Kaby Lake Pentiums will now have 4 threads

>>58414456
explain yourself mister

>>58415143
By giving Pentiums hyperthreading they are effectively moving their entire lineup one slot down.

Expect i3s with 4 cores and i5s with hyperthreading by the end of this year, mark my words.

>>58414456
THANK YOU BASED AMD

>>58415180
Thanks mister.

>>58415180
theres already an i3 getting hyperthreading

>>58415180
what would be the point of i7?

>>58415197
>theres already an i3 getting hyperthreading
.......................
......................
.....................
....................
...................
..................
.................
................
...............
..............
.............
............
...........
..........
.........
........
.......
......
.....
....
...
..
.
wot

Cool maybe I can get a mainstream 8 core when I upgrade in 2020.

They are nervous.

>>58415217
i3's have HT for awhile now

>>58415210
6-10 cores, obviously

>>58415143
Kaby Lake Pentiums are now gimped i7's.

Previous-gen Pentiums were gimped i5's and i3's.

>>58415407
Thanks mam

>>58415407
wot?
Pentiums are always two cores, so in no way is it like an i7. Pentiums have nothing in common with an i5 either, enough to be brought into the discussion. The only thing being gimped is you.
But seriously, desktops i7 should be six core, i5 should have HT, i3 can die in a fire(or be xeon only), and the pentium should take it's place. Better branding.

>>58414456
But I only see one thread in the catalogue. Explain yourself.

>>58415386
Threads are not cores.

It would still lack AVX, though.

>>58416098
Yes, thank you for that. What does that have to do with anything?

>>58415197
Then i5 would probably be running with 6 cores

>>58415197
Is this like an alternative reality thread? i3's have had hyper threading since Clarksdale, i.e. forever

>>58416114
In his post he stated expect 4 core i3s and i5 with 4 cores and hyperthread, then you said i3s have had hyperthreading. his post has to do with everything.

File: 1483812240209.png (418KB, 627x627px) Image search: [Google]

418KB, 627x627px

>>58414456

I wonder why

>>58416216
No.
I am responding to the following idiots:>>58415386
>>58415217
>>58415197
Who believe current i3's have no hyperthreading.

File: 1483618778347.png (55KB, 217x190px) Image search: [Google]

55KB, 217x190px

Here's a KabyLake Pentium G4560 running various games at 1080p
idk the settings it's running them at since it's all in chinese

Overwatch
https://www.youtube.com/watch?v=61ZTJzhz8iI
GTA 5
https://www.youtube.com/watch?v=aDML9Gq5kas
BF1
https://www.youtube.com/watch?v=Mq4Pw5guGU8

>>58414456
Great I'm building a mini PC now

>>58415217
My Ivy Bridge i3 has HT. What you talking about?

>>58414456
...and Coffee Lake is getting 6 cores on mainstream next year. Zen have really pressured Intel to stop stagnating.

so when pentium 5 will work at 5 GHz?

>>58415180
>end of this year

You mean Skylake-X? Extreme edition ones? Mainstream i7 hexacore will be implemented at Coffeelake or Cannonlake..or Icelake? Damn Intel why so many releases!

>>58420071
Not end of this year, but some time next year. That would be Coffee Lake which is before Cannonlake.

>>58420071
Coffeelake? That is a thing?
>https://en.wikipedia.org/wiki/Coffee_Lake
>Wikipedia citing WTFTECH as a source
Can I get a credible non-rumour source? Google only shows "maybe" and "rumoured" from around the world.

>>58415302

Will I need to upgrade my 2500k by then?

>>58414456
Does that mean I can finally game on a pentium box? nice

What's the main difference between Pentiums and i3s now?

>>58415180
My glorious Pentium 4 has hyperthreading.

>>58420157
Coffee lake is confirmed by reputable sources a long time ago.
https://semiaccurate.com/2016/12/28/coffee-lake-says-dire-things-intels-10nm-problems/
http://pc.watch.impress.co.jp/docs/column/ubiq/1010830.html

>>58421207
Higher clock speed, larger cache.

Well I just got a skylake i3 I guess I'll return it and get the new pentium.

>>58415210
The i3, i5 and i7 branding will stay, they'll just bump number of cores and threads. Pentiums will become 2c 4t, i3 are going to have 4c 4t (they'll pretty much become i5), i5 are going to have 4c 8t and entry level i7 are going to have 6/8c with 12/16t. High end i7 are probably going to have 10/12 cores and 20/22 threads.

