AMD TO RELEASE CHIP TO RVIAL NVIDA AND ATI - Graphic & Displays
  Tom's Guide Forums » Graphic & Displays » Nvidia » AMD TO RELEASE CHIP TO RVIAL NVIDA AND ATI
 




Word :   Username :  
 
Bottom
Author
 Thread : AMD TO RELEASE CHIP TO RVIAL NVIDA AND ATI
 
More Information

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

 

AMD enter the GPU market!

Now it might be said that where Intel lead, AMD follow, but if we're honest,
that's all just clever marketing and it has long been known that AMD run
faster and cooler than certain Intel chips. But GPUs were an area that Intel
did foray into and AMD wisely left well alone, probably believing it better
to focus on CPUs instead.

But suddenly, out of nowhere, AMD have delivered a hammer blow to ATI,
NVIDIA and Intel with a stunning new GPU that is set to light up the
accelerator scene in a fashion not seen since 3Dfx brought out the Voodoo.

The new chip, codenamed Apollo Fir, has SIXTEEN, yes that's right, SIXTEEN
pixel pipelines. Throw in just about every damn shading technique and
hardware feature you can think of and it would seem that on paper, AMD has
just won the entire GPU market with its first chip.

And no, this isn't just hyperbole, or me getting carried away with a press
release. Taking a cue from the XFX SLI system with two NVIDIA 6800 Ultra
512MB boards, AMD have produced an engineering board with a massive 2GB of
ultra-fast DDR2 memory.

HEXUS' Technical Editor, Rys, has this to say:
"I've just taken a look at a leaked engineering spec for the Fir and at
first glance it looks almost too good to be true, for a first graphics
processor from AMD. Supporting a 256-bit memory bus and being fully
compatible with WGF1.0 with its four quad fragment, and six vertex unit,
setup means the chip won't be small, even using AMD's 90nm SOI process at
their fabs in Dresden.

It's going to have roughly the transistor count of a pair of current 130nm
Athlon FX processors. It also looks like it can share vertex load with a
dual core processor, too, letting you use one of the cores entirely for
geometry processing, to go with the vertex units on the Fir core. It looks
like impressive stuff, with it hinting at future virtualisation of the
render core in a possible Apollo Fir 2006 refresh, to go with the WGF2.0
spec.

And 2GB of framebuffer will definitely come in handy for the upcoming
version of Playboy: The Mansion that uses masses of uncompressed normal maps
for even better breasticle fidelity."

At the time of writing, AMD are remaining tight-lipped about any further
specs and details such as core speeds were, unfortunately, not on our leaked
spec sheet. From what we've seen though its likely that 3DMark scores will
be somewhere up in the 30k range, putting the board way ahead of a Sun
system. By the way, the world record is roughly 35K, so the Apollo is only
5K off as an 'out of the box product' and those figures are on an
engineering sample board! All this power means that even high-end business
users are going to be grabbing the board on release.

When asked to comment, ATI's spokesperson, Louis Frappe-Mocha, had this to
say;
"We've long known about the Apollo Fir as one of the AMD engineers is
notoriously gobby after two pints and a curry, so we've been working on a
similar system based on 8 R480 chips. To save space we're stacking them two
high in four piles on each board. we think that we'll be able to push at
least 24 pipelines as soon as the damn thing stops blowing all the fuses in
the building when we plug it in."

In a similar vein, NVIDIA remain officially unconcerned, as their
spokeswoman, Sandy Beaches, stated:
"We've been working on a new version of the NV45, which effectively doubles
the triple filtering pixel overlay relay quad pass scan of each geometric
shape BEFORE the GPU even has the information needed. We've designed a way
to create small wormholes within the GPU and we pass information through
that way. Admittedly, there are stability issues and as soon as we get the
main part of Taiwan back in this dimension, we'll be going ahead with
production"

Intel's Hugh Jass had this to say about the AMD board:
"Look, stop bothering me, I KNOW I have a funny name, and oh yes, ha ha, it
is so very funny, isn't it? So now you've had your little giggle, why not go
and bother someone else?"

