First Ryzen 7 iMac benchmarks?

[Dayman1225]

Stock Vs Stock is in those images too.
8700k - 96w - 23.7 minutes
1700 - 80w - 32.9 minutes

 

[TahoeDust]

Stock Vs Stock is in those images too.
8700k - 96w - 23.7 minutes
1700 - 80w - 32.9 minutes

I was just looking at the same thing. 38% higher performance for 20% more power consumption. Hmmm...

 

[FIVR]

Stock Vs Stock is in those images too.
8700k - 96w - 23.7 minutes
1700 - 80w - 32.9 minutes

I don't have the free time to go cherry-pick graphs to refute you. I'm sure somebody else will soon enough. If you want to believe the 8700k is more efficient... be my guest. You won't convince me or anybody else here who actually has used a Ryzen or even just read the Anandtech review of the 1700.



I encourage people reading this thread to go read the anandtech article on Ryzen and efficiency make their own decisions. While you are at it, read the review of the 8700k and check power consumption. Definitely do not take cherry picked graphs from Dayman without a metric ton of salt.

 

[PhonakV30]

Stock Vs Stock is in those images too.
8700k - 96w - 23.7 minutes
1700 - 80w - 32.9 minutes

1700 running at 3Ghz on all cores while 8700K running at 4.3Ghz on all cores.

 

[Dayman1225]

Jesus christ someone is jumping to assumptions eh? I was simply pointing out STOCK VS STOCK was on that graph so your little rant about them "NOT BEING COMPARABLE" IS false. You really don't like it when someone disproves you, eh?

I'm running a 1700 my self I have no reason to shill or cherry pick for Intel.

1700 running at 3Ghz on all cores while 8700K running at 4.3Ghz on all cores.

He asked for STOCK VS STOCK and thats what HE GOT

 

[PeterScott]

Been reading "how to lie with statistics" I see....
You are comparing an OC 1700 running way outside its efficiency peak to a stock 8700k. Nobody is fooled by this.

That is exactly what Ryzen Fans do when making their silly claim that 8700K uses twice as much power.

Which is exactly the nonsene I was replying to.

If you look at stock vs stock, 8700K has better perf/watt.

Not twice as bad as some Ryzen fans are claiming.

 

[PhonakV30]

He asked for STOCK VS STOCK and thats what HE GOT

It's still stock , What's different?

 

[Dayman1225]

That is exactly what Ryzen Fans do when making their silly claim that 8700K uses twice as much power.

Which is exactly the nonsene I was replying to.

If you look at stock vs stock, 8700K has better perf/watt.

Not twice as bad as some Ryzen fans are claiming.

I'd love to live in the magical place where Ryzen uses half the power while preforming the same! Would be euphoric

 

[Dayman1225]

It's still stock , What's different?

@FIVR Complained that @PeterScott pointed out OC'ed 1700 vs Stock 8700k in blender was less efficient than the 8700k, so I pointed out STOCK VS STOCK scores and power consumption where the 1700 is still less efficient, seems he didn't like that.

 

[FIVR]

@FIVR Complained that @PeterScott pointed out OC'ed 1700 vs Stock 8700k in blender was less efficient than the 8700k, so I pointed out STOCK VS STOCK scores and power consumption where the 1700 is still less efficient, seems he didn't like that.

That is exactly what Ryzen Fans do when making their silly claim that 8700K uses twice as much power.

Which is exactly the nonsene I was replying to.

If you look at stock vs stock, 8700K has better perf/watt.

Not twice as bad as some Ryzen fans are claiming.

So explain why the Anandtech bench says the 1700 wins in every single power metric except idle power.

https://www.anandtech.com/bench/product/2016?vs=2025



I only wish I could live in some euphoric intel dream world like you guys who believe the 8700k uses less power than the 1700. It is literally a dream world in which your arguments are made. It must be wonderful to have no competition in x86.

 

[Dayman1225]

You're showing me 5 power consumption tests that don't have a ending (Work done - I.E points to represent score) so we can't exactly draw efficiency from that, can we?

