AMD Carrizo Pre-release thread

[frozentundra123456]

1 month later...
http://www.amazon.com/HP-Touchscree...8&qid=1459272513&sr=8-1&keywords=amd+fx-8800p

15.6" tablet with...1366*768 display.

Arrrgh, Hulk hold tablet in one hand!!!

 

[SPBHM]

Don't forget the weird mismatched 6GB RAM configuration, I'm sure that will really help graphics performance

if it's 4+2GB it should be fine, with performance more like dual channel, it's a lot better than single channel.

 

[ShintaiDK]

15.6" tablet with...1366*768 display.

For the visual impaired but strong upper body?

 

[frozentundra123456]

I actually love the concept of these detachables. But no matter what the hardware, it never seems quite right for one or the other. Either it is too big for a tablet (like this case obviously) or too small for the best laptop experience. Probably better to get a nice laptop or ultrabook and a cheap tablet rather than something that tries to be all things and doesnt quite fit anything.

 

[nismotigerwvu]

I actually love the concept of these detachables. But no matter what the hardware, it never seems quite right for one or the other. Either it is too big for a tablet (like this case obviously) or too small for the best laptop experience. Probably better to get a nice laptop or ultrabook and a cheap tablet rather than something that tries to be all things and doesnt quite fit anything.

I've been super happy with my Dell Venue 11 Pro. Granted I did do a BIOS modification to unlock the TDP and raise the thermal throttling point of the i5-4300Y a bit. Dell capped the TDP at 6 watts, but device runs well at 9. Additionally the throttle point was set awfully low at like 58~60C and I raised this to 72. Both of these modifications are still below the specs given by intel so there's no reason to believe they aren't safe, it just means the fans spin up under heavy use to a noticeable level where with factory settings I wasn't even sure if there were fans. Anyways, I use this machine as my primary computer at my office (I've got an external monitor well as a keyboard and mouse connected to the dock) and I carry the tablet half with the stylus to meetings for jotting down notes. When I travel, the keyboard works well enough. The extra battery life is a major plus and while the keyboard is decent to type on, I still really dislike trackpads in general and usually just poke/prod the screen unless I REALLY have to (or use a bluetooth mouse). I'd say it's a solid 8/10 device.

 

[NTMBK]

Oh look, Bristol Ridge has the same number of GPU CUs as Carrizo.

Looks like Bits n Chips were talking out of their hat again: http://www.bitsandchips.it/english/52-english-news/6676-rumor-bristol-bridge-could-have-a-16cu-gpu

 

[The Stilt]

It is identical to Carrizo

The polyshot appears to be bit different, but just because of different cropping (BR shot shows various PHYs while Carrizo does not).

Basically the only difference is the higher clocks, higher TDP options and AM4 availability.

 

[rainy]

Oh look, Bristol Ridge has the same number of GPU CUs as Carrizo.

It's hardly surprising - idea of adding more CU without solving memory bandwith issue would be insane.

Basically the only difference is the higher clocks, higher TDP options and AM4 availability.

Have they improved power management as well (similar to Richland vs Trinity)?

 

[AtenRa]

So it is official that BristolRidge will be launch in Computex 2016 (31 May), only two months to go.

edit. AT article
http://www.anandtech.com/show/10223/amd-bristol-ridge-in-notebooks

 

[deasd]

Not the same die, exactly. Having DDR4 controler might change something and not only die floorplan. But additionally BR has so much circuit which other than DDR4 controller, this might be related to AM4 implementation which has 400 more pins socket. (USB?SATA?PCIE?)

 

[AtenRa]

Not the same die, exactly. Having DDR4 controler might change something and not only die floorplan. But additionally BR has so much circuit which other than DDR4 controller, this might be related to AM4 implementation which has 400 more pins socket. (USB?SATA?PCIE?)

From the two pictures above it looks to be exactly the same die.

 

[deasd]

From the two pictures above it looks to be exactly the same die.

You cannot exclude those additional circuit around, although they're not core and cache.....

