OcUK Ryzen 5 9600X and Ryzen 7 9700X review thread

[Tommy_Here]

I don’t know… people talking about the issues with the 7900x3d it seriously is a CPU I have got my eye on. Zen5 is a disappointment in my view - plus it would be good for the multi threaded performance… 90% of a 7800x3d gaming performance but 35-50% extra multi thread muscle I think I’ll be okay with…. I’m waiting till Black Friday but yes can’t come soon enough. 7900x3d for around £300-£330 I’d take.

 

[Gerard]

That is aside from those who are still trying to cope with these CPUs not being the massive gains certain people like MLID were touting... why anyone still uses the likes of certain YouTubers for information boggles my mind after the number of times they've been exposed as a shame, but I guess it fits with what they want to believe rather than the reality.

That's ALWAYS the way especially on here, be it cpu or gpu supposed rumours turn into "AMD\Intel\Nvidia said this or that" over the many pages of a thread. Then by the end you get people asking why AMD\Intel\Nvidia said it would be x amount faster when AMD\Intel\Nvidia said nothing of the sort.

This happens year in year out, launch after launch and people still trust these dribbling gimps on YouTube purporting to have 'sources' yet somehow nobody seems to notice the 'leakers' are basically firing out every possible combination of spec\perf increase for a product, then by the end they point to the 1 spec\perf increase of of 85 they 'reported on' that was vaguely close to reality, and people go 'yaaay you were right'.

It's like going to a medium, they throw out 50 names of a deceased relative before getting one right and people somehow get selective amnesia and forget the dozens of names that were wrong.

 

[KompuKare]

Not just the YouTube leakers though.
I know the pre-launch event was only a few weeks ago, but I'm hard pressed to see any of the AMD marketing slides from that event now in reality.

These is marketeers cherry-picking and then there's just making things up. This launch seems more like the latter.

Since Zen launched and the RDNA3 launch, AMD marketing has actually been pretty spot on often understating things.

Now? Well this week we found out that Radeon and Ryzen marketing slides have about the same credibility - which is not a lot!

 

[Perfect_Chaos]

That's ALWAYS the way especially on here, be it cpu or gpu supposed rumours turn into "AMD\Intel\Nvidia said this or that" over the many pages of a thread. Then by the end you get people asking why AMD\Intel\Nvidia said it would be x amount faster when AMD\Intel\Nvidia said nothing of the sort.

Except I'm almost certain a guy in AMD said he wishes he could show us how good Zen 5 is, and he said it within the last 1-2 years.. he hyped it to sounding better than the usual uplift, and it's looking like we don't even get that now.

 

[Emlyn_Dewar]

We saw some very reasonable looking leaks that showed pretty much the performance we are seeing now, but those ones weren't interesting enough. People only listen when it's some wild percentage increase.
I don't doubt that the 9000 series are much faster than the 7000 series in avx512.

 

[Gerard]

Except I'm almost certain a guy in AMD said he wishes he could show us how good Zen 5 is, and he said it within the last 1-2 years.. he hyped it to sounding better than the usual uplift, and it's looking like we don't even get that now.

Id be surprised if that ever happened, wouldn’t surprise me if it was some ‘leaker’ bs that got slapped in with all the other waffle and somehow became a thing that ‘amd said’ over the months or years.


And when it comes down to it, even if it did happen the mantra has always been, don’t trust the companies, wait for the reviews. The reviews are what matter, not the marketing pish.

 

[Grim5]

We saw some very reasonable looking leaks that showed pretty much the performance we are seeing now, but those ones weren't interesting enough. People only listen when it's some wild percentage increase.
I don't doubt that the 9000 series are much faster than the 7000 series in avx512.

Its a good thing then that thousands and thousands of games rely heavily on AVX-512... as long as its an emulator

 

[MartinPrince]

False. My Ryzen 9 pro chips are sub 9 watts idle and those are 16 cores all with SMT. As I said, AMD are years ahead.

Just because you say "False" doesn't make it so, facts and evidence do. Moreover I won't believe your 9 watts without you providing evidence as you seem to have a history of making stuff up.

Everybody here for the 7700x etc says idle power is ~20w and another user here for the 7950x states:

In fact the idle power consumption is reversed because of the chiplet design. A 13900k will idle well below 10W while 7950X is more in the 20-30W range due to the chiplet design. I/O die power consumption is better than on the previous Ryzen CPUs but it's still consistently 10-15W that's burned doing nothing.

If one of the most efficient CPUs currently in the 7800X3D idles at ~20w how can AMD be years ahead in idle power of say a 13700k that idles around ~8w or the 13600K which is IIRC ~5w. Even if what you said were true about the AMD Ryzen 9 PRO 7945 et al idling at 9w (though we see what you tried to do there comparing a workstation CPU with a gaming CPU, nice try!) then it would still not be years ahead of Intel it would be about the same or slightly more. So what you said - "AMD are years ahead of Intel in idle power", is still a provable falsehood.

FFS, why you need to make things up is beyond me.