So there you have it, the Apollo Fir looks to be the board of choice with
no-one else even being able to get close until ATI can stop blowing
electricity substations and NVIDIA can get Taiwan back within the same
space/time envelope as us. Set for an April release you'd have to be a fool
to miss out on this one!



























APRIL FOOLS!

Related Product

Register or log in to remove.

More Information

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

 

"Anthony" <askme@aevansjr3@cox.net> wrote in message
news:R8f3e.11$Iq6.9@okepread01
> AMD enter the GPU market!
>
> Now it might be said that where Intel lead, AMD follow, but if we're
> honest, that's all just clever marketing and it has long been known
> that AMD run faster and cooler than certain Intel chips. But GPUs
> were an area that Intel did foray into and AMD wisely left well
> alone, probably believing it better to focus on CPUs instead.
>
> But suddenly, out of nowhere, AMD have delivered a hammer blow to ATI,
> NVIDIA and Intel with a stunning new GPU that is set to light up the
> accelerator scene in a fashion not seen since 3Dfx brought out the
> Voodoo.
> The new chip, codenamed Apollo Fir, has SIXTEEN, yes that's right,
> SIXTEEN pixel pipelines. Throw in just about every damn shading
> technique and hardware feature you can think of and it would seem
> that on paper, AMD has just won the entire GPU market with its first
> chip.
> And no, this isn't just hyperbole, or me getting carried away with a
> press release. Taking a cue from the XFX SLI system with two NVIDIA
> 6800 Ultra 512MB boards, AMD have produced an engineering board with
> a massive 2GB of ultra-fast DDR2 memory.
>
> HEXUS' Technical Editor, Rys, has this to say:
> "I've just taken a look at a leaked engineering spec for the Fir and
> at first glance it looks almost too good to be true, for a first
> graphics processor from AMD. Supporting a 256-bit memory bus and
> being fully compatible with WGF1.0 with its four quad fragment, and
> six vertex unit, setup means the chip won't be small, even using
> AMD's 90nm SOI process at their fabs in Dresden.
>
> It's going to have roughly the transistor count of a pair of current
> 130nm Athlon FX processors. It also looks like it can share vertex
> load with a dual core processor, too, letting you use one of the
> cores entirely for geometry processing, to go with the vertex units
> on the Fir core. It looks like impressive stuff, with it hinting at
> future virtualisation of the render core in a possible Apollo Fir
> 2006 refresh, to go with the WGF2.0 spec.
>
> And 2GB of framebuffer will definitely come in handy for the upcoming
> version of Playboy: The Mansion that uses masses of uncompressed
> normal maps for even better breasticle fidelity."
>
> At the time of writing, AMD are remaining tight-lipped about any
> further specs and details such as core speeds were, unfortunately,
> not on our leaked spec sheet. From what we've seen though its likely
> that 3DMark scores will be somewhere up in the 30k range, putting the
> board way ahead of a Sun system. By the way, the world record is
> roughly 35K, so the Apollo is only 5K off as an 'out of the box
> product' and those figures are on an engineering sample board! All
> this power means that even high-end business users are going to be
> grabbing the board on release.
> When asked to comment, ATI's spokesperson, Louis Frappe-Mocha, had
> this to say;
> "We've long known about the Apollo Fir as one of the AMD engineers is
> notoriously gobby after two pints and a curry, so we've been working
> on a similar system based on 8 R480 chips. To save space we're
> stacking them two high in four piles on each board. we think that
> we'll be able to push at least 24 pipelines as soon as the damn thing
> stops blowing all the fuses in the building when we plug it in."
>
> In a similar vein, NVIDIA remain officially unconcerned, as their
> spokeswoman, Sandy Beaches, stated:
> "We've been working on a new version of the NV45, which effectively
> doubles the triple filtering pixel overlay relay quad pass scan of
> each geometric shape BEFORE the GPU even has the information needed.
> We've designed a way to create small wormholes within the GPU and we
> pass information through that way. Admittedly, there are stability
> issues and as soon as we get the main part of Taiwan back in this
> dimension, we'll be going ahead with production"
>
> Intel's Hugh Jass had this to say about the AMD board:
> "Look, stop bothering me, I KNOW I have a funny name, and oh yes, ha
> ha, it is so very funny, isn't it? So now you've had your little
> giggle, why not go and bother someone else?"
>
> So there you have it, the Apollo Fir looks to be the board of choice
> with no-one else even being able to get close until ATI can stop
> blowing electricity substations and NVIDIA can get Taiwan back within
> the same space/time envelope as us. Set for an April release you'd
> have to be a fool to miss out on this one!
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> APRIL FOOLS!

