igor_kavinski
Lifer
- Jul 27, 2020
- 16,817
- 10,765
- 106
Oh no! I'm Neo and I don't know it yet! How will I sleep at night now???We both are, you just haven't accepted it yet.
Oh no! I'm Neo and I don't know it yet! How will I sleep at night now???We both are, you just haven't accepted it yet.
But this does not address the issue of already degraded CPUs and the future of 13/14th gen CPUs.
Hard to see how efficiency cannot improve with this, but a lot of K - and even more so, KS - buyers mostly claim they don't care about efficiency - not even from being able to cool their CPUs.HUB benchmarks the ASUS version of the baseline profile (equivalent to "Extreme"):
Verdict: -4% in gaming versus 7800X3D while using an average of +119W system power.
They previewed a bit of the Gigabyte profile ("default" baseline), and performance takes a big hit though I'm sure the efficiency gets a bit better.
HUB benchmarks the ASUS version of the baseline profile (equivalent to "Extreme"):
Verdict: -4% in gaming versus 7800X3D while using an average of +119W system power.
They previewed a bit of the Gigabyte profile ("default" baseline), and performance takes a big hit though I'm sure the efficiency gets a bit better.
They better not deny RMAs. Looks like Bench Life is reporting even the baseline does not completely solve the problem
LOLZ at 3DCenter
A recap from HUB of the profiles we're expected to get by the end of this month. The default profile Intel is supposedly going to enforce is the Baseline with PL2 = 188W, so the one that Gigglebyte introduced in their boards already.
View attachment 98520
That's a great point. As many reports of crashing were due to Unreal Engine games Oodle Data decompression performing extra integrity checks, you would hope that is what they were using. Regardless of the workload though, if the CPUs are still crashing that's meaningful in itself. Probably means Igor was right about the CPUs being broken.It would be much more meaningful if we knew how they were testing for stability.
Crazy he was getting PSU coil whine and the PSU fan was louder than the rest of the system using the Intel part. He also took 3 games out of the test suite that the 3D is faster in. Yet we already have some implying he is shilling for AMD. 🤣 Not using MS Flight Sim over account login issues, was when I knew he was being a try hard at not getting labeled AMD shill. Yet he gets hit with it anyways. All it would have taken is paying another $10 for game pass with a throw away email Xbox account.Ok, so I run the Intel system in the fall/winter and AMD one in the spring/summer.
Edit: A good comparison of power usage by HUB:
View attachment 98526
That's a great point. As many reports of crashing were due to Unreal Engine games Oodle Data decompression performing extra integrity checks, you would hope that is what they were using. Regardless of the workload though, if the CPUs are still crashing that's meaningful in itself. Probably means Igor was right about the CPUs being broken.
I am no lawyer, but I have seen how the E.U. reacts to anti-consumer behavior from Intel. Chipzilla kicks the can down the road like that, and it could be much more expensive than the RMAs and the free double whopper coupon Americans will get.If even the new Intel default/baseline spec truly doesn't get them to stability, I don't know what Intel can do except move on to ARL as soon as possible and get ready to pay the class action lawsuit settlement. Other option would be to recall all effected CPUs (seems to be i7 and above 13th and 14th gen CPUs but maybe even on the mobile side?) but it's probably way cheaper and less of a PR hit if they just settle the class action down the road.
I am no lawyer, but I have seen how the E.U. reacts to anti-consumer behavior from Intel. Chipzilla kicks the can down the road like that, and it could be much more expensive than the RMAs and the free double whopper coupon Americans will get.
That's to go along with continued brand damage and likely even fewer sales of the highest ASP parts. Everyone pretends DIY/retail doesn't matter, yet everything these companies do says different/speaks volumes to its importance. LTT, GN, Jayz, HUB, will all rake them over the coals if they don't handle this right. Aussie Steve in that update at the end of the vid called it a Cluster. That's exactly where we are now. Intel has to fix this with its users or there's is no telling how bad it could get. Do the right thing, salvage this, that's what Intel's best course is.
And the warranty? How's the replacement doing?I'm shocked that there aren't more problems with them, my 12700F was wonderful for the year and a half that it lived... First the warping happened, then I lost PCIe lanes to the CPU (leaving my GPU at x4), and then one day early this year it just up and died on me. Not a single sign of life to be had, all at base locked power settings (125-150W). Though funnily enough, it seems that the 12th gen i7s are some of the least effected... Makes me wonder if it has to do with how inefficient those "E" cores are when juiced to the gills.
Recall isn't necessary. Just have a streamlined no hassles RMA process like they always have. Intel has been a top tier vendor to deal with over RMAs going back as far as I can remember. Just don't goofed that up the way they have everything else of late.I don't know about the EU, but if the U.S. market drives their reaction, I don't think having to settle a class action years down the road will have much effect on them. Maybe the E.U. would step in with much harsher penalties if they tried to get away with that, but it would have to be a really harsh penalty to make it the less attractive option than a recall.
Recall isn't necessary. Just have a streamlined no hassles RMA process like they always have. Intel has been a top tier vendor to deal with over RMAs going back as far as I can remember. Just don't goofed that up the way they have everything else of late.
I think the reports of continued stability issues is only with CPUs already affected. Not all of them. Igor as I recall was saying the ones that crash are broken.I'm saying recall under the assumption that Intel can't get the CPUs to reach stability with even the default profiles, i.e., as you mentioned, Igor saying they are basically just broken. In that case, a streamlined RMA isn't going to help.
And the warranty? How's the replacement doing?
I think the reports of continued stability issues is only with CPUs already affected. Not all of them. Igor as I recall was saying the ones that crash are broken.
You don't need any disclaimers, that's all any of us can do is speculate. And judging by those Reuters reports, Intel is almost Nintendo tier in court kung fu.Yeah, I have no idea how many are effected or how all of this will play out, I should have emphasized that I was just speculating based upon a worst case scenario. It's obviously wide spread enough to become a serious problem but who knows the statistics in the end. Intel's PR will take a hit short term but as for long term effects, we'll just have to see how everything plays out.
What is Intel Adaptive Thermal Monitor? (with both frequency/voltage control and clock modulation)
What is Thermal Velocity Boost? (with both Ratio Clipping and Voltage Optimization)
The hardware and software tools are there, engineers did their job. It's the decision makers who played fast and loose, unlike with AMD where somebody in engineering had the last word when it came to enforcing limits.