Rumor ! - Intel to Launch 12c/24t skylake-x cpu's to crush AMD RYZEN

[FPSwithaWacomTablet]

Need to embed your opinion in the initial post. At least that's LTT's rules.

[kladzen]

7 minutes ago, FPSwithaWacomTablet said:

Need to embed your opinion in the initial post. At least that's LTT's rules.

just updated OP with my own speculation and opinion 

[-BirdiE-]

9 hours ago, tomotomov92 said:

Actually Intel is going to change their prices a lot sooner - I think in about 3 to 7 days maybe?

And this is based on your expert analysis?

 

I doubt it. They've already sold millions of chips to computer manufacturers and retailers. They'll either have to wait for all those to sell out, OR offer significant rebates to the manufacturers and retailers and go explain to their investors why the revenue they reported is no longer accurate. Not to mention suddenly dropping the price to be competitive with Ryzen would piss off consumers and, more importantly, the companies that just bought 20,000 Intel machines at a much higher price the day before.

 

The only real remedy for this is to release new/rebranded chips, or drop prices slowly over time.

[Sauron]

that's pretty much irrelevant if it costs 1700$

[Fetzie]

On 27/02/2017 at 3:09 PM, SpaceGhostC2C said:

Indeed, if they make such a move, most likely it won't be to crush AMD at the segment in which they compete, but rather to keep having a very pricey CPU in the segment in which they don't. After all, such product will be attractive to whoever was buying the 10-core 2011-3 CPUs, and irrelevant to the people buying 1151 i7s. The former don't care about absurd prices if they get the top multi-core performance, the latter already can build 3 gaming PCs with the cost of the CPU alone, without using AMD.

Except the people that don't care about money can already buy 22 core 44 thread CPUs (that you can put 4 of on a single motherboard).

 

15 minutes ago, Sauron said:

that's pretty much irrelevant if it costs 1700$

 

If it costs 1700 dollars it'll eat a big hole into their ultra-high margin Xeons.

[Sauron]

19 minutes ago, Fetzie said:

If it costs 1700 dollars it'll eat a big hole into their ultra-high margin Xeons.

All they need to do is sell it without ecc support

[SpaceGhostC2C]

26 minutes ago, Fetzie said:

Except the people that don't care about money can already buy 22 core 44 thread CPUs (that you can put 4 of on a single motherboard).

Not at 6950X clocks, though.

[RagnarokDel]

18 hours ago, Space Reptile said:

its called naples , and goes up to 32c/64t

can you imagine if Naples is overclockable?

[Space Reptile]

9 hours ago, RagnarokDel said:

can you imagine if Naples is overclockable?

maybe , would be silly if tho 

[cj09beira]

14 minutes ago, Space Reptile said:

maybe , would be silly if tho 

If it is i would be interested. Imagine the 3dmark cpu scores

They would be over 9000 literally 

[Space Reptile]

29 minutes ago, cj09beira said:

If it is i would be interested. Imagine the 3dmark cpu scores

They would be over 9000 literally 

imagine the wattage for a 3.5ghz 32c/64t , 9590 aint got nothing on that 

[leadeater]

49 minutes ago, Space Reptile said:

maybe , would be silly if tho 

Depends, if the only SKUs are based on cores then there is no big danger in allowing it. Servers simply won't do it unless it's highly custom ones that need that kind of specialization, Intel actually has high frequency low core count Xeons for these use cases where you would consider doing it. Wouldn't take much for a hardware vendor like HPE, Supermicro or Fujitsu to offer their own server SKUs that are overclocked by default in the firmware.

[Space Reptile]

Just now, leadeater said:

Depends, if the only SKUs are based on cores then there is no big danger in allowing it. Servers simply won't do it unless it's highly custom ones that need that kind of specialization, Intel actually has high frequency low core count Xeons for these use cases where you would consider doing it. Wouldn't take much for a hardware vendor like HPE, Supermicro or Fujitsu to offer their own server SKUs that are overclocked by default in the firmware.

well i expect them to have a core utilization/TDP dependent boost , but i dont think they will have an open multi 

[leadeater]

Just now, Space Reptile said:

well i expect them to have a core utilization/TDP dependent boost , but i dont think they will have an open multi 

Neither but we can dream 

[cj09beira]

11 minutes ago, Space Reptile said:

imagine the wattage for a 3.5ghz 32c/64t , 9590 aint got nothing on that 

I buy a new psu. Power is not a problem bring me the scores!!!

I am being pushed back by my i5 right now

[cj09beira]

5 minutes ago, leadeater said:

Neither but we can dream 

It seems old opterons had unlocked multiplier going down. and you could change bclk thats a start

[FratStar]

The amount of money it would cost any business to produce chips  then test each one then overclock them to make sure it's a golden sample, and finally send it out to reviewers would be astronomical. Especially since they are send quite a few samples out to reviewers.

 

It is very hard to believe that they could afford to do that with their current operating income. These conspiracy theories are just ridiculous. Now do I believe it can be done, yes. Is it so likely to have been done that you can't trust any reviewer that was sent a sample, you are out of your mind if you think that.

 

There's a thin line between healthy skepticism and tin foil hat. Ultimately they are selling a product, but why would they risk it all on their last lifeline people? That's literally the worst tactic in a situation like that.

