• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Ryzen "2" ?

Steveocee said:
Kind of meant just announcing the new lineup, you think one day in April AMD are gonna tweet “FYI #newcpus”?
Click to expand...

Oh i see, while they didn't go into specifics about branding or naming conventions "Ryzen 2600/X 2700/X" and all that... they have let the cat out of the bag that we will get new Ryzen CPU's in April, that much is official, April 19 :) Edit the exact day in April is rumor, don't know where it came from.

https://www.pcper.com/news/Processo...eration-Ryzen-CPUs-April-Threadripper-2H-2018
 
Why are we even talking about CanardPC "review". 2933 vs 2600 gives atleast 2% performance over multiple games on 1080p + 2700X vs 1800X would have higher clock even with A320 motherboard. So you are seriously saying with those advantages 2700X is only 3,4% faster than 1800X. That would in absolute sertain mean Ryzen 2 has lower IPC than Ryzen 1.
 
B12B6 said:
Why are we even talking about CanardPC "review". 2933 vs 2600 gives atleast 2% performance over multiple games on 1080p + 2700X vs 1800X would have higher clock even with A320 motherboard. So you are seriously saying with those advantages 2700X is only 3,4% faster than 1800X. That would in absolute sertain mean Ryzen 2 has lower IPC than Ryzen 1.
Click to expand...

Yeah i have also made this point. :)

In their review: Ryzen 1800X is running 2666Mhz RAM, 2700X 2933Mhz, the gaming performance difference is 3.4%, that accounts for the difference in Ram speed, its bang on if they are running the same 3.7Ghz CPU clocks.
The difference between 3.7Ghz and 4.35Ghz is 17%, they say the gaming performance on B/X400 series boards is <8%... again now THAT accounts for the clock speed difference.
Click to expand...

And expanded on it here..... because... well. best not quote what it was in response to.

humbug said:
Then we need to assume CanardPC got +3,4% gaming performance from the 2666Mhz vs 2933Mhz ram speed and the <4.35Ghz all core boosting (+17%) clocks... no, but for the sake of argument let us assume that.

So, when they say they get <8% extra gaming performance from using the 400 series motherboards, and they claim those motherboard are MCE stile overclocking what would those clocks be to achieve +8% gaming performance? well 4.35Ghz +8% is 4.7Ghz, but wait..... they only got 3.4% from a 17% higher clock, so what must the clocks actually be in order to achieve twice that? you would double the 17% for scaling, so 4.35Ghz + 2x 17% = 32% = 5.75Ghz....

Please think about this :)
Click to expand...

I'm off for some lunch :)
 
Steveocee said:
Just me or has AMD missed the chance to whip up some serious hype in advance of the new chips?
Click to expand...

I don't think the changes in Ryzen 2 are worth a load of hype, it's a slight incremental improvement mainly due to a new process that should ultimately lower production costs.

Yeah its going to be better, but its not like the jump between FX and Ryzen 1, theres no revolution this time, just refinement, this is no bad thing - we need AMD to have a good release candence across mainstream consumer, HEDT, embedded and Datacentre lines in order to keep the pressure on Intel - but more importantly for them to grow and look like a reliable and strong partner to the major industry players and hyperscalers.

Zen2 should have more architectural changes - I could see them hyping that more.
 
Last edited:
B12B6 said:
Why are we even talking about CanardPC "review". 2933 vs 2600 gives atleast 2% performance over multiple games on 1080p + 2700X vs 1800X would have higher clock even with A320 motherboard. So you are seriously saying with those advantages 2700X is only 3,4% faster than 1800X. That would in absolute sertain mean Ryzen 2 has lower IPC than Ryzen 1.
Click to expand...

I suggest you use the cpu benchnarks to determine IPC for obvious reasons. Using games which don't scale with CPU power is stupid.

Using gaming as the IPC benchmark I can otherwise say according to benchmarks the 1700X has better IPC than the 1800X, which is obviously not true.

