Intel Pentium lineup now split into two: Pentium Gold and Pentium Silver...

[jpiniero]

that problem was solved the moment they made the i3s have 4 cores, starting with 4c 3.6GHz for $120, the G4560 for its launch price makes more sense now, increasing the price and adding "gold" without changing anything else doesn't.

Remember that Intel has the 4600 and 4620. The latter has a RCP of $86 - $93. I suppose Intel could have those two go away, and the 4560 become the highest priced Pentium at the highest priced bracket.

 

[Arachnotronic]

It remains to be seen, if they will pull something similar with the i3-8100. But I figure that's where they dump all of their failed 6-core dies.

i3-8100 is based on a quad core Coffee Lake die, not a failed 6-core die.

 

[VirtualLarry]

i3-8100 is based on a quad core Coffee Lake die, not a failed 6-core die.

What proof do we have of that? I would rather believe that Intel made only one new mask set, a 6-core, and uses that same die for the 8700K, 8700, 8600K, 8400, 8350K, 8100.

 

[Arachnotronic]

What proof do we have of that? I would rather believe that Intel made only one new mask set, a 6-core, and uses that same die for the 8700K, 8700, 8600K, 8400, 8350K, 8100.

You have a chip in hand, no? Delid it and see for yourself

But in case you don't want to do that...

 

[dullard]

What proof do we have of that? I would rather believe that Intel made only one new mask set, a 6-core, and uses that same die for the 8700K, 8700, 8600K, 8400, 8350K, 8100.

I was wondering, is it possible that Intel can sell the 8350K as just a rebranded 7700K with hyperthreading turned off and the clocks lowered? Same with the 8100 just a rebranded 7600 with the clocks lowered. The specs are virtually identical otherwise.

That would explain why the 8350K and 8100 have TDP and memory speeds that match Kaby Lake and not Coffee Lake processors.

 

[Arachnotronic]

I was wondering, is it possible that Intel can sell the 8350K as just a rebranded 7700K with hyperthreading turned off and the clocks lowered? Same with the 8100 just a rebranded 7600 with the clocks lowered. The specs are virtually identical otherwise.

That would explain why the 8350K and 8100 have TDP and memory speeds that match Kaby Lake and not Coffee Lake processors.

Nope. CFL has some media engine improvements not present in Kaby Lake.

 

[VirtualLarry]

Nope. CFL has some media engine improvements not present in Kaby Lake.

Are you suggesting that the "UHD 630" graphics on CFL, is in fact somehow different than the "HD 630" present on KBL? I thought that it was just re-branded, and they removed one EU for yield purposes, and compensated by upping the graphics clock by like 5%.

 

[Arachnotronic]

Are you suggesting that the "UHD 630" graphics on CFL, is in fact somehow different than the "HD 630" present on KBL? I thought that it was just re-branded, and they removed one EU for yield purposes, and compensated by upping the graphics clock by like 5%.

Actually, I'm probably wrong. Intel added Rec. 2020 and HDR capability with Kaby Lake. So the media engine appears the same, my bad.

However, the new package design (LGA 1151 v2 or whatever) and the new process (14nm++ vs 14nm+) still distinguish the old 7600K from the new 8350K.

 

[LTC8K6]

Yes, CFL is the first line on 14nm++, but ordinary mortals have no way to confirm that.

KL still supported DDR3, so if the 8100 and 8350K were rebrands, that support would likely still be there.

No idea how we'd check that, either.

Die shots would probably reveal all, of course.

So a rebrand would be hard to hide.

 

[VirtualLarry]

Actually, I'm probably wrong. Intel added Rec. 2020 and HDR capability with Kaby Lake. So the media engine appears the same, my bad.

However, the new package design (LGA 1151 v2 or whatever) and the new process (14nm++ vs 14nm+) still distinguish the old 7600K from the new 8350K.

It's looking more and more like these "new" 4C/4T CPUs on CFL, are simply rebranded KBL silicon, on a new package. Remember, the package pinout, is independent on the silicon "pinout", more or less. (That's kind of like how AMD was able to convert Carrizzo dies into AM4 Bristol Ridge CPUs.)

 

[Jan Olšan]

Based on this thread, it seems that the Coffee quadcores have identical array of components at the bottom as Kaby Lake quads, while hexa Coffee Lakes have different and beefier set. This starts to increasingly look like a straight rebadge, possibly with only the BIOS/ucode locking these chips from working in Z270 and Kaby/Skylakes working in Z370.

7700K
http://hw-museum.cz/data/cpu/pic/Intel_Core_i7-7700K_B.jpg

8350K
https://www.techpowerup.com/reviews/Intel/Core_i3_8350K/images/cpu2.jpg

8700K
https://www.techpowerup.com/reviews/Intel/Core_i7_8700K/images/cpu2.jpg

Also CPU-Z shows 8350K having stepping B0, same as reported for 7700K... 8700K has stepping U0. You can say with 99% certainty that the new i3s are just Kaby Lake slightly modified to work on Z370 platform.

Edit - I wonder if delid would reveal this. Even if Coffee 4c was a new chip, it might look very similar to the Kaby 4c without doing a real (destructive) macro die shot like Fritzchens Fritz does.
AFAIK the document from BenchLife promised a 126mm2 big 4c+gt2 die, unless they meant rebranded Kaby by that mention. Maybe that chip will only come later? Or maybe it was cancelled? The B0 stepping on the Core i3s is quite a strong evidence.