Well said :) Fun read!
McG.

More Information

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

 

I hate April fools day.

DaveL


"McGrandpa" <McGrandpaNOT@NOThotmail.com> wrote in message
news:mwj3e.23677$1H3.16584@tornado.texas.rr.com...
> "Anthony" <askme@aevansjr3@cox.net> wrote in message
> news:R8f3e.11$Iq6.9@okepread01
>> AMD enter the GPU market!
>>
>> Now it might be said that where Intel lead, AMD follow, but if we're
>> honest, that's all just clever marketing and it has long been known
>> that AMD run faster and cooler than certain Intel chips. But GPUs
>> were an area that Intel did foray into and AMD wisely left well
>> alone, probably believing it better to focus on CPUs instead.
>>
>> But suddenly, out of nowhere, AMD have delivered a hammer blow to ATI,
>> NVIDIA and Intel with a stunning new GPU that is set to light up the
>> accelerator scene in a fashion not seen since 3Dfx brought out the
>> Voodoo.
>> The new chip, codenamed Apollo Fir, has SIXTEEN, yes that's right,
>> SIXTEEN pixel pipelines. Throw in just about every damn shading
>> technique and hardware feature you can think of and it would seem
>> that on paper, AMD has just won the entire GPU market with its first
>> chip.
>> And no, this isn't just hyperbole, or me getting carried away with a
>> press release. Taking a cue from the XFX SLI system with two NVIDIA
>> 6800 Ultra 512MB boards, AMD have produced an engineering board with
>> a massive 2GB of ultra-fast DDR2 memory.
>>
>> HEXUS' Technical Editor, Rys, has this to say:
>> "I've just taken a look at a leaked engineering spec for the Fir and
>> at first glance it looks almost too good to be true, for a first
>> graphics processor from AMD. Supporting a 256-bit memory bus and
>> being fully compatible with WGF1.0 with its four quad fragment, and
>> six vertex unit, setup means the chip won't be small, even using
>> AMD's 90nm SOI process at their fabs in Dresden.
>>
>> It's going to have roughly the transistor count of a pair of current
>> 130nm Athlon FX processors. It also looks like it can share vertex
>> load with a dual core processor, too, letting you use one of the
>> cores entirely for geometry processing, to go with the vertex units
>> on the Fir core. It looks like impressive stuff, with it hinting at
>> future virtualisation of the render core in a possible Apollo Fir
>> 2006 refresh, to go with the WGF2.0 spec.
>>
>> And 2GB of framebuffer will definitely come in handy for the upcoming
>> version of Playboy: The Mansion that uses masses of uncompressed
>> normal maps for even better breasticle fidelity."
>>
>> At the time of writing, AMD are remaining tight-lipped about any
>> further specs and details such as core speeds were, unfortunately,
>> not on our leaked spec sheet. From what we've seen though its likely
>> that 3DMark scores will be somewhere up in the 30k range, putting the
>> board way ahead of a Sun system. By the way, the world record is
>> roughly 35K, so the Apollo is only 5K off as an 'out of the box
>> product' and those figures are on an engineering sample board! All
>> this power means that even high-end business users are going to be
>> grabbing the board on release.
>> When asked to comment, ATI's spokesperson, Louis Frappe-Mocha, had
>> this to say;
>> "We've long known about the Apollo Fir as one of the AMD engineers is
>> notoriously gobby after two pints and a curry, so we've been working
>> on a similar system based on 8 R480 chips. To save space we're
>> stacking them two high in four piles on each board. we think that
>> we'll be able to push at least 24 pipelines as soon as the damn thing
>> stops blowing all the fuses in the building when we plug it in."
>>
>> In a similar vein, NVIDIA remain officially unconcerned, as their
>> spokeswoman, Sandy Beaches, stated:
>> "We've been working on a new version of the NV45, which effectively
>> doubles the triple filtering pixel overlay relay quad pass scan of
>> each geometric shape BEFORE the GPU even has the information needed.
>> We've designed a way to create small wormholes within the GPU and we
>> pass information through that way. Admittedly, there are stability
>> issues and as soon as we get the main part of Taiwan back in this
>> dimension, we'll be going ahead with production"
>>
>> Intel's Hugh Jass had this to say about the AMD board:
>> "Look, stop bothering me, I KNOW I have a funny name, and oh yes, ha
>> ha, it is so very funny, isn't it? So now you've had your little
>> giggle, why not go and bother someone else?"
>>
>> So there you have it, the Apollo Fir looks to be the board of choice
>> with no-one else even being able to get close until ATI can stop
>> blowing electricity substations and NVIDIA can get Taiwan back within
>> the same space/time envelope as us. Set for an April release you'd
>> have to be a fool to miss out on this one!
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> APRIL FOOLS!
>
> Well said :) Fun read!
> McG.
>

