Next gen AMD mobile part - Griffen

[Hulk]

According to Anand Griffen will be heavily based on K8. I'm not really getting how this strategy will help AMD in the mobile market. The current K8 core is getting pasted by C2D today clock-for-clock. And from what I just read I don't think the improvements made to create Griffen will enable to catch up to C2D IPC-wise.

Add to that the fact that by the Griffen release time Penryn will be widely available at 45nm. More power efficient than current mobile C2D offerings, faster, and with better IPC. Even if Griffen can somehow match Penryn IPC will it really be able to scale with Penryn. It just doesn't seem like they can make K8 compete with C2D. If they could then there wouldn't be Barcelona right?

Like I said, I don't get it? I'm sure I'm not reading this correctly.

 

[myocardia]

Yeah, it's very obvious that buying ATI was the worst mistake in AMD's history. Now, they don't even have the money for R&D for their primary business venture. Of course, it won't be the first company to go down the drain because of one arrogant CEO.:Q

 

[Zstream]

Originally posted by: Hulk
According to Anand Griffen will be heavily based on K8. I'm not really getting how this strategy will help AMD in the mobile market. The current K8 core is getting pasted by C2D today clock-for-clock. And from what I just read I don't think the improvements made to create Griffen will enable to catch up to C2D IPC-wise.

Add to that the fact that by the Griffen release time Penryn will be widely available at 45nm. More power efficient than current mobile C2D offerings, faster, and with better IPC. Even if Griffen can somehow match Penryn IPC will it really be able to scale with Penryn. It just doesn't seem like they can make K8 compete with C2D. If they could then there wouldn't be Barcelona right?

Like I said, I don't get it? I'm sure I'm not reading this correctly.

You are incorrect on many things, so let me start with the first.

95% of people do not buy a mobile pc to play games on, this number came from vendors such as MSI/Asus/Acer etc....

Second, most people including myself will be happy with even a 3800+ dual core that can last for say 5-6 hours on battery. This is plenty of power for what I do on a laptop.

Third, if K8 can come out of a fab with even better numbers then yes they will have a hit.

 

[MarcVenice]

http://realworldtech.com/page.cfm?ArticleID=RWT051607033728

read it, I have't read it completely but it starts of telling how Barcelona is also based quite a bit on K8, and why it is that way, very enlightening. And I don't know how much of an engineer you are, but I doubt anyone not on AMD's development team could tell into how much speedgain the current improvements will translate.

 

[Hulk]

Originally posted by: Zstream

Originally posted by: Hulk
According to Anand Griffen will be heavily based on K8. I'm not really getting how this strategy will help AMD in the mobile market. The current K8 core is getting pasted by C2D today clock-for-clock. And from what I just read I don't think the improvements made to create Griffen will enable to catch up to C2D IPC-wise.

Add to that the fact that by the Griffen release time Penryn will be widely available at 45nm. More power efficient than current mobile C2D offerings, faster, and with better IPC. Even if Griffen can somehow match Penryn IPC will it really be able to scale with Penryn. It just doesn't seem like they can make K8 compete with C2D. If they could then there wouldn't be Barcelona right?

Like I said, I don't get it? I'm sure I'm not reading this correctly.

You are incorrect on many things, so let me start with the first.

95% of people do not buy a mobile pc to play games on, this number came from vendors such as MSI/Asus/Acer etc....

Second, most people including myself will be happy with even a 3800+ dual core that can last for say 5-6 hours on battery. This is plenty of power for what I do on a laptop.

Third, if K8 can come out of a fab with even better numbers then yes they will have a hit.

I don't want to get into a debate based on speculation but I never mentioned gaming? I don't game on my laptop but I still want a fast/efficient processor. If most people are like me then I think they have just about all of programs on their lappy that they have on their desktop and only expect a slight performance hit on the laptop.

If you can choose system "A" or system "B" that will both run 5 to 6 hours but system "A" is faster then which will you choose

Just based on AMD being a generation behind in process technology I think it will be very, very difficult for them to beat Intel in the performance/watt category. Not impossible of course but very difficult.

