Intel’s surprise Ryzen killer

[smilingcrow]

My 3600 barely uses 40 Watts and that's in a high load game like Star Citizen, in Left4Dead its about 10 Watts because the CPU isn't doing anything.
I think this is probably accurate, this is a workload where the Uncore probably uses more power than the cores.

Wow, I didn't realise how lightly CPUs are being used in gaming.
I read so many people saying, "You gotta have 8 cores because the consoles are 8 cores", yet based on what you are saying and the two reviews showed, the CPU isn't being stressed that much.
It's gonna vary by game, resolution and the GPU used so I'm not going to extrapolate too much from all this.

I understand the argument here is about gaming power consumption, but 10 watts vs 15 watts is not the whole story and if it were to be used like that its incredibly misleading.
In reality where the CPU is loaded up properly Intel's efficiency is very poor, using well over 200 Watt, compared with about 150 watts of the faster 12 core 3900X, and the performance per watt comparison is about to get even worse for Intel.

The data clearly speaks for itself and the linked pages show the consumption under full load also so nothing misleading about any of that!

"Loaded up properly" is a strange term to use!
Games, DAW use and other types of loads are variable and don't load all cores continuously. That is appropriate and proper for those types of tasks.

I might dig out my power meter and see how many watts the system uses when my DAW is running.
Possibly much less than I imagined.
That's great news as it means I can use a higher TDP CPU and still maintain a zero noise desktop.

 

[Shaolin Chicken]

I might dig out my power meter and see how many watts the system uses when my DAW is running.
Possibly much less than I imagined.
That's great news as it means I can use a higher TDP CPU and still maintain a zero noise desktop.

Will be interested to see the results please

 

[chrcoluk]

AMD needed 10 quarters in order to increase its servers market share from 0.8% to 5.8%.
With this growth speed, AMD will need 88 quarters or 22 years only to reach 50-50 share parity with Intel..

https://www.tomshardware.com/news/amd-vs-intel-highest-overall-x86-chip-market-share

Datacenter's are not like twitch happy consumers who might replace their hardware every gen, they would be in serious financial trouble if they did that, they expect purchases to last upwards of 5 years. So when they buy new hardware, it is a gradual process.

I am seeing serious shifts from xeon to epyc, but it will take a decade or so of that to have a full bearing on figures.

 

[4K8KW10]

Datacenter's are not like twitch happy consumers who might replace their hardware every gen, they would be in serious financial trouble if they did that, they expect purchases to last upwards of 5 years. So when they buy new hardware, it is a gradual process.

I am seeing serious shifts from xeon to epyc, but it will take a decade or so of that to have a full bearing on figures.

Yes, they are looking for return on investment which means that AMD must give performance for low price in order to offer something of value for these customers.

And to break those relationships between Intel and everyone..

 

[jigger]

AMD needed 10 quarters in order to increase its servers market share from 0.8% to 5.8%.
With this growth speed, AMD will need 88 quarters or 22 years only to reach 50-50 share parity with Intel..

https://www.tomshardware.com/news/amd-vs-intel-highest-overall-x86-chip-market-share

You don’t seem to understand the server D.C. market very well. It take time to see movement, but AMD will grow rapidly over the coming years.

 

[4K8KW10]

You don’t seem to understand the server D.C. market very well. It take time to see movement, but AMD will grow rapidly over the coming years.

AMD "will" grow ONLY if there are conditions. It's never nice to speak untruths or false predictions with the imaginary help of a crystal ball.

 

[jigger]

AMD "will" grow ONLY if there are conditions. It's never nice to speak untruths or false predictions with the imaginary help of a crystal ball.

So maybe you should stop posting for a while?

 

[Pleiades]

So maybe you should stop posting for a while?

Hehe, brilliant!

 

[Roadrunner_668]

i like how the fanboyism hasnt changed even since core 2 vs phenom 64

 

[Grim5]

You don’t seem to understand the server D.C. market very well. It take time to see movement, but AMD will grow rapidly over the coming years.

market share in that area is a snowball, takes time to get moving but once it does its an avalanche

 

[4K8KW10]

market share in that area is a snowball, takes time to get moving but once it does its an avalanche

The historic data doesn't support your statement. AMD has 5% more market share than in 2017 and there are no signs that it will accelerate.

 

[jigger]

market share in that area is a snowball, takes time to get moving but once it does its an avalanche

Yeah generally. The market is a little more offset at the hyperscale end because of the massive scale.

 

[jigger]

The historic data doesn't support your statement. AMD has 5% more market share than in 2017 and there are no signs that it will accelerate.

Took Intel over a decade to recover. Proves the point exactly.

 

[Mr Oblong]

Took Intel over a decade to recover. Proves the point exactly.

Fair point.

 

[4K8KW10]

Took Intel over a decade to recover

Intel doesn't recover. It's actually and in reality AMD's tiny market share ever eroding.
The only thing on that historic data visible is one relatively short hill, which was in its peak 3 years after the corresponding product launch.
Now, we are well over 3 years since EPYC's launch and there is nothing even remotely comparable.

 

[jigger]

Intel doesn't recover. It's actually and in reality AMD's tiny market share ever eroding.
The only thing on that historic data visible is one relatively short hill, which was in its peak 3 years after the corresponding product launch.
Now, we are well over 3 years since EPYC's launch and there is nothing even remotely comparable.

Back then AMD could barley fill 20% of the demand.

The same thing. Intel gaining sales, AMD losing sales.

 

[d_brennen]

Intel doesn't recover. It's actually and in reality AMD's tiny market share ever eroding.
The only thing on that historic data visible is one relatively short hill, which was in its peak 3 years after the corresponding product launch.
Now, we are well over 3 years since EPYC's launch and there is nothing even remotely comparable.

Every sale that isn't an Intel sale is more than likely an AMD sale. Yes there's ARM and RISCV, but that's just nibbling away at Intel, AMD and to a lesser extent IBM. In what way is this a bad thing? Unless you have Intel shares, of course

 

[Roadrunner_668]

be interesting to see how the desktop cpu market share has changed over the last 15 years.

 

[FoxEye]

Took Intel over a decade to recover. Proves the point exactly.

We looking at the same graph?

The other way to interpret that data is that AMD lost almost 1/2 of their market share from Jun 06 to Jun 07, and the rest of the chart is AMD slowly bleeding out, from almost insignificant to almost 0% market share.

I'm not seeing anything positive for AMD in that graph.

 

[jigger]

60 FPS on a 60hz monitor isn’t a terrible place to be, todays monitor quality is another topic. Most people today are looking to run games at double or even quadruple those rates and firms like Intel and Nvidia plus the likes Asus and EVGA along with anus tech tips have been pushing that narrative for a decade.

People want da best and the PC gaming market is built around that.