- Mar 3, 2017
- 1,623
- 5,894
- 136
This has nothing to with my text. The 17 percent are NOT taken out of a fake leak. He said, an AMD employee communicated to him in april, that AMD says to partners that they should expect around 17% IPC increase.Did you miss the part where the leak was fake and MLID fell for it?
It was a ruse.This has nothing to with my text. The 17 percent are NOT taken out of a fake leak. He said, an AMD employee communicated to him in april, that AMD says to partners that they should expect around 17% IPC increase.
mss is worthless unless you maintain the margin target.
well, fatter PCH with x8 uplink for once.
But again, you can do platform change for truly arbitrary reasons.
Oh my god am I confused right now... guess I will wait for official numbers and call it a day.
I dislike MLID as well but I won't dismiss his leaks totally cause as for any leaks, things could change. Now we have two parties which are providing two predictions...
Unlike RDNA3, there is no wiggle room here, only one party will be right here...we should know more in coming months and see which one is correct. Hint from Allthewatts:
- MLID +20% IPC
- @adroc_thurston and @Kepler_L2 +40% IPC with $999 price tags
Performance and IPC should not differ much with Zen 5 because clocks are rather similar.Without going back through the thread I am sure that the 40% figure is performance not IPC and is specific to SIR2017 ST. If it is correct we wait and see, if it translates to other workloads we also wait and see but this kind of deviation from the original claim which will then get propagated is not great. Stick to the claim as stated rather than generalising because I am sure there are people who are waiting to jump in and shout 'told you they were lying' because it gets a different result in a different benchmark or test.
Without going back through the thread I am sure that the 40% figure is performance not IPC and is specific to SIR2017 ST. If it is correct we wait and see, if it translates to other workloads we also wait and see but this kind of deviation from the original claim which will then get propagated is not great. Stick to the claim as stated rather than generalising because I am sure there are people who are waiting to jump in and shout 'told you they were lying' because it gets a different result in a different benchmark or test.
N31 does clock comfortably over 3Ghz in games.
Just at 500+W, not the planned 350. Nobody saw it coming, AMD caught it way too late. A respin costs 100M+ and would take 6 months or so, so AMD decided to take the L, do a big investigation and reorganise the BU and roadmap.
Most stuff from 3 years ago is out of date, every unit in AMD has been reorganised to some degree since then.
Agreed fully, the fact that RDNA 3 could easily handle OC with large uplifts showed that it was meant to run at more than 2.5Ghz...but there's little excuse left for how misleading the marketing claims were then.
Question though, that respin, couldn't it apply to N32? It came out almost 9 months after the rest.
View attachment 96414
The IPC I referred is what AMD shown us when they announced Zen4...AMD claimed +13% IPC uplift compared to Zen3 at specified clock speed (4GHz in this case)...
No, but I have worked for so long at this point for so many companies that if you have lived in the U.S. at least, your life has been somehow been influenced by my code.Who are you? The Dumbledore of software devs?
I have to speak up and say that the price tag is most certainly a prediction and not any type of leak or anything like that. AMD will charge what they think they can get away with. If Intel or competitors do not have a good counter, no price tag will be too high.I dislike MLID as well but I won't dismiss his leaks totally cause as for any leaks, things could change. Now we have two parties which are providing two predictions...
Unlike RDNA3, there is no wiggle room here, only one party will be right here...we should know more in coming months and see which one is correct. Hint from Allthewatts:
- MLID +20% IPC
- @adroc_thurston and @Kepler_L2 +40% IPC with $999 price tags
Sorry, you must be new here.Unlike RDNA3, there is no wiggle room here, only one party will be right here...we should know more in coming months and see which one is correct.
- MLID +20% IPC
- @adroc_thurston and @Kepler_L2 +40% IPC with $999 price tags
But I want my vita-ramen or wherever that food is called…..I wish people would stop watching that bozo so he doesn't get monetization by spewing fecal matter. People can't help themselves it seems. Hopefully this will change some minds.
Yeah, I got what you meant. I am creating Excel sheet to calculate percentage improvement. What formula should I use to normalize clock speed to 4GHz??Yet here is the specint result.
When you normalise for clockspeed, calculate the delta in each test and geomean the result you get a 13% IPC increase.. Funny that isnt it. Same thing happened with Zen 2 to Zen 3. Almost as if AMD take the nT SpecInt IPC uplift as their target marketing number and slap in a bunch of real world results that geomean to that figure.
Quick edit to add. This is why remembering the original claim is important because when AMD do release their IPC or performance uplift figures at some point they may differ from SpecINT 2017 ST... since chances are that marketing number is the nT figure.
I am not making any guesses as to the accuracy of the claims made but I am saying that when we revisit this after we have actual benchmarks in hand make sure you revisit on the basis of the claim made, not the basis of a generalised version that is not accurate.
I love this, however Zen 4 does have killer IPC compared to the competition and RDNA3 was supposed to clock higher. It was a busted product and they have since solved the issue.Sorry, you must be new here.
"My sources got old data". "My source tested an old firmware". "My IPC figure didnn't count games". "My IPC figure got no AVX-512".
The same clowns pushed the Zen 4 killer IPC or RDNA3 brutal leakz. Navi 33 having 4+k stream processors, being a RV770 reborn, with a "whole new uArch" argument countering the questions.
Yeah, I got what you meant. I am creating Excel sheet to calculate percentage improvement. What formula should I use to normalize clock speed to 4GHz??
I love this, however Zen 4 does have killer IPC compared to the competition and RDNA3 was supposed to clock higher. It was a busted product and they have since solved the issue.
Hype most certainly got out of control, however.
??I love this, however Zen 4 does have killer IPC compared to the competition
This. Also, increasing IPC is hard, 13% is a great result to say nothing of the accelerated AVX-512 instructions, which technically provide a much bigger uplift.What is the IPC increase from alder to raptor to raptor refresh vs the ipc uplift from Zen 3 to Zen 4? Practically zero vs 13%.
Basically on par with Raptor Lake (maybe a few % lower) while having used less resources. Intel current Cove cores are so bloated it's not even funny.