 

[dahorns]

So explain why the Anandtech bench says the 1700 wins in every single power metric except idle power.

https://www.anandtech.com/bench/product/2016?vs=2025



I only wish I could live in some euphoric intel dream world like you guys who believe the 8700k uses less power than the 1700. It is literally a dream world in which your arguments are made. It must be wonderful to have no competition in x86.

I dont think anyone has made the claim that the 8700k uses less instaneous power at its peak.

The power numbers you are comparing relate to peak power consumption, not work efficiency. Yes, the 8700k consumes more power at its peak, but it also does more work. Until you measure both power consumption AND performance, you arent really demonstrating anything. In the tests I have seen that have actually done that, the 8700k generally comes out ahead except in very highly threaded scenarios (e.g., cinebench if I recall).

 

[FIVR]

Here are some of the metrics, and remember this is for a bone-stock 1700 shortly after release. Bottom of the barrel silicon (NOT what Apple would get).


Power Total Package 1T: 8700k uses 1.31x 1700
Power Total Package Full Load: 8700k uses 1.29x 1700
Power Cores Only 1T : 8700k uses 2.14x 1700
Power Cores Only Full Load: 8700k uses 1.66x 1700


So I guess I it wasn't "Euphoria" to say 1700 uses half the power after all. Unless you want argue Anandtech Bench is euphoric.

 

[FIVR]

AT WORST you get a 30% increase in efficiency going from 8700k to 1700. At best it is MORE THAN 2X AS EFFICIENT

It's too bad for intel that anandtech bench exists. I might have been convinced by their cherrypicked graphs if it didn't.

 

[Dayman1225]

Are you not reading what we are saying? Or are you just ignoring what you want so you can live in Ryzen land? The 8700K will use more power at PEAK. We are talking about efficiency of the CPU in actual workloads. Where you calculate the work done by the processor with its power usage and score at the end of a test, such as blender.

 

[dahorns]

Here are some of the metrics, and remember this is for a bone-stock 1700 shortly after release. Bottom of the barrel silicon (NOT what Apple would get).


Power Total Package 1T: 8700k uses 1.31x 1700
Power Total Package Full Load: 8700k uses 1.29x 1700
Power Cores Only 1T : 8700k uses 2.14x 1700
Power Cores Only Full Load: 8700k uses 1.66x 1700


So I guess I it wasn't "Euphoria" to say 1700 uses half the power after all. Unless you want argue Anandtech Bench is euphoric.

All youre measuring is peak power. That does not make your point.

 

[FIVR]

Are you not reading what we are saying? Or are you just ignoring what you want so you can live in Ryzen land? The 8700K will use more power at PEAK. We are talking about efficiency of the CPU in actual workloads. Where you calculate the work done by the processor with its power usage and score at the end of a test, such as blender.

Quote an andandtech bench that supports your argument. You claim 8700k is more efficient and use graphs from some fly-by-night GAMERX.com website but why can't you find reputable sources that say the 8700k is more efficient?

All I see is "muh workloads". Aandtech did the testing. Intel lost, get over it.

 

[dahorns]

All youre measuring is peak power. That does not make your point.

Follow up,

Two robots:

Robot A consumes 5w continously for 5 minutes while moving a 5kg rock 1 mile.

Robot B consumes 20w continously for 5 minutes while moving a 5kg rock 10 miles.

Which is more efficient?

 

[TahoeDust]

Quote an andandtech bench that supports your argument. You claim 8700k is more efficient and use graphs from some fly-by-night GAMERX.com website but why can't you find reputable sources that say the 8700k is more efficient?

All I see is "muh workloads". Aandtech did the testing. Intel lost, get over it.

Nothing you posted my Anandtech has anything to do with power efficiency. It only shows power consumption. You need to look at performance per watt to judge efficiency.

That is like judging a car's efficiency in gallons per hour and not miles per gallon.

How much work is being done per watt? I do not know which is better, but that should be the question.