Why would it be about idle power consumption the thing that you care about when your CPU is slower in games and uses 3X as much power when its doing anything? This is strawmaning to the highest degree.

That is such an inane statement I'm almost lost for words. "Strawmaning" would be arguing a point that Jigger didn't make whereas I am directly refuting the totally false point he made that "AMD are years ahead of Intel in idle power". His exact words not mine.

The fact is I personally don't care about idle power because it's peanuts at the end of the day; power consumption in use is of far more consequence which AMD totally dominate (my 5950x is a wonderful example of this) but that doesn't change the fact that Jiggers statement was totally false.
I could just let it go by and ignore it "because it's normal Jigger" but when it comes to a point of fact and not opinion then why should anyone let that misinformation go without setting the record straight?
This is how misinformation and lies get a foothold and before long you have people repeating his same made up lies and false information.

I can use the Trump analogy. He throws so much misinformation and lies in with facts that people are almost immune to it and say well that's Trump. So should people stop fact checking him and accept everything he says? Of course not.

In so many areas AMD are years ahead of Intel but idle power is not one of them and to say otherwise is a simple falsehood and should not be simply accepted as fact.

So if you agree with Jigger say so (providing evidence) or even better if you disagree then come from behind any AMD bias and show some objectivity and state that it's false but don't try to make this about something other than his original and initial false statement:
"AMD are years ahead of Intel in idle power" as that would really be "strawmanning to the highest degree"!

 

[Troezar]

Seems clear that AMD are milking the market whilst there's not a lot of competition. Intel and Nvidia taught them well Leaves them plenty of scope to up the performance at a later date if necessary.

 

[jigger]

Seems clear that AMD are milking the market whilst there's not a lot of competition. Intel and Nvidia taught them well Leaves them plenty of scope to up the performance at a later date if necessary.

Intel still destroy AMD. AMD aren’t completion for Intel.

 

[KompuKare]

Except I'm almost certain a guy in AMD said he wishes he could show us how good Zen 5 is, and he said it within the last 1-2 years.. he hyped it to sounding better than the usual uplift, and it's looking like we don't even get that now.

I also recall seeing all the recent pre-release interviews with Mike Clark - "father of Zen", chief Zen 5 architect - making big claims about Zen 5.

Now, being a CPU engineer on such a big redesign I'm sure he had plenty to be excited about but that may very well have little do to with overall end performance.

In danger of hyping Zen 6 onwards, one of the big goals of Zen 5 may have been to have a solid foundation to build on.

Also, internally Zen 6's goal might have been for Linux server code or the HPC market where recompiling is common.

 

[jigger]

I also recall seeing all the recent pre-release interviews with Mike Clark - "father of Zen", chief Zen 5 architect - making big claims about Zen 5.

Now, being a CPU engineer on such a big redesign I'm sure he had plenty to be excited about but that may very well have little do to with overall end performance.

In danger of hyping Zen 6 onwards, one of the big goals of Zen 5 may have been to have a solid foundation to build on.

Also, internally Zen 6's goal might have been for Linux server code or the HPC market where recompiling is common.

Zen was expected to be on 3nm and had to be pushed to 4nm and those two nodes have different characteristics. So far the two parts we’ve seen from the Zen5 stack look pretty good and time will tell for the rest.

 

[JediFragger]

Will be interesting to see if the X3D parts are on 3nm...

 

[jigger]

Just because you say "False" doesn't make it so, facts and evidence do. Moreover I won't believe your 9 watts without you providing evidence as you seem to have a history of making stuff up.

Everybody here for the 7700x etc says idle power is ~20w and another user here for the 7950x states:


If one of the most efficient CPUs currently in the 7800X3D idles at ~20w how can AMD be years ahead in idle power of say a 13700k that idles around ~8w or the 13600K which is IIRC ~5w. Even if what you said were true about the AMD Ryzen 9 PRO 7945 et al idling at 9w (though we see what you tried to do there comparing a workstation CPU with a gaming CPU, nice try!) then it would still not be years ahead of Intel it would be about the same or slightly more. So what you said - "AMD are years ahead of Intel in idle power", is still a provable falsehood.

FFS, why you need to make things up is beyond me.



That is such an inane statement I'm almost lost for words. "Strawmaning" would be arguing a point that Jigger didn't make whereas I am directly refuting the totally false point he made that "AMD are years ahead of Intel in idle power". His exact words not mine.

The fact is I personally don't care about idle power because it's peanuts at the end of the day; power consumption in use is of far more consequence which AMD totally dominate (my 5950x is a wonderful example of this) but that doesn't change the fact that Jiggers statement was totally false.
I could just let it go by and ignore it "because it's normal Jigger" but when it comes to a point of fact and not opinion then why should anyone let that misinformation go without setting the record straight?
This is how misinformation and lies get a foothold and before long you have people repeating his same made up lies and false information.

I can use the Trump analogy. He throws so much misinformation and lies in with facts that people are almost immune to it and say well that's Trump. So should people stop fact checking him and accept everything he says? Of course not.

