Who here hyped for Another Massive Disappointment? I know I am
my predictions
Ryzen 3000 series 10-14% IPC jump but only 4.4ghz max boost and memory latency issues but power efficient
Navi gpu will be slightly above rx590 with only a slight decrease in power usage.
based retardposter
Prepare to be wrong cuck.
fuck off rudeposter
I'm an AMD owner I just know better to get my hopes up. these are reasonable predicitions
The 12 core has 4550mhz all core and 5ghz single core boost.
>Navi gpu will be slightly above rx590
So the most expensive Navi will be 100-150$? Pretty hard to believe.
not talking about costs probably be 50-100 more MSRP than current radeon series on initial release
straight asspull from jimbob
>I can't show source or anything ;)
can't believe so many people fall for that cock sucking faggots lies
k see you in the next thread
most likely that
I don't get it.
whe exactly is computex?
end of this month/beginning of next
>2 weeks until the absolute shitstorm
can't wait
read the image. I made sure to crop that part for people like you
Who the fuck gets hyped over CPU announcements?
uhhh computer hobbyists
We do.
>laughs at 13% IPC improvement. When this should mean that AMD will now have better performance per watt.
>keep saying that it only will clock to 4.5GHz when it it'll clock that way, with way more efficiency.
Please remember that Athlon 64's were way more efficient than a Pentium 4, and they had slower clocks, but had an IPC and architectural advantage. History is repeating itself.
> 4.4GHz
thats threadripper 2950X user... Considering the jump from zen1's 4.1GHz to zen2's 4.4GHz, expect at least 4.7GHz.
>zen2
I meant zen+
where do you think you are?
expect disappointment then. the architecture of Ryzen won't allow it to go past 4.5ghz I guarantee it
>Ryzen 3000 series 10-14% IPC jump but only 4.4ghz max boost and memory latency issues but power efficient
so.... substantially faster than anything intel has? 4.4ghz + a 10% IPC boost puts it at an effective 4.8Ghz, and AMD is already highly competitive at 4.4 this generation.
Why not? The node jump from 12nm to 7nm is much more significant than zen1's 14nm to 12nm.
> I guarantee it
souce: My dad at Nintendo
because the way it was designed was for a certain target of power and efficiency. it can't go past that even with node shrink. you'll see
>Ryzen won't clock past 3Ghz!
ends up at 4.1Ghz on a MOBILE PROCESS meant for smartphones, licensed from Samsung
>Zen won't hit their goal of +40% IPC improvement!
they achieved +52% IPC
>Ryzen won't allow it to go past 4.5Ghz I guarantee it!
everyone and their mother are talking about an SKU hitting 5Ghz single core boost for days but here's this little faggot talk shit
now enlighten us: why couldn't it clock to 5Ghz on a single core? or hitting an all core similar to 9900k(elvin)?
I never made any prediction about original ryzen but dont let me stop you from constructing that strawman
Node jump doesn't means higher frequencies. Take Ivy bridge or Broadwell/Skylake for example.
Ryzen frequency was limited by arch and not thermals, so unless they've changed something big about it, it shouldn't be much further than Zen+.
>target of power and efficiency
well, 7nm uses much less energy, therefore being much more efficient. Just look at the node specs bro.
> you'll see
kek
14nm was a tweaked mobile node, 7nm is HPC
>Ryzen frequency was limited by arch and not thermals
[citation required]
His ass
16 cores or no buy.
you can tell the AMD fanboys itt secretly know this is probably closer to the truth than not and just have to sit there and seethe
Buildzoid has talked about 16 core CPUs and 4000MHz memory in his latest video about the Biostar X570 leak. He supposedly has seen more X570 boards that he can't talk about because they aren't public knowledge, so likely has insider info.
He also talks about how boards will be required to have active cooling (fans) because of "certain features" that he didn't disclose. AMD is chasing the 5.0GHz, make no mistake.
Try overclocking it, fucktards. You'll hit the voltage wall way faster then you'll hit unsafe temperature.
probably just overheats from trying to make the memory latency not so shit
GIMME THIRDREICHER
Then how do you explain the jump fron 4.1GHz in zen1 to 4.4GHz in zen+?
Literally how new are you? You'll hit the overclocking limit before 1.3V on most current Ryzen chips, you could cool those with the stock cooler if you don't mind 75ºC+ temps.
>AMD is chasing the 5.0GHz, make no mistake.
good to hear.
positive baseless speculation: ahh yes this must be true
realistic fact based speculation: source your dad an nintendo XD (im from /v/ btw do I fit in yet!?)
also make predictions bitches. I will compile them and send a special gift to who is the closest to predicting the release :)
>positive baseless speculation: ahh yes this must be true
>realistic fact based speculation: source your dad an nintendo XD (im from /v/ btw do I fit in yet!?)
You're retarded. These are industry insiders talking about public leaks. They know more than you do, they're just sandbagging details because what they know is under embargo and they'd be fucked hard if they publicly gave out that info.
>also make predictions bitches. I will compile them and send a special gift to who is the closest to predicting the release :)
Back to plebbit with you mong.
The 16 core ryzen will be 5ghz on air, can oc to 5.5ghz with an aio. All mobos support 5000mhz ram.
Navi is 4096 cores @ 2500mhz. Performance is around 1.8x vega for the 400$ flagship.
Source: a few key people
im not from reddit you projecting faggot ass
why the fuck would the ARCHITECTURE prevent it from hitting those clockspeeds, you fucking mongoloid?
change your fucking tone and I might respond to that question
>5ghz on air
That's believable, a big Noctua could pull it off
cuz amd sucks lol
haha yup thats basically it
Just overclock 9900k to 6GHZ, AMD BTFO
seething retard
K
>change your fucking tone and I might respond to that question
we're hitting r*ddit levels that shouldn't even be possible
t. barely functioning retard who can now wipe his own ass
Hello Jow Forums!