More Information

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

 

i pretty much guessed as much from the "beyond 3d" law suit

i'm slow but thankfully not that slow, plus i'm an offical amd reseller so i
think i would of kinda heard

www.verisys.8k.com

"Anthony" <askme@aevansjr3@cox.net> wrote in message
news:R8f3e.11$Iq6.9@okepread01...
> AMD enter the GPU market!
>
> Now it might be said that where Intel lead, AMD follow, but if we're
> honest, that's all just clever marketing and it has long been known that
> AMD run faster and cooler than certain Intel chips. But GPUs were an area
> that Intel did foray into and AMD wisely left well alone, probably
> believing it better to focus on CPUs instead.
>
> But suddenly, out of nowhere, AMD have delivered a hammer blow to ATI,
> NVIDIA and Intel with a stunning new GPU that is set to light up the
> accelerator scene in a fashion not seen since 3Dfx brought out the Voodoo.
>
> The new chip, codenamed Apollo Fir, has SIXTEEN, yes that's right, SIXTEEN
> pixel pipelines. Throw in just about every damn shading technique and
> hardware feature you can think of and it would seem that on paper, AMD has
> just won the entire GPU market with its first chip.
>
> And no, this isn't just hyperbole, or me getting carried away with a press
> release. Taking a cue from the XFX SLI system with two NVIDIA 6800 Ultra
> 512MB boards, AMD have produced an engineering board with a massive 2GB of
> ultra-fast DDR2 memory.
>
> HEXUS' Technical Editor, Rys, has this to say:
> "I've just taken a look at a leaked engineering spec for the Fir and at
> first glance it looks almost too good to be true, for a first graphics
> processor from AMD. Supporting a 256-bit memory bus and being fully
> compatible with WGF1.0 with its four quad fragment, and six vertex unit,
> setup means the chip won't be small, even using AMD's 90nm SOI process at
> their fabs in Dresden.
>
> It's going to have roughly the transistor count of a pair of current 130nm
> Athlon FX processors. It also looks like it can share vertex load with a
> dual core processor, too, letting you use one of the cores entirely for
> geometry processing, to go with the vertex units on the Fir core. It looks
> like impressive stuff, with it hinting at future virtualisation of the
> render core in a possible Apollo Fir 2006 refresh, to go with the WGF2.0
> spec.
>
> And 2GB of framebuffer will definitely come in handy for the upcoming
> version of Playboy: The Mansion that uses masses of uncompressed normal
> maps for even better breasticle fidelity."
>
> At the time of writing, AMD are remaining tight-lipped about any further
> specs and details such as core speeds were, unfortunately, not on our
> leaked spec sheet. From what we've seen though its likely that 3DMark
> scores will be somewhere up in the 30k range, putting the board way ahead
> of a Sun system. By the way, the world record is roughly 35K, so the
> Apollo is only 5K off as an 'out of the box product' and those figures are
> on an engineering sample board! All this power means that even high-end
> business users are going to be grabbing the board on release.
>
> When asked to comment, ATI's spokesperson, Louis Frappe-Mocha, had this to
> say;
> "We've long known about the Apollo Fir as one of the AMD engineers is
> notoriously gobby after two pints and a curry, so we've been working on a
> similar system based on 8 R480 chips. To save space we're stacking them
> two high in four piles on each board. we think that we'll be able to push
> at least 24 pipelines as soon as the damn thing stops blowing all the
> fuses in the building when we plug it in."
>
> In a similar vein, NVIDIA remain officially unconcerned, as their
> spokeswoman, Sandy Beaches, stated:
> "We've been working on a new version of the NV45, which effectively
> doubles the triple filtering pixel overlay relay quad pass scan of each
> geometric shape BEFORE the GPU even has the information needed. We've
> designed a way to create small wormholes within the GPU and we pass
> information through that way. Admittedly, there are stability issues and
> as soon as we get the main part of Taiwan back in this dimension, we'll be
> going ahead with production"
>
> Intel's Hugh Jass had this to say about the AMD board:
> "Look, stop bothering me, I KNOW I have a funny name, and oh yes, ha ha,
> it is so very funny, isn't it? So now you've had your little giggle, why
> not go and bother someone else?"
>
> So there you have it, the Apollo Fir looks to be the board of choice with
> no-one else even being able to get close until ATI can stop blowing
> electricity substations and NVIDIA can get Taiwan back within the same
> space/time envelope as us. Set for an April release you'd have to be a
> fool to miss out on this one!
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> APRIL FOOLS!
>