In so many areas AMD are years ahead of Intel but idle power is not one of them and to say otherwise is a simple falsehood and should not be simply accepted as fact.

So if you agree with Jigger say so (providing evidence) or even better if you disagree then come from behind any AMD bias and show some objectivity and state that it's false but don't try to make this about something other than his original and initial false statement:
"AMD are years ahead of Intel in idle power" as that would really be "strawmanning to the highest degree"!

I take it you have googled really hard and the realisation I’m right has really upset you. The 7800X3D isn’t even close to most efficient Zen parts.

The issue with constantly talking nonsense is Intel will probably listen.

 

[Rroff]

I take it you have googled really hard and the realisation I’m right has really upset you. The 7800X3D isn’t even close to most efficient Zen parts.

The issue with constantly talking nonsense is Intel will probably listen.

14th gen can be tuned down to less than 2 watt idle (without any features disabled) for the CPU - FACT - nominally around 4-5 watt, you can't realistically get the SoC on any recent Zen parts below ~5 watt plus another ~3+ watt for the CPU and that is with features disabled. Anything sub around a dozen watts isn't realistic for Zen 4, Zen 3 typically idle half of Zen 4 but they still can't touch Intel (they still have around the same minimum as Zen 4 if tuned).

Some people have Zen 4 parts which are impossible for some reason to get below ~35 watt idle due to the IF and SoC voltages.

That said whole system idle power consumption at the wall isn't hugely different, in fact can even go the other way sometimes depending on motherboard as some have poor idle power management.

 

[Scotster]

Was the launch delayed across the globe? I know there was an initial delay with the release to the 8th August, but did anywhere in the UK actually have any for sale.... or was the 8th of August always representative of a pre-order date?

Just wondering if anyone physically has their hands on any retail yet.

 

[jigger]

14th gen can be tuned down to less than 2 watt idle (without any features disabled) for the CPU - FACT - nominally around 4-5 watt, you can't realistically get the SoC on any recent Zen parts below ~5 watt plus another ~3+ watt for the CPU and that is with features disabled. Anything sub around a dozen watts isn't realistic for Zen 4, Zen 3 typically idle half of Zen 4 but they still can't touch Intel (they still have around the same minimum as Zen 4 if tuned).

Some people have Zen 4 parts which are impossible for some reason to get below ~35 watt idle due to the IF and SoC voltages.

That said whole system idle power consumption at the wall isn't hugely different, in fact can even go the other way sometimes depending on motherboard as some have poor idle power management.

First off see post 153, secondly stability would be questionable a 2 watts as would measuring power use. If Intel made a single cluster all Atom part with minimal IO and qualified the chip to run that low, maybe.

The fact is Zen architecture would be better and to claim Intel are ahead is false. What uscool probably should have said is these two particular Ryzen* chips pull more power at idle than said Intel parts.

 

[Rroff]

First off see post 153, secondly stability would be questionable a 2 watts as would measuring power use. If Intel made a single cluster all Atom part with minimal IO and qualified the chip to run that low, maybe.

The fact is Zen architecture would be better and to claim Intel are ahead is false. What uscool probably should have said is these two particular Ryzen* chips pull more power at idle than said Intel parts.

There is a review here showing whole system idle https://www.guru3d.com/review/amd-ryzen-7-8700g-processor-review/page-4/ though they don't include Intel parts equivalent to the 8700 for some reason and some motherboards are much more efficient than others - my 14700K on the Aorus Master is about 10 watt lower at stock than that article.

Sub 2 watt idle for the whole CPU package is measured the same way as AMD for the whole CPU package so is comparable and is completely stable.

As I've posted before this is an actual real 14700K idle wattage and completely stable with no features disabled:

You won't get close to that with Zen for the package power though as above whole system draw is another story. Probably possible to beat that as my chip doesn't like undervolting.

But as above Intel you can tune down to much lower power consumption, completely stable with all features enabled. The IF/SoC are stumbling blocks on Zen when it comes to pure idle optimisation.

 

[jigger]

There is a review here showing whole system idle https://www.guru3d.com/review/amd-ryzen-7-8700g-processor-review/page-4/ though they don't include Intel parts equivalent to the 8700 for some reason and some motherboards are much more efficient than others - my 14700K on the Aorus Master is about 10 watt lower at stock than that article.

Sub 2 watt idle for the whole CPU package is measured the same way as AMD for the whole CPU package so is comparable and is completely stable.

But as above Intel you can tune down to much lower power consumption, completely stable with all features enabled. The IF/SoC are stumbling blocks on Zen when it comes to pure idle optimisation.

Maybe, email Intel with your findings as they might need the help. Shrugs.

 

[Rroff]

Maybe, email Intel with your findings as they might need the help. Shrugs.

Did it really need the petulant stomping feet reply?

Fact is if/when you include all the low desktop power and laptop derived chips if you really want to go down that route for the actual CPU power Intel is better, when you include motherboard and RAM, etc. the whole system power consumption at the wall isn't much different and comes down to PSU and motherboard efficiency.