Your one and only Jen Hsun Huang here!
How's it going? I hope you're doing better than me, cause my favorite niece Lisa just gave me some pretty bad news. Looks like I might have to leave business soon.
Thank you for supporting me and let me know if you have any questions!
Sincerly,
Jensen
kek
>silicon voltage requirements are determined by the architecture
>too stupid to understand that "voltage wall" is actually a thermal wall
Are you fucking retarded?
it sort of can, depends on pipeline
not x86 CPUs though, and it's easy to check going by OC records.
Realistically how bad can navi be? Knowing AMD they can release Vega three times in a row too. That can be worse than 10% perf improvement.
Is that Jack Ma in drag?
>Navi is 1.8 Vega $400
More like 0.8 Vega for $200.
Friendly reminder:
- Zen 2 will have a 12 cores 24 thread chips with a 5Ghz boost
- Navi will be 1080 Ti performance for $250
we have same prediction every single year, and it never happens every single year
zoomer
Only $680.
Incel is finished!
Why does AMD threadripper not have a normal square shape?
So is it still pretty much confirmed that their 12c24t chip will be the fastest clocked? I have a 1700X already that I use to encode a lot with and I'm undecided between the 12c or the 16c32t model. Even with slower clocks, would the IPC uplift offer better gaming over gen1 Ryzen? I'm also worried about the 16c32t chip being memory bandwidth starved due to only having dual channel on AM4.
I have a 2080Ti and 2x16GB of 3200MHz memory at 14-14-14-34 timings.
zen+ already offers better performance so zen2 certainly will
>So is it still pretty much confirmed that their 12c24t chip will be the fastest clocked?
nobody knows. it's all bullshit until confirmed
>I have a 2080Ti
my condolences.
> 16c/32t
> $680
if that's true, Intel is truly finished
it's been that way since GNC got introduced
it is holding Radeon back.
>fastest consumer GPU available
>my condolences
wut?
>voltage wall is actually a thermal wall
Mind expanding on that line of reasoning?
What's wrong with the 2080Ti other than being expensive?
Bad value and despite spending so much on a GPU he kvetched on his CPU
sizeable chunks of the silicon are devoted to a meme?
also price is kinda important
It's the same hype, rumours, disappointment cycle as every year.
Personally I am more than happy if the power efficiency really is as good as they claim. But if you believe any of the rumours I got some bad news for you.
>Kvetch
Lol no. I had gen1 Ryzen at launch when the only Intel option that didn't cost $1500 for a cpu and $400 for a board was the i7-6700k. Had a 980Ti then, have a 2080Ti now while I wait for new CPUs. If 7nm RyZen is a flop, I'll get something from Intel
Price is subjective. All about what you're willing to spend for your end result of choice . Even with portions dedicated to "meme tech", it manages to out performed every single consumer card on the market in DX11/12
We need a /v/toddler hardware board
meanwhile intel 10 core 1000$
I will not buy a mobo with a loud as fuck fan on it.
Navi with 2070 performance for $450, which is actually quite good but still more expensive than we'd like and not at all competitive and all the shills will be out in force then we forget about it and wait until the new architecture next year where they can hopefully offer 2080ti levels for $550
why would it be loud?
is intel damage control starting this early?
Ready for those 64 threads at 6ghz for 150$?
I wish
But vega 64 is already 2070 performance for 400$.
>silicon voltage requirements are determined by the architecture
Frequency scaling with voltage is determined by architecture (silicone quality plays role here too, but arch is pretty damn important). Try reading posts before throwing your shit around.
>too stupid to understand that "voltage wall" is actually a thermal wall
We're not talking about sub-zero stuff here. On normal ambient temps voltage wall is a fucking voltage wall.
>40mm proprietary "low profile" shitfan
>spinning at several thousand rpm
Yeah, I have no idea why it'd be loud either. Do you not actually remember how awful those little fans were?
Engineering samples already reaching 4.4ish. 10% IPC increase but 4.4ghz max is already disproven by Ryzen 3000's ES beating the 9900k in January. Next?
>Who the fuck gets hyped over CPU announcements?
Mostly gamers, at least on Jow Forums.
Did everyone on this shithole board just memoryhole that AMD publicly demonstrated a 65w TDP Zen 2 8-cote beating a 9900k in CineBench R15?
The previous 65w TDP Zen+ 8-core (the 2700) scores 1561 on that test. The Zen 2 version scored 2057, which is ~30% faster.
The publicly available data are already incompatible with your prediction.
>2400g being weaker than 2200g
>Both are comparable to the ages old i5 2500
>Pcie 4 delayed
>iGPU series next year rather than this
>No Thunderbolt or any cool on-motherboard features till usb4
>Power consumption higher than intel's when taking into consideration the system with motherboard
Yeah, I'm giving up. I waited a year to give AMD a chance.
>Frequency scaling with voltage is determined by architecture (silicone quality plays role here too, but arch is pretty damn important).
So you're decided to just make stuff up then? Frequency scaling with voltage is dependent first and foremost on manufacturing process. Literally everyone knows that Glofo's 14nm LP was a low power design not intended for high clocks and that GloFo's 12nm is based on it. It is common knowledge that the voltage wall is from those process nodes.
>4.4 max boost
reasonable prediction? Nah.
only Intel shills pretend it never happened
I just want my 3400G meme machine
I'm in contact with AdoredTV.
Their 12c 32t chip can boost to at least 5Ghz on half the cores (6)
$680 for that 16 core speed demon is too good to be true.