I just hope AMD isn't shooting too low with Griffen. I'm worried because the release is over a year away and that gives Intel a lot of time to pull further ahead.

 

[TanisHalfElven]

Originally posted by: Hulk

Originally posted by: Zstream

Originally posted by: Hulk
According to Anand Griffen will be heavily based on K8. I'm not really getting how this strategy will help AMD in the mobile market. The current K8 core is getting pasted by C2D today clock-for-clock. And from what I just read I don't think the improvements made to create Griffen will enable to catch up to C2D IPC-wise.

Add to that the fact that by the Griffen release time Penryn will be widely available at 45nm. More power efficient than current mobile C2D offerings, faster, and with better IPC. Even if Griffen can somehow match Penryn IPC will it really be able to scale with Penryn. It just doesn't seem like they can make K8 compete with C2D. If they could then there wouldn't be Barcelona right?

Like I said, I don't get it? I'm sure I'm not reading this correctly.

You are incorrect on many things, so let me start with the first.

95% of people do not buy a mobile pc to play games on, this number came from vendors such as MSI/Asus/Acer etc....

Second, most people including myself will be happy with even a 3800+ dual core that can last for say 5-6 hours on battery. This is plenty of power for what I do on a laptop.

Third, if K8 can come out of a fab with even better numbers then yes they will have a hit.

I don't want to get into a debate based on speculation but I never mentioned gaming? I don't game on my laptop but I still want a fast/efficient processor. If most people are like me then I think they have just about all of programs on their lappy that they have on their desktop and only expect a slight performance hit on the laptop.

If you can choose system "A" or system "B" that will both run 5 to 6 hours but system "A" is faster then which will you choose

Just based on AMD being a generation behind in process technology I think it will be very, very difficult for them to beat Intel in the performance/watt category. Not impossible of course but very difficult.

I just hope AMD isn't shooting too low with Griffen. I'm worried because the release is over a year away and that gives Intel a lot of time to pull further ahead.

i would choose the cheaper one. not the faster one. hell my current laptop is a p3 900. works great for me.

 

[Finnkc]

Originally posted by: Hulk

Originally posted by: Zstream

Originally posted by: Hulk
According to Anand Griffen will be heavily based on K8. I'm not really getting how this strategy will help AMD in the mobile market. The current K8 core is getting pasted by C2D today clock-for-clock. And from what I just read I don't think the improvements made to create Griffen will enable to catch up to C2D IPC-wise.

Add to that the fact that by the Griffen release time Penryn will be widely available at 45nm. More power efficient than current mobile C2D offerings, faster, and with better IPC. Even if Griffen can somehow match Penryn IPC will it really be able to scale with Penryn. It just doesn't seem like they can make K8 compete with C2D. If they could then there wouldn't be Barcelona right?

Like I said, I don't get it? I'm sure I'm not reading this correctly.

You are incorrect on many things, so let me start with the first.

95% of people do not buy a mobile pc to play games on, this number came from vendors such as MSI/Asus/Acer etc....

Second, most people including myself will be happy with even a 3800+ dual core that can last for say 5-6 hours on battery. This is plenty of power for what I do on a laptop.

Third, if K8 can come out of a fab with even better numbers then yes they will have a hit.

I don't want to get into a debate based on speculation but I never mentioned gaming? I don't game on my laptop but I still want a fast/efficient processor. If most people are like me then I think they have just about all of programs on their lappy that they have on their desktop and only expect a slight performance hit on the laptop.

If you can choose system "A" or system "B" that will both run 5 to 6 hours but system "A" is faster then which will you choose

Just based on AMD being a generation behind in process technology I think it will be very, very difficult for them to beat Intel in the performance/watt category. Not impossible of course but very difficult.

I just hope AMD isn't shooting too low with Griffen. I'm worried because the release is over a year away and that gives Intel a lot of time to pull further ahead.

Thats if sys A and B are the same price ... I am not paying 400 bucks for a few more benchmark points. For a laptop atleast.

