- Mar 3, 2017
- 1,747
- 6,598
- 136
Both. Zen 4 X3D is thermally limited to 89C no matter the voltage, which tells you too much power is also a no go.isnt X3D issue voltage and not power ?
Precisely-- now do you think that bodes well for X3D boost clocks?
Was there talk in the past about hoping they can separate the voltage that goes to the V-cache from the regular voltage going to the rest of the core to help with this.As pointed above the problem was the V-cache limited voltage.
Seems that Zen 5 require 1.3V at 5.3GHz, wich is the safe limit of the cache, so in principle the X3D variant could be clocked a little higher, and along with 10-12% better INT IPC this could yield 15-16% better perf than the 7800X3D although at this point it s impossible to do a sure estimation.
Likely that AMD wont bother implementing separate voltages rails, this would require a regulator on die and the market is just too small to be worth the effort and added cost, after all at the basis that s a server die with V-cache and those SKUs work at low frequencies and voltages anyway.Was there talk in the past about hoping they can separate the voltage that goes to the V-cache from the regular voltage going to the rest of the core to help with this.
More like from Zen 1 (Zen 4 in this case) to 1+ (aka 2000 series, Zen 5 here) at first glance, but I see TPU reporting on an updated agesa already with lifted power limits...Zen 5 reminds me of Zen 1 except it doesn’t bring a massive uplift vs past generation. But It seems to have so much potential for future iterations.
Price, OEM and office Pc's.If they somehow manage to hit clock parity with X variant, I genuinely think there’s no point for the existence of lower end X variants. People who are into productivity tasks naturally gravitate toward R9s anyway.
and why would you buy a 16 core 32 thread CPU if not for very heavy use ? I know there are windows uses, but more uses that are like server uses.I do not agree with him at all, linux test are more like sever grade test, linux is not used as windows as home user do use them, with xx and xxy service and appz running at same time...
I have one 13700k desktop use
and 12 ryzen linux cli more server kind of load, he do the same its not fair to compare it with desktop windows use at all
I'd personally add users who can't afford X3D chips as well & like to OC or game at 4K+ high refresh rates where the diff in fps is 1-2% avg compared to non X3D.Price, OEM and office Pc's.
Nah. I think there are thousands, if not millions, of Ryzen Linux servers out there. So it's a sizeable percentage of users.but I am in that 0.0x % group of the use case
lol and some people had the gall to question my assertion that this was a dumpster fire of a launch when it got pushed back a couple weeks ago. Now this? Only a hardcore fanbot would dare claim otherwise at this point. The 9700X, at least, should have been 105W TDP from the get go.
Have we found out exactly why AMD delayed the launch? Related to the Windows/Linux disparity?
This launch shares many parallels with Bulldozer, although the absolute performance and power characteristics of the product are nothing like Bulldozer.
Anyways, both got:
* Horrible hype and projections
* Slow and painful hype train derailing process
* Incompetent marketing fluff
* Unorthodox x86 architecture solutions
* Last minute software "adjustments"
* Linux performance gains more impactful than the Windows ones
* etc.
But probably the worst part having striking resemblance to Bulldozer are the expectations for future products - "fixing the shortcomings" and "unlocking the true potential". This sounds exactly like all those people talking about the great foundation for the future. It remains to be seen.
Assuming a fix is coming in the future, now's a good time to start categorizing benchmarks where the inter-core latency is hurting 9950X. Previously that would have required some serious profiling to identify applications that depend extensively on inter-thread communication for parallelization but now just a mere glance at the difference between 7950X and 9950X benchmark scores will tell us whether an application is doing heavy inter-thread communication.Cuz whatever its going with the intercore latency on 2CCD chips and/or Windows core scheduling
It doesn't make any sense. That's like taking one step further and then going two steps back. There must be a solution. Now it's all making sense why AMD held Zen 5 back for a year. Maybe they were trying to fix whatever the issue is.I doubt any fixes are coming. Most likely is that's just how Zen 5 works.
1.1vcore set for static 5.3ghz allcore in R23 MT on my Z5 vanilla ES sample for ~46k pointsAs pointed above the problem was the V-cache limited voltage.
Seems that Zen 5 require 1.3V at 5.3GHz, wich is the safe limit of the cache, so in principle the X3D variant could be clocked a little higher, and along with 10-12% better INT IPC this could yield 15-16% better perf than the 7800X3D although at this point it s impossible to do a sure estimation.
Next time for example don't believe everything @adroc_thurston(Bondrewd) said or pointed to.God, I HATE leakers!
There wasn't any, that place wasn't worth much for eons.Because his track record at beyond3d was not good.
Unrelated places with different naming conventions.Why do you think he used a different nick here?
To be fair, I don't think he was the one who said anything about ladder cache. It was probably Red Gaming Tech and their stupid ladder cache diagram.Next time for example don't believe everything @adroc_thurston(Bondrewd) said or pointed to.
Skymont being as strong as it is supposed to be and if Intel learned any lesson from MTL, I think their inter-core latency isn't going to be as bad as Zen 5. AMD basically invited Intel to eat their lunch in parallel workloads with heavy inter-thread traffic. Maybe there's an Intel double agent engineer at AMD who is sabotaging their designs.I would say ARL would be the clear winner, but considering all the talk about no HT I am not sure.