- Mar 3, 2017
- 1,747
- 6,598
- 136
Why do you always jump to anti-AMD conspiracy theories? I doubt he’s maliciously “tricking” the numbers to make AMD look bad.
SoC power is directly tied to memory kit and motherboard defaults. One could be using an Asus, the other Asrock both have different default vSoC target voltages. There’s a bunch of other explanations than a conspiracy.
Looking over the variance in some of the reviews, I would surmise the lower TDP (65W) is hurting some results more than others.
Why? Because it looks like some boards are putting unnecessarily high voltage on some rails (e.g. vSOC 1.25V+) which would subtract that much more power budget from the cores.
Your uncore is a decent percentage of 65W (88W PPT) while at higher power budgets the percentage becomes less significant. Seeing 9600X hit higher clocks both ST and MT versus 9700X shows the lower power budget is really hurting Zen 5 reviews and crimping scores (esp 9700X). Or leaving a lot more OC headroom (20%+ in some cases!) depending on your perspective.
I expect to see this clarified a bit next week when we see what Zen 5 will do out of the box for the higher TDP parts, especially the 9950X.
PC World pulled their launch day review because of this. They don't even know this is the reason.
They talked about how they got mysteriously 10% worse MT performane than the AMD reference data (and other reviews that came out), and after working with AMD to troubleshoot, the issue was enabling EXPO set VSOC to 1.25V when it should have been set to 1.2V. They expressed confusion at why or how this would have caused it, but confirmed that they are getting results more in line with what they should be after correcting it.
My analysis is this was causing the IOD power consumption to increase and consume more of the core power budget, hurting perf. This was especially apparent because of how power starved the cores are in the stock configuration..
I would go 5800X3D I did from a 5800X and had a 6800XT and it was a noticeable difference.And I still dunno if my 5600X will be replaced by 5800X3D or not. Probably not. Not in a hurry to upgrade to anything lol.
Those benchmarks results really don't try me to upgrade to AM5....
Surely launch pricing is an attempt to capitalize on early adopters. I expect at least a 10% haircut on ALL SKUs prior to Arrow Lake launch / X3D launch.I found it:
7600X:70
7700X:270
7900X:230
7950X:280
Indeed 850 for All 4. And this was already considered as very bad, ZEN4 sold like 1:5 to ZEN3 for the first 3-4 months until the price cuts came. Platform was expensive and gaming only on par with 5800X3D.
The worst part is that launch week sales aren't evenly spread. The majority comes during first hours because enthusiasts are waiting for the products. I can't recall it 100%, but I think ZEN4 had like 50, 200, 180, 210 at the first day and sales start at 3pm here in germany. So my guess is at Sunday evening both ZEN5 SKUs will still be below 20, 9600X maybe even below 10 depending on how many sold right now.
Yes, I'm also expecting price cuts.Surely launch pricing is an attempt to capitalize on early adopters. I expect at least a 10% haircut on ALL SKUs prior to Arrow Lake launch / X3D launch.
They won't be launching higher count CPUs on AM5. Broken record but people keep suggesting an idea without meritYes, I'm also expecting price cuts.
Seems like the conclusion has been that people are currently not prepared to may much more for additional performance, so AMD tried to optimize perf/price with Zen5. Some milking of early adopters of course, but then price drops due to Arrow Lake DT, Zen5 X3D, etc.
The question is how long to wait until most of the price drops have materialized. 1, 3, 6, ... months?
Also, possibly AMD has something more in store for later before Zen6. Not sure what, but e.g. 24/32C SKUs. Or new IOD with better iGPU, NPU, and/or higher RAM speeds. But if so, that's likely 12+ months away.
I just cannot see that this is all we'll get with Zen5, so it'll be 4-5 years between Zen4 and Zen6 until there is meaningful perf increase, and nothing else to warrant purchasing in that time frame. That is unless they are counting on Win10 becoming obsolete soon (EOL in 2025) and people being forced to upgrade to Win12 compatible CPU, so that'll drive upgrades anyway. Therefore they can save perf improvements for Zen6, when that will be needed to drive upgrades in the following cycle.
Does something prevent it?They won't be launching higher count CPUs on AM5. Broken record but people keep suggesting an idea without merit
Who knows, we'll see. It's just an option, but something will be needed anyway. 4-5 years between Zen4 and Zen6 with nothing else than what we've seen so far just won't do. That is unless we'll see substantial price cuts along the way.They won't be launching higher count CPUs on AM5. Broken record but people keep suggesting an idea without merit
Of course since efficiency is traded for more performance than the 7700, but if you set the 9700X at same perf than the 7700 it would consume about 60W vs 80W for the 7700, so 30% better efficency at isoperf, that s remarkable given that the core is much bigger and the node jump relatively modest.Zen 5 also doesn’t seem much more efficient than Zen 4 like everyone is saying when we compare the 9700x and the 7700.
View attachment 104954View attachment 104953
Not really. It has a 5% advantage in pts/w over the 7700 on CB nT.Of course since efficiency is traded for more performance than the 7700, but if you set the 9700X at same perf than the 7700 it would consume about 60W vs 80W for the 7700, so 30% better efficency at isoperf, that s remakable given that the core is much bigger and the node jump relatively modest.
Not really. It has a 5% advantage in pts/w over the 7700 on CB nT.View attachment 104960
I completely doubt that graph.At stock the 9700X use 60W in CB 2024.
N3E would've brought 100-200 MHz tops, that would not have made any difference.don’t fool into thinking zen 5 is more efficient
Zen 5 vanilla is a flop for gaming and higher clocks would have definitely helped and oh N3E
I completely doubt that graph.
That's not the full CPU, just the cores ("CPU Core Power" in HWiNFO). The headline translates to "Power consumption of the cores...".At stock the 9700X use 60W in CB 2024.
No, the 60W number is for the cores, without SOC. Add the 27W and some change for the correct 88W PPT in total.We ll have to wait for confirmations, what is odd is that Computerbase had the 9700X idle power at 27W, wich is excessive, and if the CPU use only 60W then it means that the cores are fed with only 33W.
yep. Look at PC Watch, it adds up. No need to wait.No, the 60W number is for the cores, without SOC. Add the 27W for 87W in total.
They measure power at the 12V CPU power rail, so that include the VRMs losses wich are about 10-12%, so their CB run is at 65W or so, that s comparable to Computerbase s 60W.The TPU review has the 7700 using slightly less power pretty much across the board. Both are using more than 60W in CB 2024 nT.
That s the full Soc power since they get about 90W peaks in Blender, previously they always used the CPU package power, wich include the uncore.No, the 60W number is for the cores, without SOC. Add the 27W and some change for the correct 88W PPT in total.
And yes, the 27W indeed seems excessive, considering they used RAM at JEDEC timings, so the memory controller should not be in OC mode.