>>58424401
so pentium is the new i3, i3 is the new i5, i5 ... and the new i7 is just a waste of money

>>58415197
he means they are leveling up. the pentiums are nearly i3 performance now and they are gonna increase the performance in i3 and i5 in the next release too

>>58414456
>hyperthreading
Literally cancer and adds tons of delay

>>58424401
No, Pentiums will straddle the line between 2c/2t and 2c/4t but with 3MB of L3$. i3s will be 2c/4t with 3MB or 4MB of L3$ and higher clocks. i5s and i7s are going to stay the same. They offer much more performance than anything AMD can conjure up thanks to their massively higher clock frequencies.

If that's the case Zen is DOA.

>>58426301
> i5s and i7s are going to stay the same
Coffee lake will bring 6 cores to mainsteam i7s. No word how intel will position i5s and i3s

>>58426322
Zen is the reason for Intel doing these changes. Intel traditionally doesn't lower prices but they give more performance for the same price.

File: 1482384848920.gif (1003KB, 322x251px) Image search: [Google]

1003KB, 322x251px

>>58426322
Zen isn't even released and intel is on the defense and giving their slaves better shit they didn't get since 5 years

>>58426362
>No word how intel will position i5s and i3s
Why the fuck would intel change what they already have? An i3-7100 runs circles around the A12-9800 and will go head-to-head with any Zen APU on the horizon. Pentiums will do the same to A10s and so on. i5-6500s can beat even the full 8-core Ryzen in most games by a handful of frames. Intel need not change their pricing at all.

>>58426458
i3 and i5s will be competing against high clocked +4.0GHz 4c/8t Zens, enough reason to add hyperthreading to i5.

>>58426616
No they won't. The cheapest Ryzen will not go under $250, so the i5s are safe from competition. Plus, it's already been shown that an i5-6500 has nothing to worry about even from an eight-core Ryzen.

>>58426657
Chink OEM leaks priced 4c/8t Zen at $150 mark. If it instead priced at 250, sure you would be right.

>>58426683
Don't believe those chink leaks. They've been outright denied by the manufactures instead of being obfuscated.

>>58426275
proff

>>58420163
Nope :^)

>>58426379
>tfw amd isn't even president yet and I'm already tired of winning
No we just need all the binary blobs removed

>>58421207

Slightly higher clock speed (between 400 and 500 MHz, when comparing between Kaby Lakes), higher cache (when comparing against Core i3s that aren't Core i3-7100), AVX2 instructions, better iGPU (when comparing against Pentium G4560).

>>58414456
Yup, $75 for an intel 4 thread CPU at 3.5ghz or more

It's also why intel is going to release an overclocking i3

>Pentiums with HT
>i3's now overclockable

What else?

File: 5955af491e2a471e8e08b140208ea727[1].jpg (89KB, 750x1023px) Image search: [Google]

89KB, 750x1023px

Will they give HT to the celeron lineup?

>>58417080
>GTA V
>80 FPS
bullshit
looks jittery as fuck

>>58433626
The FPS is correct and averaged over a time period of like 500ms. That's jitter normal for hard CPU bottleneck. Like if your CPU is really tapped out completely, you get those jitters. Frame rate can be high, but frame time recording for each presentation will show the horrible variance between frame times. You will not see the jitters in frame rate because that's averaged. Unless you update it for every single frame, which you'd normally not do.

>>58415180
>Expect i3s with 4 cores and i5s with hyperthreading by the end of this year, mark my words.

Yet still not unlocked.

>>58433768
Actually, you know what, you still couldn't see them in the frame rate even if you did update every single frame since you'd need to average a few previous anyway so it wouldn't quite show the real picture.

Only recording frame time will show the jitters. Assuming it's not a video encoding problem, but it probably isn't because I used to get shit like this when I was heavily CPU bottlenecked.

>>58415180
and why people want amd to die again?

>>58433902
People don't want AMD to die, they are just butthurt that AMD hasn't released anything that doesn't suck dick in a very long time.

>>58434042
They are about to release something that does not suck dick.

>>58434258
That's what they said about Bulldozer. So pardon a little bit of cynicism.

>>58434282
Bulldozer was not made by certified shitwrecker. K7 and K8 were and they demolished Intel.

>>58434353
Amd has never demolished Intel. They had an advantage at several periods in the past, and better arch during p4 era, but never was it so bad in amds favor as it is today in intels favor

I mean I believe they have a competitive product in the works, but I'm not gonna waste any energy on caring until it is in the hands of reviewers.

>>58415558
>pentiums are always two cores
>except for all of the single core pentiums that exist
nice one you fucking idiot

>>58434441
It was already reviewed, A0 sample showed IPC parity with Broadwell-E. Now it also has higher clocks.