 

[dahorns]

Th

Quote an andandtech bench that supports your argument. You claim 8700k is more efficient and use graphs from some fly-by-night GAMERX.com website but why can't you find reputable sources that say the 8700k is more efficient?

All I see is "muh workloads". Aandtech did the testing. Intel lost, get over it.

Anandtech did not do that testing. It tested peak and idle instanteous power consumption. It did not test efficiency, not that I have seen.

 

[PeterScott]

Quote an andandtech bench that supports your argument. You claim 8700k is more efficient and use graphs from some fly-by-night GAMERX.com website but why can't you find reputable sources that say the 8700k is more efficient?

All I see is "muh workloads". Aandtech did the testing. Intel lost, get over it.

Up until you moved the goal posts, we were talking about efficiency: perf/watt.

Now you you just want to ignore that, and rant about peak power usage.

Seriously. This is like putting your fingers in your ears and yelling "I can't here you", when someone points out some truth you don't like.

Gamersnexus is not some fly by night site. They do extensive investigations of GPUs/CPUs and they actually tested CPU power while doing work. So you can get an idea about efficiency which is what the discussion was about.

 

[Dayman1225]

Heres the full GN article if you would actually like to read it and come up with an opinion of your own.

FIVR, you're just digging a hole for your self. Everyone can see you're moving of goal posts. But hey, I am the one that cherry picks and is shilling for Intel!

All I see is "muh peak wattage in a test that never ends so we cant actually calculate efficiency".

 

[FIVR]

You guys are the ones changing the goal posts. You want to ignore power usage and concentrate on some nebulous metric called "Efficiency". There are so many different workloads with that would produce vastly different results that it allows you to post stuff like that gamer.com graph which, although I doubt it is fake, is a gross misrepresentation of the actual efficiency of the 1700 vs the 8700k.


The context of this thread is the Apple iMac. Apple does not care about 20% extra performance if it comes at 30% more power usage. Apple cares first and foremost about load power usage and peak power usage in the iMac platform. It is a TDP constrained device.


Pertaining to the argument at hand, the iMac, the 8700k is a vastly poorer choice and that is supported by anandtech bench and almost every single source online. Please constrain your arguments to the one at hand and do not change goal posts to "what is the most efficient processor based on this blender workload i picked" or "what is the most efficient processor based on the highest possible score for 8700k". Which is what you are actually trying to argue.

 

[Dayman1225]

There are so many different workloads with that would produce vastly different results that i

Yes there are but that also applies to this

Apple cares first and foremost about load power usage and peak power usage in the iMac platform.

DIFFERENT workloads will change PEAK and LOAD usage.

For example the 8700k as seen above is more efficient in blender, while the 1700 is likely to be more efficient in lets say AES Encryption work loads (Which I hear Ryzen is particularly good at).

Pertaining to the argument at hand, the iMac, the 8700k is a vastly poorer choice and that is supported by anandtech bench and almost every single source online.

And why is that?
1) Intel and Apple likely have an contract for X amount of years, like AMD has one for the GPUs

2)The Macs, IMacs were recently updated no? Also the IMac Pro is coming with upto 18 Xeon cores. So a switch to AMD anytime soon is UNLIKELY.

3)Like I said before, different workloads mean different PEAK AND LOAD consumption.

"what is the most efficient processor based on this blender workload i picked" or "what is the most efficient processor based on the highest possible score for 8700k". Which is what you are actually trying to argue.

No, we were trying to argue that Intels 8700k is more efficient in certain work loads, as Ryzen would be in others. Efficiency isn't a set thing it depends on the workload at hand - the same applied to IPC.

 

[Dayman1225]

Now back to the IMAC Ryzen stuff - Its fake.

Base Freq doesn't match

Cache amount doesn't match either.

https://browser.geekbench.com/v4/cpu/4568987

https://browser.geekbench.com/v4/compute/1279306

Compare them yourself

Seriously - why wasn't this the first thing you checked?


EDIT: My mistake the first result I picked for some reason had different Cache from normal chips - But the change in Base freq still stands