SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

   Technology StocksAMD, ARMH, INTC, NVDA


Previous 10 Next 10 
To: neolib who wrote (30701)6/14/2019 12:02:49 PM
From: Jamie153
of 35017
 
Late July/early Aug has a lot of companies reporting. CDNS, AMD, Qcom, Apple, Intel, etc.

The two big ones that reported so far are NVDA and Broadcom and both were iffy at best.

MU is coming up June 25th.

OLED is up over 80% so far this year and it's in early August (I think).

I like XLNX. Up over 50% over the last year and up 24% YTD. Down 3 today.

Share KeepReplyMark as Last ReadRead Replies (1)


To: Jamie153 who wrote (30702)6/14/2019 12:26:27 PM
From: neolib
of 35017
 
XLNX is getting some of the 5G love, and of course they should have some traction in AI/ML

I really wish that XLNX and AMD would get more in bed on the AI/ML side, as AMD has the perfect architecture these days with chiplets to make custom server CPUs with XLNX die copackaged, and both are at TSMC. It would be a nice counterpoint to Intel's Altera unit.

Share KeepReplyMark as Last ReadRead Replies (1)


From: neolib6/14/2019 1:32:15 PM
of 35017
 
This article looks at Nvidia's lock on the data center accelerator market share vs Intel's CPU share in the same:

eetimes.com

Share KeepReplyMark as Last Read


To: neolib who wrote (30703)6/14/2019 2:31:18 PM
From: TGPTNDR
of 35017
 
neo, Re: AMD has the perfect architecture these days with chiplets to make custom server CPUs>

IMO the best reason for chiplets is to get heat out of the area of the CPU.

On those pictures from a week or so ago did you notice the reflection of the photographers face on one CPU? In another picture there was a reflection of a hand with middle finger extended.

I'd guess that kind of milling and polishing to get the fit of the cooler across the CPU and the auxiliary chips does not come cheap. It looked like mirror finish to me.

Mirrors and window glass are "Finished" by floating on liquid Tin. That won't work for stuff with solder which melts at lots lower temperatures.

It'll get figured out, but I doubt mirror finishes and thick coatings of silver in grease will be a long time solution.

MCMs just move the heat problem around. (as far as I can fathom, else MOBOs could do the job. ) The next real problem(IMO) is Multi chip real coolers.

Still in for around 6K shares, down from 10,

-tgp

Share KeepReplyMark as Last ReadRead Replies (2)


To: TGPTNDR who wrote (30705)6/14/2019 3:10:43 PM
From: neolib
of 35017
 
One of the articles I linked recently, directly commented on that issue, but it was wrt to the chiplet die attached to the substrate, in order to get all the chiplet surfaces at the same height for the heatspreader. The way they did it was via careful control of the copper posts they are using for the interconnects. They claimed it gave better stack control than plain solder bumps.

I think by far the big reason for chiplets is yield. It allows a 64 core chip to be implemented on 7nm but with only 70mm2 die or so. That can yield, whereas Intel's 500mm2 monsters can't.

As to whether there is any thermal advantage, that I'm less sure of, because a little thickness in the heatsink at the contact can spread that laterally too. As far as the die is concerned, the heat per unit area is approximately independent of adjacent heat sources, its assumed the heat all goes normal to the surface and into the heatsink. Its just the delta between the silicon and the heatsink that counts.

Share KeepReplyMark as Last Read


From: neolib6/14/2019 3:20:33 PM
of 35017
 
GF & Soitec in WSA:

ebnonline.com

Share KeepReplyMark as Last Read


From: neolib6/14/2019 3:52:58 PM
of 35017
 
Someone must have a very large sell at 30.50 order hanging out there:


Share KeepReplyMark as Last Read


To: TGPTNDR who wrote (30705)6/14/2019 6:28:29 PM
From: Pravin Kamdar
of 35017
 
IMO the best reason for chiplets is to get heat out of the area of the CPU.

I'd say that is a really good benefit, but the best reason is chiplet yield -- vs monolithic large dies -- and all of the binning options that come with it. Also, chiplets allow you to update one kind of chiplet in the package without having to update an entire large monolithic die.

Pravin

Share KeepReplyMark as Last ReadRead Replies (1)


To: Pravin Kamdar who wrote (30709)6/14/2019 7:18:13 PM
From: neolib
of 35017
 
Not to mention, that the same CPU chiplet is used in all of AMD CPUs from low end DT to high end server. This time around it might be true of the APUs as well, but we don't actually know that yet. AMD has saved a lot of money having one common die.

Share KeepReplyMark as Last ReadRead Replies (1)


To: neolib who wrote (30710)6/14/2019 8:24:51 PM
From: combjelly
of 35017
 
It also makes "custom" chips a lot easier to do. I have no idea what the NREs are on an organic interposer, but it has got to be a whole lot less than a new mask set for a processor. Even if a new chiplet needs to be designed, the total costs are still a lot lower.

Share KeepReplyMark as Last ReadRead Replies (1)
Previous 10 Next 10