Stock voltage is very high, Raptor Lake's P-cores doesn't need 1.2v to hit 5 GHz, and can do 4 GHz at 0.95v or less. It'll be interesting to see what undervolting does, SkatterBencher's 24-core 2495X uses 1.15v at 4.4 GHz:
Asus does use the V/F curves to generate a rating for the CPU but it's only available with their MBs. I don't think they've publicly released their methodology publicly or licensed it to other board manufacturers, though Gigabyte has recently released their own version...
Are you using a negative offset voltage to undervolt? I find that these kind of reboots are related to idle or low CPU usage voltage being too low.
In my new 13900K system, I had one idle reboot when I was using the a global negative voltage offset. I switched to using the V/F curve to only...
Igor's Lab has an article with an analysis of several hundred 13900K/KFs using the SP values read out with an Asus motherboard. Though after reading the article, it doesn't seem like their is a strong finding for one model over the other...
Raptor Lake uses less power than Alder Lake in games for better performance. The 13700K vs the 12900KS performance/watt improvement is so high you might think it was actually using a new 7 nm process instead of 10 nm.
A Chinese reviewer used a Thermalright AS 120, and undervolted their 13900K to 5.2/4.3 at 1.23V. Power use went down to 190W in Cinebench R23 but still almost hit 40K. Temperatures were only mid 70s to low 80s running the AIDA Stress FPU test.
AMD and Apple have more than 8 P-processors using a MCM approach, which Intel has so far not wanted or unable to do so for the consumer market. To use the old AMD term, neither has a CCX with more than 8 P-processors.
The used benchmark, Cinebench R23 scales almost perfectly with clock speeds. The low difference between 300W and 200W is probably from the 300W runs hitting the overheat limit and frequently throttling the P-cores while at 200W, the P-cores can run continuously at 4.8-4.9 GHz.
The 9700K didn't have HT because it was the #2 CPU in Intel's lineup and it having 8 non-HT cores was enough to trade blows with the 2700x. It didn't need to beat the 2700x in everything, that was the job for the 9900K and its higher price tag.
A nearly 29K score in R23 for the 5950X requires an overclock to 4.5 GHz all core. I'd be surprised if AMD is claiming the 40+% over that, more likely it's against a stock 5950X that scores in the 25.5K range.
Hyperthreaded Sandy Bridge and Ivy Bridge still do better in games today (and even compete pretty well against Zen 1). Their superior per-core performance meant that even two Piledriver "cores" could only match a HTed Bridge core with two threads.
In my experience with patents, the only thing that matters are the claims, which in this case seems to be only related to cache coherency and provides almost no details on the rest of the processor's architecture. Every thing else in a patent is really just fluff.
I guess one way for Intel to...
I don't think the 12900KS is meant for the 5800X3D, more for Intel to take some of the remaining benchmarks that the 12900K currently loses to the 5950X. If they worried about the 5800X3D, a 12600KS would be the better counter (given how well the 12600K currently competes with the 5800X), or...
What's the core voltage when it's running R23? If it's 1.155v that seems very high for a P-core clock of only 3.2 GHz. On a 11800H laptop that I have, the voltage at 3.2GHz all-core is only 0.9v at default settings.
As Ethereum is mostly memory bandwidth limited and requires a fair amount of memory, it isn't as easy to build an ASIC that can beat a GPU in Ethereum as easy as they do in Bitcoin. Especially as new GPU flagships often adopt the latest memory tech first.
And it could be ASIC manufacturers also...
A long time ago, I recall reading that it mainly equalizes the priority of background services so that they have a better chance of being scheduled next to run and increasing the time that a thread can run before being thread switched away and therefore increasing throughput at the cost of...
Assuming that TSMC remains capacity limited for its most cutting edge node, it could also be a competitive move. One 3nm TSMC wafer for Intel is one less wafer for AMD/Apple/Nvidia/Qualcomm/etc.
To me it sounds like you have a lot apps open, but not really loading the CPU cores simultaneously. In most cases, a moderate number of extremely powerful cores will probably benefit you more than a greater number of less powerful cores.
Yeah, the only thing I can think of is the ring bus being used still has a limit of 10 cores (or core clusters) that won't be improved until Raptor Lake. Otherwise, a 10P+8E would have effectively Core 2ed Zen3, sweeping virtually all of the benchmarks with reduced power consumption.
It depends on the particular CPU though in general for the 5000 series, the 5900x and lower appear to have quite a bit of headroom while the 5950x is already optimized by AMD (probably to make it shine even more). Here's a video about using it to overclock:
I could only get -9 on my 5950x...
It can if you also use the curve optimizer to set a negative voltage offset, which would allow for a higher clock at load but potentially result in too low of a voltage at idle or low load situations.
I think the next few weeks and months will be interesting in terms of under-volting as some users and reviewers try to optimize Alder Lake for performance/watt and see how that compares with the best under-volts for the 5900X and 5950X.
I have a 5950x too, I get a score of about 28.6K using PBO2 overclocking (by increase the package power limit to 250W) and a -9 offset in the curve optimizer which results in all-core clock speeds of 4.45GHz when running CB23:
28K score requires an overclock on the 5950X (and all-core clock speeds around 4.4 GHz). At stock, its score is in the 25K range
https://www.anandtech.com/bench/CPU-2020/2976
I think so, but I don't think it's too unfair. The 5950x is already pretty well-tuned by AMD, you can't use a very high negative offset (compared to the lower tier 5000 models) with PBO curve optimizer before a 5950x becomes unstable.
Also impressive is the 25244 score with package power of 117W, which is better than the 5950x.
Given the high performance/watt of Alder Lake at reasonable clock speeds, we may see AMD "recommend" 200+W power settings in reviews for its 16-core CPUs going forward.
The point is that it didn't matter that Sweclockers used only 2.8 GHz to compare the cores in Cinebench R20 because GC had the same percentage higher score when Tweaker.nl tested it at 3.5 GHz or TweakTown did at 4 GHz.
Yeah, the recent Cinebenchs are pretty unaffected by memory speed and cache sizes. It's probably why a few sites picked them for IPC testing for this benchmarking round.
https://tweakers.net/reviews/9472/9/intel-12th-gen-alder-lake-core-i9-12900k-i7-12700k-en-i5-12600k-ipc-test.html...
And in CB R20, at 2.8 GHz, GC is about 16-17% faster then Zen 3. 22% faster than Cypress Cove and 44% faster than Skylake.
https://www.sweclockers.com/test/33062-intel-core-i9-12900k-och-core-i5-12600k-alder-lake/11
Interestingly, Hardwaredeluxx testing suggest that AVX-512 can be activated by disabling the E-cores:
https://www.hardwareluxx.de/index.php/artikel/hardware/prozessoren/57430-core-i9-12900k-und-core-i5-12600k-hybrid-desktop-cpus-alder-lake-im-test.html?start=16
I think in hindsight given that Intel finally mastered 10nm with Tiger Lake, they should have used all that resource into getting the the 8-core Tiger Lake released earlier and on the desktop. But it's understandable why Intel would be gun shy with a 10nm product given the trouble they had for...
It's a rough approximation but it's something for the curious and enthusiasts to do while waiting for the review embargo to lift.
Based on the behavior on Tiger Lake-H that I see. On an Inspiron 16 Plus laptop, at 8x3.4GHz (undervolted) results in 56W maximum package power in AIDA64 FPU, 50W in...
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.