 

[tvdang7]

well they said the new platform has support for the new HT so when the new amd cpu comes out they will prob just drop it in like intel didwith coreduo and coreduo 2

 

[CurseTheSky]

I have a laptop sitting next to me with a Mobile Athlon 64 3200+, Windows Xp, 512MB of RAM, and a 100GB 5400RPM hard drive...

The thing is a champ. It slows down when I'm multi-tasking (mainly because of the memory), but it can handle just about anything I throw at it. If I could stick something like an X2 3800+ into it, I'd probably keep it for another two years.

 

[SlowSpyder]

Our laptop at home is a PM @ 1.6ghz, everyday I work on Core 2 Duo laptops. You can't tell the difference running business apps in XP. MS Office installs and runs the same on both. Most people buy a laptop based on it's mobility features, not e-penis CPU processing power.

 

[f4phantom2500]

why would you want to pay more for a faster intel based laptop when you can get a much cheaper amd laptop that does everything you want to do on a laptop?

 

[Kur]

If it's not broke, don't fix it.

 

[Hulk]

Originally posted by: f4phantom2500
why would you want to pay more for a faster intel based laptop when you can get a much cheaper amd laptop that does everything you want to do on a laptop?

Yeah okay. Do you realize that the 1.66 to 2Ghz C2D mobile parts are really cheap now?

You can get a C2D Dell lappy for llike $600 these days.

Hey if you're happy with the AMD laptop more power to you. I like a laptop with some power. My fully loaded Dell 640m with 2GHz C2D was $900 shipped.

Like I initially said I'm worried that Griffin will be too little too late for AMD.

But I really hope you guys are right and I'm wrong. The way Intel is storming the market right now makes me wonder though. There were a lot of guys poo pooing Intel's preliminary benches for C2D this time last year. And now look what happened. AMD is all but heading down the toilet...

 

[formulav8]

AMD isn't going anymore anytime soon. They have absolutely more than $2 Billion dollars at their disposal for the next 2 years. I don't know why people keep saying their dead and gone anyday now.

And their new Mobile should very well be a hit. Most people buy a lappy based on mobility, features, and battery life. CPU performance is down the chain. Besides, the performance of the new mobile is going to increase over the K8 anyways. Not a barc increase, but it will be higher performing AND be lower wattage. Makes for a perfect mobile cpu. I don't see what the problem with this is?

Just MO.


Jason

 

[Accord99]

Originally posted by: formulav8
And their new Mobile should very well be a hit. Most people buy a lappy based on mobility, features, and battery life. CPU performance is down the chain. Besides, the performance of the new mobile is going to increase over the K8 anyways. Not a barc increase, but it will be higher performing AND be lower wattage. Makes for a perfect mobile cpu. I don't see what the problem with this is?

The problem would be Intel's mobile CPUs remaining faster and lower wattage.

 

[JackPack]

Originally posted by: http://www.reghardware.co.uk/2007/05/18/amd_peeks_into_griffin/
AMD Fellow Maurice Steinman told Register Hardware the CPU cores are almost transistor-for-transistor the same cores found in the 65nm Turion processors AMD launched earlier this month.

So bascially, it's the same as the current Turions? Zero or negligible performance increase.

TDP for Griffin remains 35W for dual-core and it's a 65nm device.

Since this comes out in mid-2008, it's against Intel's mature 45nm Penryn products and likely Nehalem as well.

 

[JackPack]

Originally posted by: formulav8
And their new Mobile should very well be a hit. Most people buy a lappy based on mobility, features, and battery life. CPU performance is down the chain.

If that were true, Transmeta-based notebooks would be ruling the roost today.

Consumers care about battery life, but not enough to sacrifice too much performance. Griffin has essentially zero IPC improvements. When it comes out in mid-2008, it'll offer early-2006 performance.

The problem with battery life is that the processor isn't the major consumer of power -- it's the display. AMD needs to work with notebook ODMs the way Intel is to improve battery life. People simply aren't willing to give up too much performance for battery life.

 

[Hulk]

Originally posted by: JackPack

Originally posted by: formulav8
And their new Mobile should very well be a hit. Most people buy a lappy based on mobility, features, and battery life. CPU performance is down the chain.

If that were true, Transmeta-based notebooks would be ruling the roost today.

