I wouldn't call CPB as default behavior an overclock, since this is a bios setting offered and promoted by AMD, and this explains why it is initially enabled. The cpu still adheres to its base-boost frequency range.
The point a few of us were trying to make is that the 3700x can be set to stay...
But the data available shows the 2700 does not behave like a 2700x or 3700x when it comes to power usage with default settings.
2700x did go well over its 95w, that was pretty well known to use its 142w cap unless it hit other limits. The review I linked shows that in the wall power...
It'd be nice if I had a 2700 to do a back to back comparison. However, I only have access to my 3600 and possibly a 3300x machine with similar specs.
For the 3 b450 boards I have used this year, (asus b450-i Strix, B450M-tuf, asrock b450m pro4) CPB was enabled in cpu sub menus. I believe this...
False, but why bother looking it up when you can just state opinion.
Lowest power draw from the wall was the 2700 in blender, at 126w:
https://techreport.com/review/33719/amds-ryzen-7-2700-cpu-reviewed/
I understand the setup is different, however mine is pretty well optimized for high eff at...
I put my watt meter back on my 3600 system, and the only way it stays close to the 65w rated tdp when doing avx2 style workloads is CPB off, base frequency (3.6ghz).
Idle 40w wall, CPB off, ram at 3200, y-cruncher pulled 119w peak. PSU is 450w 80+ gold.
All stock bios (well disabled the rgb eye...
Have you given the board a visual inspection for any evidence of a component gone bad? Vrm area is likely hidden by heatsinks, though the back of the board might show something here if it went really bad.
I'm guessing if it smelled like the angry pixies escaped somewhere this wouldn't be much of...
I shouldn't need to jump in to say the mods here are being a bit defensive when the user just got here. Theres a reason this is a discussion and not a bunch of stickies where the mods tell us what the situation is. Maybe I'm one of the few here who have actually had to figure out how to get a...
What do you think laptop renior has? A real psu? Though maybe if one tries to do a static oc like you did to your 3600 you'd be past a picopsu wattage.
It seems likes some of you guys responding need to go outside and get some fresh air. This poster is definitely just looking for correct info...
I'm not sure where the early guys responding got a defensive nature from, I dont see it.
Whats your use case where a 2700 and a 3700x are somewhat interchangeable? For a lot of compute workloads, the 3600 will beat a 2700(x) due to the higher ipc, avx2, and more efficient node. Though it is...
I would check to see what bios version you need for your mb to support a 3000 series, but most b450 boards should be able to.
However, the 1600 is prob fine, might see more benefits from a gpu upgrade if they mostly game.
I believe the gpu core is based around openMM, and the code is optimized far better for nvidia in general, even when cuda is not used. Cuda has been the dominant api in that field, so it makes sense that if majority of your cards are already nvidia, thats what you optimize for. Its also a self...
Folding at home with gpu is OpenCL only right now, using both SP and DP precision as needed (a22 core). Cpu core is avx-256. Most other projects are still cpu only, though occasionally a gpu project will run for a while. Nearly all cpu projects that I know of utilize avx2 when available.
I'm...
I was merely pointing out to other users who seemed to think the 10900K had avx512, and that there might be a benefit to that over a zen2 cpu.
I think most DC type applications would sooner move to opencl type work, as the gpu power available would be far more than a HEDT/server only...
I've seen it mentioned a few times here, but 10900K doesnt support anything past avx2. Its still rehashed skylake round 4.
You'd need the 10900X hedt for versions of avx-512.
My base overclocked 3850 easily got higher cpu scores than the 5800k and 6790k I played with after. (3870k didn't arrive till later).
Way overvolted from factory. 1.4ish, it ran 100% stable with a -0.2v offset. It was a lot of fun to learn how to OC on. Definitely fried the NB vrms on that itx...
It opens up a lot more doors than just pcie connected devices, if you consider that IF connected gpu has been in the works as well. At least, to me it looks like they went for HSA at the exascale level, instead of consumer.
I'm really hoping that next gen APU is actually CDNA2 based, allowing...
They may have stopped making as big a deal about it from the consumer side, but they are still working towards it.
CDNA2 is supposed to get cache-coherent memory access between cpu + gpu:
https://www.techpowerup.com/264536/amd-announces-the-cdna-and-cdna2-compute-gpu-architectures
Start slow, figure out what helps and what doesn't (both in terms of increased freq and voltage), especially with a program that provides a score. cinebench for short runs, blender/handbrake (without quicksync or whatever its called) for longer ones.
Theres more to stability than just passing x...
Didn't expect it to only see 1/2 your threads on linux, though after some reading it seems the x265 encoder is limited to 64 threads unless told otherwise:
https://x265.readthedocs.io/en/default/cli.html (check out the pools info section)
Also, from the log file :x265 [info]: Coding QT: max CU...
If you're comfortable with CLI, you might be able to get away without a gui:
https://handbrake.fr/docs/en/latest/cli/cli-options.html
Now, I am not great at it so cant say it will work 100%, but it does seem like its possible.
As for using a GUI linux, I'm on Linux Lite, which is based on...
I do believe this a windows problem and not a bios setting based on the 3990x review here on anandtech.
Handbrake is available on linux if you want to see how well that 3990 scales.
What? That makes no sense. Roll out a new platform with significant upgrades to connectivity, but then make extra effort to make a few processors backwards compatible for upgrades? Why not just roll it all out on x399 in the first place then?
I recall reading that majority of TR sales were the...
The avx2 of the 3600x might make it faster for some editing post processes, even though its down cores compared to the 2700x. Its why I'm waiting for the zen2 apus.
Since you say its probably similar performance to a 3970x, how does it compare in perf/watt?
I believe in another thread you said the MB allows for 7002 series cpus, what cpu models are you going to keep an eye out for in the future to upgrade?
All IHS factory installs I have seen, do not seal to the pcb 100%. They always leave a small gap to prevent any pressure differentials during heat cycles. Therefor there is no filling the cavity with any inert gases.
If there isnt a specific reason to use the x2, and dont want to spend the...
But that $750 high end consumer is 4x what you can get from amd on the low/mid end, and the midrange for intel - so its price is justified. Its not like a high clocked quad for 4x the price (lga 775 extreme edition)
Edit: If you dont need it, sure the 3950x seems pricey. But for those who can...
I've seen a few AMD mounting brackets that use the center strap method, but allow some rotational leeway - up until the heatpipes hit the mounting strap. Plenty of ways to get worse cooling than having the tower cooler at an odd angle.
Maybe some of you guys should try and start a class action about being ripped off by class action lawyers.
Clearly this BD 8 core class action was a lawyer money grab, to the point where AMD just decided it was cheaper to settle than to keep paying the scumbags - I mean lawyers.
There are pro cpus with and without integrated graphics - any based on the 8c die wouldn't have any.
Models are split between having no gpu and those with gpus.
Edit: Ha, seems maddie beat me to it with the same source.
You initially stated the 7252 was a 16c and I assumed that was correct. I'm not sure what seems to be confusing you but the core counts and cache make sense at each level to me. (I had it all laid out and the post got eaten)
There was a leaked list a while back down to 2 chiplets, so I'm...
32MiB per chiplet, so the 7252 is most likely a 2 chiplet 16c just like the 3950x;
the 7282 most likely a 4 chiplet 16c.
Now, the 7282 can either be 2 cores per ccx, or 1 ccx disabled per chiplet. I like the latter based on latency tests of zen2 core - core from same ccx vs to other ccxs.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.