 

[AtenRa]

You cannot exclude those additional circuit around, although they're not core and cache.....

Ehm no, those are parts of other Carrizo/BristolRidge dies from the wafer

 

[deasd]

Ehm no, those are parts of other Carrizo/BristolRidge dies from the wafer

no. Watch carefully again, those circuit/unit are not from other die.

 

[NTMBK]

no. Watch carefully again, those circuit/unit are not from other die.

I think it's you who needs to look again Ignore the coloration, look at the structures. They're repeated from the opposite edge of the die.

 

[AtenRa]

no. Watch carefully again, those circuit/unit are not from other die.

They are the same dies from the wafer, AMD just painted them with different colors to make the PDF picture different than the 6th Generation (Carrizo).

 

[The Stilt]

Have they improved power management as well (similar to Richland vs Trinity)?

As far as I know, no.
They can alter the clock stretching and L2 cache downsize parameters to make the power management more aggressive, however Carrizo is already using all the tricks which are supported by the hardware.

 

[The Stilt]

Having DDR4 controler might change something and not only die floorplan.

It's the same exact die. Carrizo already has a DDR4 controller (DDR3 & DDR4 combo).

 

[Triskain]

They did in fact improve the power management, where else would the clock increases materialize from? This presentation from the ISSCC this year has more info, for those who have access. For those who don't, here are some of the highlights:

- Reliability Tracker (+100MHz)

- DLDO on the cores for C-State Boost (+6%)

- improved STAPM

- Boot-time calibration for the power supply (saves 30mV)

- Shadow P-States (+100MHz)

Most of these changes were implemented through changes in the SMU Firmware and required no special hardware, except the already existing.

 

[deasd]

Oh, it seems that I figure it wrongly. Looks like they're same die. Makes me wonder what's the usage of those extra pins on socket.

Looks like Bits n Chips were talking out of their hat again: http://www.bitsandchips.it/english/52-english-news/6676-rumor-bristol-bridge-could-have-a-16cu-gpu

The author explain 16CU SKUs might be for console......maybe.

 

[NTMBK]

Makes me wonder what's the usage of those extra pins on socket.

My theory is that they are there for extra PCIe lanes on future chips. Carrizo only has PCIe3x8 for attaching graphics. Hopefully Zen has a full x16!

 

[deasd]

So many BristolRidge geekbench score appears:
https://browser.primatelabs.com/geekbench3/search?q=AMD+family+21+model+101

 

[Flash831]

As far as I know, no.
They can alter the clock stretching and L2 cache downsize parameters to make the power management more aggressive, however Carrizo is already using all the tricks which are supported by the hardware.

According to Sweclockers, AMD hade activated some features that were inactivated in Carrizo.

They did not say which features, though.

 

[monstercameron]

According to Sweclockers, AMD hade activated some features that were inactivated in Carrizo.

They did not say which features, though.

Probably whatever uvd functionality that was/is in stoneyridge.

 

[monstercameron]

http://www.pcper.com/news/Processors/Early-Availability-AMDs-Bristol-Ridge-7th-Gen-APU

broke the nda?

AMD's 7th Generation A-Series Processors will not be released until Computex, which starts on May 31st. That said, they surprised us with “early availability” in the HP Envy x360 laptop. Bristol Ridge is based on their Excavator core architecture, introduced with Carrizo last year, although AMD claims that this revision has even higher performance, both x86 and graphics. They don't say exactly where it came from, but they cite results of about 12% over Carrizo in PCMark 8, both at 15W.

amd-2016-bristolridge.png
And that's... really about it. It's interesting how AMD could squeeze that much extra performance without changing the wattage or moving to a different fabrication process, but that is what they claim. Carrizo was already intended to be a highly-refined part when it launched. That said, a year is a year, and engineers will continually find better solutions if they are given the opportunity.

While this is partially a pre-announcement for the processor's launch at Computex, it would be interesting to know when the HP Envy x360 will ship with the new processor. It sounds like it will be soon, so we'll be sure to keep a look out.