Consumers care about battery life, but not enough to sacrifice too much performance. Griffin has essentially zero IPC improvements. When it comes out in mid-2008, it'll offer early-2006 performance.

The problem with battery life is that the processor isn't the major consumer of power -- it's the display. AMD needs to work with notebook ODMs the way Intel is to improve battery life. People simply aren't willing to give up too much performance for battery life.

Exactly my point. Thank you.

 

[Viditor]

Originally posted by: JackPack

Originally posted by: http://www.reghardware.co.uk/2007/05/18/amd_peeks_into_griffin/
AMD Fellow Maurice Steinman told Register Hardware the CPU cores are almost transistor-for-transistor the same cores found in the 65nm Turion processors AMD launched earlier this month.

So bascially, it's the same as the current Turions? Zero or negligible performance increase.

TDP for Griffin remains 35W for dual-core and it's a 65nm device.

Since this comes out in mid-2008, it's against Intel's mature 45nm Penryn products and likely Nehalem as well.

The first Penryn notebooks will start coming out at the same time (mid 2008)...ODMs always do their designs in January for a mid-year release (back to school). Nehalem won't even start until either mid 2009 or mid 2010 for notebooks.
If you look at the platform as a whole, the CPU is no longer the most power hungry part...memory is taking over that honour.
That said, HT uses far less power than a FSB, and it's cheaper to design around and manufacture.

The 4 key metrics for a purchase decision on laptops (in order of importance):

1. Price...easily the most important for most everyone
2. Features...important for many, but not all. The key is that the notebook has the right features for the customers needs (not necessarily their wants).
3. Battery Life...also important to many but not all customers
4. Performance...literally the last thing that 95% of customers care about, as long as they can meet their usage needs.

 

[Viditor]

Originally posted by: JackPack

Originally posted by: formulav8
And their new Mobile should very well be a hit. Most people buy a lappy based on mobility, features, and battery life. CPU performance is down the chain.

If that were true, Transmeta-based notebooks would be ruling the roost today.

Consumers care about battery life, but not enough to sacrifice too much performance. Griffin has essentially zero IPC improvements. When it comes out in mid-2008, it'll offer early-2006 performance.

The problem with battery life is that the processor isn't the major consumer of power -- it's the display. AMD needs to work with notebook ODMs the way Intel is to improve battery life. People simply aren't willing to give up too much performance for battery life.

Transmeta sold every chip they could make for laptops! The problem was that they couldn't make a profit on them because of COGS, and that they didn't meet the minimum standard for modern software in many cases (Turions are nowhere near this category).

 

[Regs]

I don't care if people don't play games on a mobile unit. Performance is king, and performance speaks for itself. People buy the better performing product and I bet you more than half the time a person would rather buy the better performing CPU playing 3D games than the other product that saves you trivial time of battery life.

Hmm brand new Core Duo that saves me battery life over my old Intel and performs tremendously better....or.... AMD's K8 which offers...wait what?

My worst fears are becoming a reality at AMD. They simply do not want to compete in performance anymore. Whatever their definition of "innovation" is, it isn't the ones we share. Performance sells and the K8 proved that. Performance starts the buzz and ripples down to the customers. Did Hector and his lackeys think that their "true" dual core or power saving features made the sales? Do you think people know what the hell a dedicated power plane is? Because over the past 4-5 years that's the only thing I've been seeing; improvements in niche markets that only target the minorities. It's "the safe bet" moves like this from AMD that makes me worry what they plan to do with Barcelona against Intel's onslaught.

Ah hum bug im going to bed.

 

[jhtrico1850]

"Nehalem won't even start until either mid 2009 or mid 2010 for notebooks."

A whole 1-2 years after launch?

"If you look at the platform as a whole, the CPU is no longer the most power hungry part...memory is taking over that honour."

With that logic, DDR3 with Intel should be smooth sailing ahead. How do you come to this? Perhaps when idle, but definitely not under load. And if the notebook is idle, it might as well be off.

"That said, HT uses far less power than a FSB, and it's cheaper to design around and manufacture."

Do you have numbers?