 

[VirtualLarry]

Based on this thread, it seems that the Coffee quadcores have identical array of components at the bottom as Kaby Lake quads, while hexa Coffee Lakes have different and beefier set. This starts to increasingly look like a straight rebadge, possibly with only the BIOS/ucode locking these chips from working in Z270 and Kaby/Skylakes working in Z370.

That would also indicate, that the whole 14nm / 14nm+ / 14nm++ thing is primarily a marketing label, for what is really just a single, continuously-improved process. Which also suggests, current-production 7700K CPUs may well be 14nm++.

Or if they are in fact "different" processes, then if the 8100 and 8350K are re-badged KBL, with a different package pinout, then Intel claiming that 8th Gen / Coffee Lake is all 14nm++, would be stretching the truth a bit.

So, I think that they ARE re-badges, and that the first is true, that there is really only a singular, continuously-improved 14nm, and that the +, ++ labels are all a creation of the marketing dept., and not engineering.

Edit: I'm not claiming that there AREN'T transistor-size / density changes to the mask layout for the 6C/12T dies, they probably did tweak a few parameters for the 6C CFL die.

 

[kwalkingcraze]

Kind of shameful, because the only reason that prices went so high on that CPU, because Intel intentionally held back production, and thus caused prices to rise, then discontinued that CPU, then re-introduced it with re-branding at that newer higher price. Talk about shenanigans from team Blue.

The gap between G4560 at $64.99 originally and i3-6100 at $114.99 is too significant and confusing, so rebranding it as Pentium Gold G4560 at $84.99 or $89.99 will help consolidate the 2C/4T group.

And then Newegg can reliably sell the Gold G4560 for $79.99 everyday price (maybe $84.99).

 

[Abwx]

I was wondering, is it possible that Intel can sell the 8350K as just a rebranded 7700K with hyperthreading turned off and the clocks lowered?

That s a KBL die on the improved process, so same as a 7700K but it should be more efficient.

http://www.hardware.fr/articles/970-3/overclocking-pratique.html

 

[kwalkingcraze]

Remember that Intel has the 4600 and 4620. The latter has a RCP of $86 - $93. I suppose Intel could have those two go away, and the 4560 become the highest priced Pentium at the highest priced bracket.

Pentium Gold G4600 and G4620 are valid models, still available and will be re-launched:

https://www.intel.com/content/www/us/en/products/processors/pentium/g4600.html
https://www.intel.com/content/www/us/en/products/processors/pentium/g4620.html

I'm estimating MSRP to be around $89.99 and $104.99. respectively. Newegg sells them for $86.99 and $99.99.

 

[dark zero]

Stop Intel.

Just make Atom and Celeron the same uArch

And Pentiun and Core the top tiers.

 

[kwalkingcraze]

Now in-stock at Micro Center with Gold packaging. Price increased from $56.99 to $74.99. Since Micro Center sells CPUs for 12% off MSRP, the new MSRP with Gold packaging is now $84.99.

http://www.microcenter.com/product/485530/Pentium_G4560_Kaby_Lake_350GHz_LGA_1151_Boxed_Processor

 

[VirtualLarry]

I don't see any evidence of the "gold" packaging changes in the pictures, nor in the specs, on that linked page.

 

[dave_the_nerd]

Well, they have i9 now, maybe the Pentiums should be i1.

 

[PingSpike]

Well, they have i9 now, maybe the Pentiums should be i1.

Should Intel call the soon to be launched single core kabylake cpu the 486 30th anniversary edition or i0.5?

 

[Jan Olšan]

https://www.anandtech.com/show/12057/intel-to-use-additional-assembly-lines-for-coffee-lake-cpus

So Coffee Lake chips are getting second assembly line - but it only affects the hexa-core versions. Probably another proof that i3-8100/i3-8350K are Kaby Lakes and that's why they show better supply situation.

 

[LTC8K6]

Looks like December 15th is the expected arrival:

Date Customer Must be Ready to Receive Post-Conversion Material: December 15, 2017

 

[VirtualLarry]

https://www.anandtech.com/show/12057/intel-to-use-additional-assembly-lines-for-coffee-lake-cpus

So Coffee Lake chips are getting second assembly line - but it only affects the hexa-core versions. Probably another proof that i3-8100/i3-8350K are Kaby Lakes and that's why they show better supply situation.

Your post should probably go into the "big" CoffeeLake thread, if it hasn't already, but I have another question - since this change order, specifies an additional assembly/test plant, will overclockers want to stick with the "known" Malaysian packaging plant, or will they willingly accept chips made in the new Chengdu, China plant? (Labeled "Made in China" on them.)

I just know how well overclockers from back in the day loved them some Malay chips, and there were actually differences in terms of overclockability, from say, the Costa Rica chips. (Which, I think, the Costa Rica package/test plant is no longer used for that anymore?)

 

[LTC8K6]

Your post should probably go into the "big" CoffeeLake thread, if it hasn't already, but I have another question - since this change order, specifies an additional assembly/test plant, will overclockers want to stick with the "known" Malaysian packaging plant, or will they willingly accept chips made in the new Chengdu, China plant? (Labeled "Made in China" on them.)

I just know how well overclockers from back in the day loved them some Malay chips, and there were actually differences in terms of overclockability, from say, the Costa Rica chips. (Which, I think, the Costa Rica package/test plant is no longer used for that anymore?)

I think it is just assembly? The chips are the same chips?

 

[dark zero]

If at the end the Quads CL are rebranded KBL... They will leave AMD proud since other people does the same.