In an embarrassing presentation on Friday morning, Intel chief performance strategist Ryan Shrout presented a group of tech reporters with the aim of taking AMD’s Zen 2 (Ryzen 4000 Series) laptop CPU under a peg.
Intel’s newest laptop CPU design, Tiger Lake, is actually a compelling release – but it comes on the heels of some crushing upsets at that location, giving Intel an angle to block its rival’s market share Looks for Early Tiger Lake systems performed incredibly well — but they were configured for 28W cTDP, rather than the usual 15W TDP seen in production laptop systems — and critics were barred from testing battery life Was.
Left-wing reviewers like you have compared Intel’s i7-1185G7 to 28W cTDP at half the power consumption of the AMD Ryzen 7 system — and although Tiger Lake generally came out on top, the power discrepancy made it AMD’s decider. Or saved from decisive trauma. Market share is increasing with OEM vendors who are actually Buying Laptop CPU in the first place.
Intel’s original Tiger Lake launch presentations sought to draw attention to the battery versus off-battery discrepancies in AMD’s performance, but those efforts mostly went unheard. Shrout’s presentation was an attempt to retell that story on Friday, this time with enough additional information to evacuate people.
We can easily see this discrepancy between battery and off-battery performance in the PCMark 10 application benchmark, as well as many of Intel’s Rugs-scripted workloads that are based around production applications, which the company calls a “realistic usage ride” is. However, a similar discrepancy between on-off and off-battery performance does not appear in commonly used industry benchmarks, such as Cinebench, Passmark, or Geekbench.
Intel’s engineering team explains the reason Why the We don’t see a discrepancy in the CineBench in the final image of the gallery above – in Intel’s test, the Ryzen 4000 CPUs didn’t ramp up power and voltage to their maximum position anywhere between a maximum of 11 seconds. After The heavy workload began.
We were able to confirm Intel’s findings over the weekend, working with the Acer Swift 3 SF314-42 laptop (with a Risen 7 4700 CPU) and an MSI Prestige 14 Evo laptop (with a Core i7-11852-2) Were staying. In the chart above, we repeatedly compress small portions of the Linux 5.3 kernel source and graph throughput on each CPU.
The 4-core / 8-thread i7-1185G7 easily transforms the 8-core / 8-thread Risen 7 4700u into both single and quad-thread workloads, even after the Ryzen 7 4700u has its full performance around 12-seconds Emphasis for Mark. In unlimited workloads, where the Raizen 7 is allowed to flex its full octa-core muscles, things are much closer – and the 4700u evens out a narrow win in the last four seconds.
However we need to mention a few things here. First and foremost – Intel is 100 percent correct in its claims that AMD’s Zen2 laptop CPUs delay ramping power and voltage to their maximum states. This causes a faster, consistent, and reduced performance during those first few seconds.
We reached out to AMD representatives to comment on this design decision. Although AMD representatives asked further questions about our comments, we have not yet received a response to the record at press time.
The devil is in the details – it’s so hot
But Intel is still playing games with its power consumption. In the screenshot above, we can see the MSI Prestige Evo 14 with a Core i7-1185G7 during a Cinebench R23 run. We don’t have enough to review this laptop in its entirety – and in particular, to review its battery life, which we’re very curious about because it’s forbidden to test that stat two First i7-1185G7 system.
But we can see that instead of dialing the cTDP of i7-1185G7 bottom To estimate the typical Ryzen 7 4000 cTDP of 15W, as widely expected — MSI chose to dial it into this laptop that we saw in earlier prototypes. This production i7-1185G7 system has a variable PL1 that hits as high as 36W during the Cinebench R23 run – 51W in addition to its PL2, which is unchanged from the prototype.
During this Cinebench R23 run, the laptop has its first 10 to 15 seconds full PL2 power limit running at 51W, with a blistering temperature of up to 98 ° C. After the initial, extremely high-performance, power and heat-generating burst, the CPU dropped down to maintain an average power consumption of 34W. In contrast, an 8 core / 16 thread Ryzen 7 Pro 4750U consumes an average of 27.9W, with a cTDP of 25W – higher than 29.9W.
When we’re distancing ourselves from the CPU and in the laptop-design realm, it’s probably worth noting that system-fan activity also differed significantly between the MSI Prestige 14 Evo — which was almost immediately the gaming-laptop level. Arrived at. And the HP EliteBook, which took more than a minute on the ramp to maximize its fans and was much quieter than MSI during the entire run.
The fight is on
While Intel didn’t specifically tell us what we should conclude from Tiger Lake in immediate-on-performance versus Zen 2 laptop CPUs, performance delays, the company is clearly “AMD Gaming the Benchmark” and something in between. Was hoping for. ” It turns out, Intel was the winner all along. “
We do not think there is such a cut-and-dry conclusion to draw here. Intel’s conclusions about the slow performance ramp of the AMD Zen2 laptop CPU are, frankly, true in the facts – we had no trouble confirming this, and point out that many of Intel’s favorite benchmarking techniques were Team Blue Large performance delays in favor of. Compared to widely used industry benchmarks such as Cinebench, PassMark, and so forth.
But this ignores the greater efficiency of AMD systems above and beyond the delayed shift in maximum performance (and battery consumption) states in the CPU. When we run the Cinebench R23 for a full five minutes, a Ryzen 7 Pro 4750u system offers more visuals than an Intel i7-1185G7. And It does this with low total power consumption. There is no clever trick to explain that away.
We also believe that both sides have a tuning logic. Intel’s more rapid shift to the highest-performing state carries with it some real-world benefits, but we’re not sure they’re as attractive as the charts seem to them. In practice, we’ve spent a lot of time with both the Zen2 and Tiger Lake laptops – and the Tiger Lake system isn’t really Feel Fast in terms of seat-of-the-pants subjective experience. It argues strongly that it often makes no sense to ramp up the CPU power profile, which is probably better to preserve the battery instead if the agar-system does not notice an improvement in human pilot latency.
… and continues
The best news for consumers, we suspect, is that the “which system is better” argument is so difficult to answer conclusive in the first place. This level of competition means that neither team gets to rest on their praise, and consumers are less likely to buy the system None Want to be informed about the differences completely.