"1. Price...easily the most important for most everyone
2. Features...important for many, but not all. The key is that the notebook has the right features for the customers needs (not necessarily their wants).
3. Battery Life...also important to many but not all customers
4. Performance...literally the last thing that 95% of customers care about, as long as they can meet their usage needs."

Agreed.

 

[BigDH01]

Originally posted by: formulav8
AMD isn't going anymore anytime soon. They have absolutely more than $2 Billion dollars at their disposal for the next 2 years. I don't know why people keep saying their dead and gone anyday now.

And their new Mobile should very well be a hit. Most people buy a lappy based on mobility, features, and battery life. CPU performance is down the chain. Besides, the performance of the new mobile is going to increase over the K8 anyways. Not a barc increase, but it will be higher performing AND be lower wattage. Makes for a perfect mobile cpu. I don't see what the problem with this is?

Just MO.


Jason

Two billion is nice and absolutely needed when losing about 500m a quarter. However, it doesn't do much to pay off their massive debt. AMD is in trouble right now and mostly because of the ATI purchase. And, of course, their inability to execute on product launches. We'd better hope they get Barcelona right.

 

[Zstream]

Originally posted by: Regs
I don't care if people don't play games on a mobile unit. Performance is king, and performance speaks for itself. People buy the better performing product and I bet you more than half the time a person would rather buy the better performing CPU playing 3D games than the other product that saves you trivial time of battery life.

Hmm brand new Core Duo that saves me battery life over my old Intel and performs tremendously better....or.... AMD's K8 which offers...wait what?

My worst fears are becoming a reality at AMD. They simply do not want to compete in performance anymore. Whatever their definition of "innovation" is, it isn't the ones we share. Performance sells and the K8 proved that. Performance starts the buzz and ripples down to the customers. Did Hector and his lackeys think that their "true" dual core or power saving features made the sales? Do you think people know what the hell a dedicated power plane is? Because over the past 4-5 years that's the only thing I've been seeing; improvements in niche markets that only target the minorities. It's "the safe bet" moves like this from AMD that makes me worry what they plan to do with Barcelona against Intel's onslaught.

Ah hum bug im going to bed.

You are silly, the AMD will have voltage regulations of each core this makes for a longer battery life. So, with that being said the people who buy laptops do not buy for performance. They buy for battery life.

If you are at a certain performance ratio which we are at right now buying a better CPU means nothing. More ram is often a better buy then a cpu. Then to top it off you have to buy a GPU which cost $$$$$.

"That said, HT uses far less power than a FSB, and it's cheaper to design around and manufacture."

It is on the AMD website. That is why Intel is copying it and putting it on the new cpu's.

 

[Viditor]

Originally posted by: jhtrico1850
"Nehalem won't even start until either mid 2009 or mid 2010 for notebooks."

A whole 1-2 years after launch?

Launch is expected to be the end of 2008. ODM designs are done in January for release 6 months later...in other words, if you don't have the supply locked up by January, you'll wait another year and a half for the product to hit the market. So in effect it will be 6-18 months after launch for volume sales...

"If you look at the platform as a whole, the CPU is no longer the most power hungry part...memory is taking over that honour."

With that logic, DDR3 with Intel should be smooth sailing ahead. How do you come to this? Perhaps when idle, but definitely not under load. And if the notebook is idle, it might as well be off.

Notebooks spend most of their time at idle or very low usage...in fact the percentage of time spent under full load for the vast majority of people is in the low single digits.
DDR3 will indeed help Intel's power profile, as it will AMD's. I expect that Intel will have a good 6 month lead on AMD for DDR3 usage, but remember that notebook designs are very cyclical.

"That said, HT uses far less power than a FSB, and it's cheaper to design around and manufacture."

Do you have numbers?

Not handy (sorry), but I'm sure if you find the spec sheets you'll be able to determine this for yourself...

"1. Price...easily the most important for most everyone
2. Features...important for many, but not all. The key is that the notebook has the right features for the customers needs (not necessarily their wants).
3. Battery Life...also important to many but not all customers
4. Performance...literally the last thing that 95% of customers care about, as long as they can meet their usage needs."

Agreed.