[TidaLWaveZ]

58 minutes ago, FratStar said:

The amount of money it would cost any business to produce chips  then test each one then overclock them to make sure it's a golden sample, and finally send it out to reviewers would be astronomical. Especially since they are send quite a few samples out to reviewers.

 

It is very hard to believe that they could afford to do that with their current operating income. These conspiracy theories are just ridiculous. Now do I believe it can be done, yes. Is it so likely to have been done that you can't trust any reviewer that was sent a sample, you are out of your mind if you think that.

 

There's a thin line between healthy skepticism and tin foil hat. Ultimately they are selling a product, but why would they risk it all on their last lifeline people? That's literally the worst tactic in a situation like that.

 

You do realize that it's no secret that both intel and AMD have Quality departments that test processors and bin different quality performers, right? No, I don't believe that every processor they produce is extensively tested but I can guarantee a nice sample pool is taken especially prior to a new release. There's no risk involved, the cost of benchmarking processors is not tremendously high or something, like I mentioned earlier they absolutely have a quality department that has been devoted to this type of thing since the companies inception.

 

It's not even close to far-fetched or a "conspiracy theory" to believe that AMD, Nvidia, and Intel all three give out the highest quality review products that they can because they absolutely do.

[MakeAMDGreatAgain]

All they have to do is slash some of their i5 prices. And no, microcenter's IN-STORE ONLY bologna doesn't count.

 

Right now would be the perfect time as people like me don't want to fucking wait til quarter 2 (could be JUNE!) just to upgrade from an FX 6300. When an i5 7600 would be a huge upgrade and be absolutely fine for gaming.

[FratStar]

Just now, TidaLWaveZ said:

 

You do realize that it's no secret that both intel and AMD have Quality departments that test processors and bin different quality performers, right? No, I don't believe that every processor they produce is extensively tested but I can guarantee a nice sample pool is taken especially prior to release. There's no risk involved, the cost of benchmarking processors is not tremendously high or something, like I mentioned earlier thry absolutely have a quality department that has been devoted to this type of thing since the companies inception.

 

It's not even close to far-fetched or a "conspiracy theory" to believe that AMD, Nvidia, and Intel all three give out the highest quality review products that they can because they absolutely do.

You totally misunderstood my entire point. It is very far-fetched to risk your corporate integrity pretty much on your deathbed since you've been bleeding money for over 10+ years. Just to get good reviews because you picked out golden samples then risk, misleading customers is the fastest way to lose in court. Also notice I said for AMD's operating income something like that is just not worth it. Paying people to do something dishonest like that is a waste of time and money when your back is against they wall , it's just bad business. 

 

Why would they open themselves up to:

A. A lawsuit

B. People post the pre-order season not purchasing the product

C. People not having trust in the Ryzen 5 and Ryzen 3 processors due to them royally screwing up with Ryzen 7

D. Losing consumer trust before launching Naples. They need HPE, Dell, and Lenovo to adopt test and validate the product. What do you think would happen if they see that they processors have been misrepresented. They wouldn't even want to do PoC testing on Naples at that point

 

Like I said there's a very thin line between heavy skepticism and tin foil hats. Any one that has even a slight understanding of how business relationships work between partners, JVs, suppliers, etc. Understands that even suggesting something like that is a reach.

 

It's unlikely, period get over it.

 

[TidaLWaveZ]

6 minutes ago, FratStar said:

You totally misunderstood my entire point. It is very far-fetched to risk your corporate integrity pretty much on your deathbed since you've been bleeding money for over 10+ years. Just to get good reviews because you picked out golden samples then risk, misleading customers is the fastest way to lose in court. Also notice I said for AMD's operating income something like that is just not worth it. Paying people to do something dishonest like that is a waste of time and money when your back is against they wall , it's just bad business. 

 

Why would they open themselves up to:

A. A lawsuit

B. People post the pre-order season not purchasing the product

C. People not having trust in the Ryzen 5 and Ryzen 3 processors due to them royally screwing up with Ryzen 7

D. Losing consumer trust before launching Naples. They need HPE, Dell, and Lenovo to adopt test and validate the product. What do you think would happen if they see that they processors have been misrepresented. They wouldn't even want to do PoC testing on Naples at that point

 

Like I said there's a very thin line between heavy skepticism and tin foil hats. Any one that has even a slight understanding of how business relationships work between partners, JVs, suppliers, etc. Understands that even suggesting something like that is a reach.

 

It's unlikely, period get over it.

 

I think you're misunderstanding a point, there's nothing illegal about cherry picking review samples.  Since we've already seen it time and time again obviously it hasn't tarnished the integrity of any of the companies or destroyed consumer trust. The silicon lottery is real, It's known that the companies cherry pick, period get over it.

[blackmambo]

On 2/28/2017 at 11:46 AM, Rika Shiguma said:

remember - Ryzen 5 and 3 are still to come. the cheaper but i guess still good ones. (I'm guessing it's gonna be 6 cores on 5 and 4 cores on 3.)

We know only 2 so far(officially) --

Ryzen R5 1600X = 6c/12t

Ryzen R5 1500X = 4c/8t
Targeted price range --  higher 100$ to about 300$
 

Source: this video Linus & others can't speak about it, coz its still under NDA!

Ryzen R3 might possibly be like i5, i.e. 4c with SMT/HT disabled but at a lower than i3 price point.