AMD Carrizo Pre-release thread

[Abwx]

so nobody really knows how good the chip is.

Lol, there s tons of infos and measurements provided by someone who has the product in hand, so if i understand well he s lying and providing made up numbers...?..

And dont forget, AMD can be sued if they provided made up datas, and they published a lot since they are 100% sure of their numbers.

As for waiting for other sources, thank you for the insight, after you took at face value numbers posted by a turkish site that had no sample in hand, i guess that it s more credible than someone who talk of what he s testing, true that the brand wasnt the same, who would had guessed..?..

Also you could as well check here in the waiting :

https://browser.primatelabs.com/geekbench3/search?q=AMD+CARRIZO

 

[mrmt]

If you're referring to the early ES Geekbench leaks, then any poster had the same problem: find the relevant information. See my blog for the latest article on those results. With more samples now there one could identify 2 clusters of FX-8800P GB results (15W and 35W), and some other cluster (e.g. Android tests). Putting the old values into that context shows, that they were some of the worst scores ever seen, barely reaching the lower of the current clusters (likely 15W settings).

Given AMD track record and AMD latest financial numbers, what should Engineer/Scientist logic dictates as the most probable cause for Carrizo mediocre sales on the notebook market:

1) OEMs committing harakiri and shunning what is a very good product from AMD.

2) A mediocre product, only good on AMD powerpoint slides and in cherry picked benchmarks, unlikely to compete with Intel offers on performance and cost structure.

I wait your answer.

 

[mrmt]

So according to some guys here Carrizo flopped already because OEMs don't use it. I mean since when has OEMs opinion been relevant to what is a good product and what isn't. They are the ones building the world's worst notebooks still using HDDs even in 1k+ notebooks. A joke they are.

You don't buy a notebook processor, you buy a notebook, the CPU is just a component. So for all intents and purposes the OEM opinion is the only opinion that matters for AMD, and guess what, OEMs said that Carrizo is a no-go.

 

[maarten12100]

What is the point of a chip? Sales
Who makes products to sell? Wow I believe it is the OEMs.
So by definition, fairly or not, the OEMs basically control AMD's success. Call them a joke if you want, but it would behoove AMD to co-operate with them, or design their own product.

Yeah as if they can compete with Intel in throwing money at OEMs and making a loss on chips. The best they could do is design something themselves and actually bring it to market.

But really we have no, zero, benchmarks on retail products from independent sources, so nobody really knows how good the chip is. And if you want to call something a joke, that Tom's article would be at the top of the list. It contained no independent test results at all, just basically a regurgitation of AMD's PR.

"We have actually seen first-hand what Carrizo can do, and feel like this will be a game changer for the mobile market. It lays the foundation for a future of very efficient, very powerful HSA enabled processors."
Hands on experience means they actually do if it is good or not at Tom's
http://www.tomshardware.com/news/amd-carrizo-apu-6th-gen-processor,29230.html

You don't buy a notebook processor, you buy a notebook, the CPU is just a component. So for all intents and purposes the OEM opinion is the only opinion that matters for AMD, and guess what, OEMs said that Carrizo is a no-go.

Well OEMs rather sell more expensive worse notebooks I get that. Carrizo will get some notebooks to keep the anti Intel anti trust at bay and that is probably where the story ends. Long live no competition. Carrizo is obviously a much more fitting product for the market than Kaveri was and even that got a couple of notebooks. While it was horrid.

 

[mrmt]

Well OEMs rather sell more expensive worse notebooks I get that. Carrizo will get some notebooks to keep the anti Intel anti trust at bay and that is probably where the story ends. Long live no competition. Carrizo is obviously a much more fitting product for the market than Kaveri was and even that got a couple of notebooks. While it was horrid.

How do you explain the fact that AMD did increase its market share with the cat family? See, when AMD get good products out of the door the market responds and buy them, when they field crappy CMT designs, the market doesn't.

 

[monstercameron]

getting poopoo'd on by the usual suspect, lets just wait for more data.

 

[mrmt]

getting poopoo'd on by the usual suspect, lets just wait for more data.

It works like a charm with AMD products, doesn't it? AMD products always end up performing better in terms of advertised performance and in terms of overall performance on the market.

 

[maarten12100]

How do you explain the fact that AMD did increase its market share with the cat family? See, when AMD get good products out of the door the market responds and buy them, when they field crappy CMT designs, the market doesn't.

They were still being put in the worst possible notebooks pretty much. People still buy because those are the entry level notebooks for people that will be running MS Word mostly.

 

[monstercameron]

It works like a charm with AMD products, doesn't it? AMD products always end up performing better in terms of advertised performance and in terms of overall performance on the market.

not every product can succeed, sometimes you are just unlucky. As for AMD I liked my bobcat netbook, kaveri desktop, mullins laptop, mullins fitlet micropc and kaveri laptop, so I am obviously biased. Carrizo looks promising so I'll probably pick it up...atleast I practice what I preach.

 

[mrmt]

They were still being put in the worst possible notebooks pretty much. People still buy because those are the entry level notebooks for people that will be running MS Word mostly.