If you insist on using games, it must be clock for clock.

www.techpowerup.com/img/vgr2mXf1M9QKkLi0.jpg

2700X is 8.7% better on real CPU benchmarks. Exactly what you'd predict given what we know about clocks. It also explains why AMD didn't want to call this the 2800X.
 
humbug said:
Thats not what it says on the chart

The chart says the following:

8 Cores 3.9
7 Cores 3.95
6 cores 4.0
5 Cores 4.05
4 Cores 4.05
3 Cores 4.1
2 Cores 4.175
1 Core 4.3
--------------------

With other leaks citing 4.35Ghz i believe those findings to be accurate.

iioaync.jpg.png

This is completely insane, everyone can read, what are you doing?
Click to expand...

So just to be clear. Given what you've just posted you don't think the all core boost boost is 3.9ghz but...
 
ltron said:
The difference between MCE and what AMD does is quite simple. AMD guarantees these speeds to work on all CPUs released with no issues while MCE is not guaranteed at all by Intel, it's simply an auto overclock done by the motherboard manufacturers.

While MCE will probably work in most cases, due to it being fairly conservative, it will not in all as evidenced by Gamers Nexus' sample falling over in certain Blender workloads. Also, if you buy and use a basic cooler expecting it to run at stock, MCE could give you a nasty surprise as your CPU will run very hot.
Click to expand...

In addition you void the warranty of your CPU, someone could buy the chip, pop it in, load optimsied defaults, assume they running the chip to intel spec and without realising they are actually running outside of spec. Also with faster degradation as well.

This is bad, but even worse is some reviewers didnt even realise what was going on, its as the oc3d guy said, if you cannot notice abnormal clock speeds and voltages whilst reviewing hardware then you shouldnt be in the reviewing business.
 
muon said:
2700X is 8.7% better on real CPU benchmarks. Exactly what you'd predict given what we know about clocks. It also explains why AMD didn't want to call this the 2800X.
Click to expand...

I'd say it's pretty clear they're harvesting the top bin for threadripper/2800(x) once this 8 core Intel turns up. It's (kinda) fair enough. They'll have a better idea where Intel went (thermals/power/mce/etc) once the 8 core turns up and they can delivery their best chip in a similar envelope so best match(/better) it.
 
Mercutio said:
I'd say it's pretty clear they're harvesting the top bin for threadripper/2800(x) once this 8 core Intel turns up. It's (kinda) fair enough. They'll have a better idea where Intel went (thermals/power/mce/etc) once the 8 core turns up and they can delivery their best chip in a similar envelope so best match(/better) it.
Click to expand...

When are intel supposedly turning up?
 
LOLN1 said:
Hi all

here amd ryzen 2700x live test
https://youtu.be/FUus2sCUE5k?t=2416
this afternoon, other live tests

Happy Easter
Click to expand...

Nice. Boosting to 4.35 on x370 and memory and L2 latency improved.
174 cinebench single core is going in the right direction too, 1786 multi core!

If this pans out that's now enough of an upgrade in single core from my Haswell to warrant an upgrade.
 
Pants said:
Nice. Boosting to 4.35 on x370 and memory and L2 latency improved.
174 cinebench single core is going in the right direction too, 1786 multi core!

If this pans out that's now enough of an upgrade in single core from my Haswell to warrant an upgrade.
Click to expand...


DragonQ said:
For comparison, stock i7-6700K (Skylake @ 4.2 GHz) scores 183 in single-threaded Cinebench.
Click to expand...

Yeah its the same 8% single threaded IPC difference between Ryzen and Coffeelake, so no IPC change on single threaded, i think we can assume the Multithreaded comparison is also the same, 99% of Coffeelake.

@Pants the single core performance is about the same higher to Haswell as what Coffeelake is to Ryzen, about 8%, its not that much higher but the multicore performance is massively higher :)

I'm late into this, i'll watch it from the beginning later.
 
You must log in or register to reply here.
Back
Top Bottom