- Mar 3, 2017
- 1,684
- 6,227
- 136
The worst part of this, by far the worst part, for either AMD or Intel fans....
is gotta be the fact that in MLID's next video, or on twitter soon, all he is gonna do is gloat 🤣
To be fair, the Kepler/Adroc team was the one most responsible for making Zen 5 IPC into a shibboleth for "legit leakers"
Lol, yes. He’s done this before. Him and Uzzi and Kepler — the whole gang are like this, he’s just the most vocal and explicit. They reset every time.After how many weeks of "just wait a few weeks/days for Computex! You'll see! You'll all see!!!!"
Oh come on.the whole gang are like this, he’s just the most vocal and explicit. They reset every time.
Should be thankful Microsoft didn't push for NPU's on the desktop.
its tested againt the 9950x. imo this is worse than Zen 4. I want to see the breakdown of ipc improvements+16% is geomean Zen 5 vs Zen 4.
Keeping in mind the TDP for 12/8/6 core models is significantly less (in case of 7700X --> 9700X is going from 105W to 65W). We also don't know if the average clocks during full loads are the same or better, either. fMax being the same doesn't mean much if the AVX2 all-core clock is significantly better...
It's funny tho, bcuz in the case of Intel, it looks like he got LNC and SKM switched (huge improvement in little core, smol improvement in big core). But I'm not gonna say anything just yet cuz Intel's official event is tmmrw...Yeah, they really made their bed here.
In the case of MLID, a broken clock is right twice a day.
I think 16% is underwhelming, and deff worse than Zen 4, but I also don't think it's atrocious.its tested againt the 9950x. imo this is worse than Zen 4. I want to see the breakdown of ipc improvements
I did 3800X - 5800X - 5800X3DCPU upgrades with the same socket only make sense when you have really cheap SKUs that you buy with the aim to keep it for the short term, until you get something more expensive that suits your needs.
For AM4, it meant buying a 5600X and waiting for either the 5800X3D for gaming or 5900X/5950X for productivity.
It's fine given the area bump being miniscule, the issue is the server perf projections.I think 16% is underwhelming
Clock rates, TDPs, you got coaxed into a snafu. Such things happen. But two in a row now...It's fine given the area bump being miniscule, the issue is the server perf projections.
its not atrocious but I was one those that believed the hype.but I also don't think it's atrocious.
Clock rates, TDPs, you got coaxed into a snafu. Such things happen. But two in a row now...
No, not that.Clock rates, TDPs, you got coaxed into a snafu. Such things happen. But two in a row now...
AMD hasn't lied or missed a single DC part in Zen history so this would be a first.Which is, I think, why humility and caution in one's public statements can be advisable.
I'm really starting to hate MS. coploit that.
No, not that.
At 16% PPC they're physically not getting that 128c socket score.
96c you can game with magical v/f curves being magically way better, but not 33% more cores ISO-ish node.
AMD hasn't lied or missed a single DC part in Zen history so this would be a first.
Everything since Rome was exactly as promised.
You can thank all the tech giants, imo. Google, MS, Meta, etc.I'm really starting to hate MS. coploit that.