Well, now Carrizo don't creep in even in these bottom of the barrel notebooks. That says volumes on how bad the product is.

not every product can succeed, sometimes you are just unlucky. As for AMD I liked my bobcat netbook, kaveri desktop, mullins laptop, mullins fitlet micropc and kaveri laptop, so I am obviously biased. Carrizo looks promising so I'll probably pick it up...atleast I practice what I preach.

The fact that you liked it doesn't say much about AMD marketing, supply chain or about the cost structure of the products, you know, the BUSINESS features of the product that matters far more than technical bullet points. What you call lack of luck is in fact AMD's incompetence.

 

[maarten12100]

Well, now Carrizo don't creep in even in these bottom of the barrel notebooks. That says volumes on how bad the product is.

The HP Pavilion and HP Envy and Acer's line up and Asus' line up and Toshiba's lineup. Albeit all probably horrible still mean that it will be at the very least be in bottom of the barrel notebooks.

Which is not a positive thing but to say that Carrizo isn't coming at all is just false.

 

[dark zero]

Well, now Carrizo don't creep in even in these bottom of the barrel notebooks. That says volumes on how bad the product is.
.

If you were on the 3rd world... even Intel U is dissapearing of the markets and I am seeing only crappy Bay Trails!

 

[Abwx]

The HP Pavilion and HP Envy and Acer's line up and Asus' line up and Toshiba's lineup. Albeit all probably horrible still mean that it will be at the very least be in bottom of the barrel notebooks.

Which is not a positive thing but to say that Carrizo isn't coming at all is just false.

Are you sure that it s bottom of the barrel or did you just board the usual urban legend bandwaggon..??..

http://forums.anandtech.com/showpost.php?p=37568426&postcount=1705

 

[cbn]

A suggestion I have for Carrizo (bundling Windows with bing on DIY boards):

http://forums.anandtech.com/showthread.php?t=2439271

This also to promote more uptake from small system integrators who could further differentiate from the large OEMs by using DDR3 2133 or DDR3 2400 RAM. (Same idea for the barebones Carrizo notebooks (sold by MSI and others)).

 

[Abwx]

How do you explain the fact that AMD did increase its market share with the cat family?

Because there was no contra revenues to render the product uncompetitive $ wise despite being better than the competition, like Mullins for instance..??.

 

[monstercameron]

Because there was no contra revenues to render the product uncompetitive $ wise despite being better than the competition, like Mullins for instance..??.

lets not derail with contrarev talk.

 

[Abwx]

lets not derail with contrarev talk.

You re right, let s get back to relevant discussions and numbers, sorry for being that easily fished....

Interesting numbers in the notes 1/9/10, who said that there s no numbers available, and so far i didnt heard that Intel did protest that AMD was making up numbers about Core M perfs...

http://www.notebookcheck.net/fileadmin/_processed_/csm_carrizo_footnotes_p2_1_84e6768f69.png

 

[cbn]

You re right, let s get back to relevant discussions and numbers, sorry for being that easily fished....

Interesting numbers in the notes 1/9/10, who said that there s no numbers available, and so far i didnt heard that Intel did protest that AMD was making up numbers about Core M perfs...

http://www.notebookcheck.net/fileadmin/_processed_/csm_carrizo_footnotes_p2_1_84e6768f69.png

AFAIK, no OEMs have used DDR3 past 1600 with Kaveri right?

 

[monstercameron]

@cbn I haven't seen anything higher that 1600 for kaveri.

@Abwx I can see how #4 on the endnotes is true because my fx7500 will eat ~40W! gaming. At first I thought it was win10 due to its beta state but it is definetely just kaveri. so that is a massive jump in efficiency.

The biggest change I hope carrizo brings is a more stable clock scaling mechanism and a better tdp balance between the cpu and gpu. Seeing the cpu jump all over the place in task manager is annoying. I doubt this will happen due to avfs and OEMs using cheaper power circuitry on AMD platforms...sigh.

 

[Arachnotronic]

Yeah as if they can compete with Intel in throwing money at OEMs and making a loss on chips. The best they could do is design something themselves and actually bring it to market.

You realize Intel makes a big profit on PC chips, so clearly they're not giving them away.

I don't know why it's so hard for some to face the reality that AMD APUs are just not very competitive right now.

 

[monstercameron]

You realize Intel makes a big profit on PC chips, so clearly they're not giving them away.

I don't know why it's so hard for some to face the reality that AMD APUs are just not very competitive right now.

nope, they are competitive.

 

[CHADBOGA]

nope, they are competitive.

Only in certain metrics that you concentrate on, but not the metrics you completely ignore.

 

[monstercameron]

Only in certain metrics that you concentrate on, but not the metrics you completely ignore.

exactly, and it goes both ways.

 

[CHADBOGA]

exactly, and it goes both ways.

The difference being the metrics you are concentrating on, just aren't as important to the intended market as the metrics you are ignoring.

That is why the Carrizo boosters are running up hard against reality.

 

[monstercameron]

The difference being the metrics you are concentrating on, just aren't as important to the intended market as the metrics you are ignoring.

That is why the Carrizo boosters are running up hard against reality.

and that is your opinion.