• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Intel Skylake / Kaby Lake

Page 478 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Could be an offlabel model that is either binned a bit better or slightly higher TDP.
Ars is reporting the new m3 in the MacBook to be Core m3-7Y32 (but with a slightly faster base clock). So it seems to be boosted a bit with a slightly higher TDP.

But the bigger point here is that last year's model is the 6Y30, with the speeds in the MacBook at 1.1 GHz / 2.2 GHz Turbo. Now it's 7Y32 at 1.2 GHz / 3.0 GHz Turbo. That represents an impressive 36% increase in max Turbo speed or an increase of 800 MHz, going from Skylake to Kaby Lake.

Also, this means that MacBook m3 in 2017 is faster than the 2016 m5, and is almost as fast as 2016 m7.
 
Last edited:
Finally an entry that looks like a real desktop product. Confirms 6C/12T for Coffee Lake-S and 12MB L3. Looks like an early stepping, final SKUs should clock better.

Yeah, if they want this SKU to fly, 4.2GHz max single core turbo isn't going to do it. They need >4.5GHz, otherwise 7700K is still probably a better choice.
 
Yeah, if they want this SKU to fly, 4.2GHz max single core turbo isn't going to do it. They need >4.5GHz, otherwise 7700K is still probably a better choice.
Roughly the same clock speeds as the 7700 is all that is needed, imo. 14nm++ will help a little and the extra cores is what will do the trick of bumping multi-thread performance for the desktop line. 7700K will be far behind in multi-thread. The complaints have always been about the number of cores, not about single thread performance. If they get single thread close to the 7700K, that will be just right, imo.

I'm thinking it will be a 7700 with 2 more cores, 3.6 / 4.2, and then everyone will complain about the clock speed, when previously they were complaining about the number of cores. 🙂
 
Are they actually selling RAID "keys" for $99? Somebody please tell me that part was trolling and not accurate.
 
Are they actually selling RAID "keys" for $99? Somebody please tell me that part was trolling and not accurate.

For CPU connected NVMe RAID modes 1/5 that people are literally willing to pay hundreds to thousands for via other means? Yes. They would literally be stupid not to.
 
They would literally be stupid not to.

So I guess they were stupid for every other platform they sold until this one, and just 'wised up'?




I'm at a loss for why anyone would think this is a good idea when their competitor already offers a cheaper and more robust platform with these same features, and doesn't charge extra for them.
 
So I guess they were stupid for every other platform they sold until this one, and just 'wised up'?




I'm at a loss for why anyone would think this is a good idea when their competitor already offers a cheaper and more robust platform with these same features, and doesn't charge extra for them.

False.

You're confusing CPU connected PCI-e RAID with SATA or chipset connected PCI-e raid. This feature did not exist until now, and we don't know if threadripper will have anything equivalent.
 
Threadripper doesn't have a PCI-e chipset, so all PCI-e already comes directly off of the SoC. So we actually do know that they do have an equivalent feature (and more of it), and we can surmise they won't charge for it because they never have before (on Ryzen, which has the same feature) and never said they would.

Lol, okay. Show me some evidence this feature exists on Threadripper and X399.

I have not been able to find anything to indicate that is the case. I've been looking for a while, so I'll be patient.

Edit: I'm glad you came around, but deleting your post makes you look bad, especially after you've been quoted.

The entire thing has been causing a lot of confusion because of this exact thing. We already did this a few days ago and a few pages ago.
 
So I guess they were stupid for every other platform they sold until this one, and just 'wised up'?




I'm at a loss for why anyone would think this is a good idea when their competitor already offers a cheaper and more robust platform with these same features, and doesn't charge extra for them.

Um, what? Literally no other CPU vendor is offering NVMe CPU connected raid. In fact the only way to get it right now is to buy literally hot out of the Fab raid controller ICs from broadcom which retail on cards for roughly $1000.

Also as an FYI, every example of either a TR or EPYC motherboard I've seen does in fact have a chipset connected via PCIe.
 
Are they actually selling RAID "keys" for $99? Somebody please tell me that part was trolling and not accurate.

no its accurate...

which is why i said its like DLC's on motherboards.
They learned from the gaming industry DLC's sell.

And if the Dongles Raid keys do well, i have a sad feeling intel will dongle overclocking on motherboards, and then dongle SLI, or make us buy Deluxe edition motherboards with all the DLC's unlocked.
 
And if the Dongles Raid keys do well, i have a sad feeling intel will dongle overclocking on motherboards, and then dongle SLI, or make us buy Deluxe edition motherboards with all the DLC's unlocked.

I have to disagree with you on this; it's reaching and a bit overreacting in my opinion. This specific CPU RAID has been explained well a few posts up and is not at all the same as the RAID we usually use (however people see it as one and the same for some reason).
 
Ars is reporting the new m3 in the MacBook to be Core m3-7Y32 (but with a slightly faster base clock). So it seems to be boosted a bit with a slightly higher TDP.

But the bigger point here is that last year's model is the 6Y30, with the speeds in the MacBook at 1.1 GHz / 2.2 GHz Turbo. Now it's 7Y32 at 1.2 GHz / 3.0 GHz Turbo. That represents an impressive 36% increase in max Turbo speed or an increase of 800 MHz, going from Skylake to Kaby Lake.

Also, this means that MacBook m3 in 2017 is faster than the 2016 m5, and is almost as fast as 2016 m7.
w00t! I just ordered one of these Kaby Lake Y m3 1.2/3.0 GHz MacBooks. I'm so glad I didn't get a 2016 refurb. This should represent a 150% speed boost over my Core 2 Duo 2.26 GHz P8400.

Plus I also just ordered a Kaby Lake i5-7500 3.4/3.8 GHz iMac. I'd expect maybe a 50% boost in performance over my Core i7 870, but the main reason I wanted it was for everything else. Computing power is secondary.

I have now satisfied my Kaby Lake hunger. 🙂
 
I don't mean to be rude, and I could've easily missed the point, but what has ^^ to do with X299? Aside from that, congrats on your purchase 🙂
 
I don't mean to be rude, and I could've easily missed the point, but what has ^^ to do with X299? Aside from that, congrats on your purchase 🙂
This isn't a thread just about X299. It's the general Skylake / Kaby Lake thread.

BTW, I have just about zero practical interest in X299 since I never buy anything in this space. Like most of the real world, I buy machines with mainstream and low power CPUs.
 
My opinion, I consider Intel Kaby Lake dead; upcoming Coffee Lake increases core count throughout the line-up. I wouldn't consider scratching a Kaby Lake itch,
Except Core m, which will likely remain 2 cores, transitioning to 10 nm Cannon Lake at end of year (Intel target).
 
Yeah, if they want this SKU to fly, 4.2GHz max single core turbo isn't going to do it. They need >4.5GHz, otherwise 7700K is still probably a better choice.

I agree, and give that that bit of info was likely from the first stepping, there is still hope. I think stock 3.1 GHz is a bit low as well. Still two months away so....more to come?
 
Back
Top