More Information

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

 

"Anthony" <askme@aevansjr3@cox.net> wrote in message
news:R8f3e.11$Iq6.9@okepread01...
> AMD enter the GPU market!
>
> Now it might be said that where Intel lead, AMD follow, but if we're
> honest, that's all just clever marketing and it has long been known that
> AMD run faster and cooler than certain Intel chips. But GPUs were an area
> that Intel did foray into and AMD wisely left well alone, probably
> believing it better to focus on CPUs instead.
>
> But suddenly, out of nowhere, AMD have delivered a hammer blow to ATI,
> NVIDIA and Intel with a stunning new GPU that is set to light up the
> accelerator scene in a fashion not seen since 3Dfx brought out the Voodoo.
>
> The new chip, codenamed Apollo Fir, has SIXTEEN, yes that's right, SIXTEEN

Apollo Fir, right. No need to read further.

You had me going there for a minute, though. I guess I'm up too late.

N.

More Information

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

 

I just subscribed to this group....what a hoot...loved the
joke....when I saw the names "Louis Frappe-Mocha" and "Sandy Beaches",
I just about hit the floor laughing...And when I saw "Hugh Jass", I
did hit the floor in a fit of laughing, I was crying....

Good job to the origianl author....


On Sat, 02 Apr 2005 00:16:25 GMT, "Veritech"
<avis.dalrymple@ntlworld.com> wrote:

