD and the data leaks. While this is a preview of the final article, which we’ll update in due course, we’ve already given some very interesting hints of what’s to come and also the issues any processor with the Zen 4 architecture can be faced. this problem, because as for consumption and efficiency per mm2, the Ryzen 7000 they don’t do well.

For some time now and if we look back we will throw our hands in our heads thinking about the consumption we had in the past and the ones we have and are going to have now. The biggest problem is that in the Intel vs. AMD war for processors, both companies are pushing each other to outdo each other and it’s reaching a limit where the level of cooling is going to be decisive.

Intel Core 13 against Ryzen 7000, TDP, PPT and PL2 with their zones on stage

The data presented indicates fairly clear information, although little comparison due to the small number of models specified. You still have to read between the lines to understand what is going on here.

In Zen 4 we can see 5 digits of watts with their respective areas per CCD and the efficiency and ratio of watts per square millimeter. In Intel models are specified, where both have their maximum consumption in PL2 TVB status as well as the whole area of โ€‹โ€‹the matrix, because it is not a MCM Architecture like AMD’s.

That said, the data 105 watt and 130 watt as well as its surface area logically refer to the TDP and PPT of the Ryzen 5 7600X and Ryzen 7 7700Xwhile the data from 170W and 230W with two identical zones refer to the Ryzen 9 7900X and Ryzen 9 7950X. Finally, the 280W are keys, because they indicate the overclock limit for the common user depending on the allowed SoC power.

That’s to say, AMD will allow 50 watts of overclocking as a safety limit for their processors something that seems like a lot but seeing frequencies and temperatures will be more than short for sure which is why the theory that they are squeezed near the electrical, thermal and so limit is reinforced. Of course, frequently. The voltage wall has a lot to say here, but it will be much better when the first non-pro overclocking data arrives.

AMD Ryzen 7000 Zen 4, lower performance per mm2 and general consumption?

AMD-Ryzen-7000

The calculations are very simple and reveal that Intel has degraded its efficiency despite having a larger area with its new Core 13. 350 watts they’re a tough wall to climb, but considering the total size of the die and talking about the Intel Node 7, the efficiency isn’t bad at all, scoring 1.36 watts per square millimeter.

Worse data than in Core 12 yes, but given the L2 and L3 cache jump as well as the higher number of cores and the increase in frequencies, we can say that this is more or less justified, because area increased by 19%while the ratio did so in a twenty one%. More than optimal scalability.

Opposite we have AMD, where the highest performance is achieved by processors with two CCDs and in your TDP, that is to say in its base frequency, something logical and which is not comparable because those of Intel are in PL2, that is to say with its frequency and its maximum consumption authorized by the brand. If we equalize the contention and go up to the 230 watts with it, what we will find is that AMD gets an efficiency per area of โ€‹โ€‹1.61W/mm2.

In other words, comparatively speaking and fully loaded, AMD is a 18% more inefficient than the Intel option, or the like, the Ryzen 9 7950X is less efficient per square millimeter than the i9-13900K. What does this imply beyond this superfluous comparison for the average user?

The problem of pushing to the limit

AMD-Zen-4-Ryzen-7000-CPU-die

Well, AMD’s worst efficiency has a second handicap and it is precisely the smallest total area of โ€‹โ€‹IHS. The Ryzen 7000s are smaller overall than the Ryzen 5000s and as we’ve seen, efficiency has deteriorated. Added to this is that due to the design of the PCB and the dies, CCDs are togetherstuck, which aggravates the Tjmax general and therefore AMD had to modify the PBO algorithm to allow a maximum temperature to be reached without lowering the frequencies.

That is to say, it has chosen the path of Intel, and for this reason we will no longer see temperatures hovering around 70 ยบC or having better frequency like better cooling. Now it all depends on consumption and temperature, where if we reach the limit set by AMD, we will start to lose performance. This particular number is 95ยบCso let’s hope Lisa Su has the right voltage and frequency hopping, and MHz balance between cores to avoid going to those extremes.

In short, the Ryzen 7000 seems not only to lose the performance battle (we’ll see in gaming) but also in efficiency per mm2 (no efficiency per performance achieved). The positive point is that in consumption they will be ahead by remaining at 230 watts against the 350 watts of the Core i9-13900K.

About the author

admin

Add Comment

Click here to post a comment

Your email address will not be published. Required fields are marked *