>i pretty much guessed as much from the "beyond 3d" law suit
>
>i'm slow but thankfully not that slow, plus i'm an offical amd reseller so i
>think i would of kinda heard
>
>www.verisys.8k.com
>
>"Anthony" <askme@aevansjr3@cox.net> wrote in message
>news:R8f3e.11$Iq6.9@okepread01...
>> AMD enter the GPU market!
>>
>> Now it might be said that where Intel lead, AMD follow, but if we're
>> honest, that's all just clever marketing and it has long been known that
>> AMD run faster and cooler than certain Intel chips. But GPUs were an area
>> that Intel did foray into and AMD wisely left well alone, probably
>> believing it better to focus on CPUs instead.
>>
>> But suddenly, out of nowhere, AMD have delivered a hammer blow to ATI,
>> NVIDIA and Intel with a stunning new GPU that is set to light up the
>> accelerator scene in a fashion not seen since 3Dfx brought out the Voodoo.
>>
>> The new chip, codenamed Apollo Fir, has SIXTEEN, yes that's right, SIXTEEN
>> pixel pipelines. Throw in just about every damn shading technique and
>> hardware feature you can think of and it would seem that on paper, AMD has
>> just won the entire GPU market with its first chip.
>>
>> And no, this isn't just hyperbole, or me getting carried away with a press
>> release. Taking a cue from the XFX SLI system with two NVIDIA 6800 Ultra
>> 512MB boards, AMD have produced an engineering board with a massive 2GB of
>> ultra-fast DDR2 memory.
>>
>> HEXUS' Technical Editor, Rys, has this to say:
>> "I've just taken a look at a leaked engineering spec for the Fir and at
>> first glance it looks almost too good to be true, for a first graphics
>> processor from AMD. Supporting a 256-bit memory bus and being fully
>> compatible with WGF1.0 with its four quad fragment, and six vertex unit,
>> setup means the chip won't be small, even using AMD's 90nm SOI process at
>> their fabs in Dresden.
>>
>> It's going to have roughly the transistor count of a pair of current 130nm
>> Athlon FX processors. It also looks like it can share vertex load with a
>> dual core processor, too, letting you use one of the cores entirely for
>> geometry processing, to go with the vertex units on the Fir core. It looks
>> like impressive stuff, with it hinting at future virtualisation of the
>> render core in a possible Apollo Fir 2006 refresh, to go with the WGF2.0
>> spec.
>>
>> And 2GB of framebuffer will definitely come in handy for the upcoming
>> version of Playboy: The Mansion that uses masses of uncompressed normal
>> maps for even better breasticle fidelity."
>>
>> At the time of writing, AMD are remaining tight-lipped about any further
>> specs and details such as core speeds were, unfortunately, not on our
>> leaked spec sheet. From what we've seen though its likely that 3DMark
>> scores will be somewhere up in the 30k range, putting the board way ahead
>> of a Sun system. By the way, the world record is roughly 35K, so the
>> Apollo is only 5K off as an 'out of the box product' and those figures are
>> on an engineering sample board! All this power means that even high-end
>> business users are going to be grabbing the board on release.
>>
>> When asked to comment, ATI's spokesperson, Louis Frappe-Mocha, had this to
>> say;
>> "We've long known about the Apollo Fir as one of the AMD engineers is
>> notoriously gobby after two pints and a curry, so we've been working on a
>> similar system based on 8 R480 chips. To save space we're stacking them
>> two high in four piles on each board. we think that we'll be able to push
>> at least 24 pipelines as soon as the damn thing stops blowing all the
>> fuses in the building when we plug it in."
>>
>> In a similar vein, NVIDIA remain officially unconcerned, as their
>> spokeswoman, Sandy Beaches, stated:
>> "We've been working on a new version of the NV45, which effectively
>> doubles the triple filtering pixel overlay relay quad pass scan of each
>> geometric shape BEFORE the GPU even has the information needed. We've
>> designed a way to create small wormholes within the GPU and we pass
>> information through that way. Admittedly, there are stability issues and
>> as soon as we get the main part of Taiwan back in this dimension, we'll be
>> going ahead with production"
>>
>> Intel's Hugh Jass had this to say about the AMD board:
>> "Look, stop bothering me, I KNOW I have a funny name, and oh yes, ha ha,
>> it is so very funny, isn't it? So now you've had your little giggle, why
>> not go and bother someone else?"
>>
>> So there you have it, the Apollo Fir looks to be the board of choice with
>> no-one else even being able to get close until ATI can stop blowing
>> electricity substations and NVIDIA can get Taiwan back within the same
>> space/time envelope as us. Set for an April release you'd have to be a
>> fool to miss out on this one!
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> APRIL FOOLS!
>>
>


  Tom's Guide Forums » Graphic & Displays » Nvidia » AMD TO RELEASE CHIP TO RVIAL NVIDA AND ATI

Go to:
 

Google ads