Intel Files Lawsuit Against Nvidia

Quote:

---

Originally Posted by 16floz470ml

I see that a number of people are saying "Why would you buy an Nvidia chipset when there is the X58." On the 775 lga the Nvidia 790i Ultra was/is the best 775 chipset. If they made a chipset that worked with Nehalem and was better I would buy it.

---

Maybe compared to the P965... Other than that no thanks. The 790i was just as terrible as the 780 IMO with it's super low clocks and pathetic raid support.

Quote:

---

Originally Posted by saaya

donnie, microsoft doesnt NEED a license key for windows and no application NEEDS a license key or serial number either, does that mean they should pay intel licensing fees for beeing able to use their plattform?

Anemone, how the h3ll is sli a monopoly?
could you please explain this in some more detail?


thats taking the trend of the past years and flipping it around 180degrees... could you please explain why the trend should change all the time?

whats driving pc upgrades and innovation? its almost entirely gaming and some internet and multi media. now whats better for this, slow cpu and fast gpu or fast cpu and slow gpu? then think again about which chip is going to be integrated into what... ;)

jensen is right about that part, and why do you think intel is going for gpus? just for fun? they know its more important than cpus and itll be even more important in the future as gpu processoring demand is growing way faster than cpu processing demand.

---

My point exactly! MS and Intel follows that same old agreement I mentioned earlier. They have a Mutual Agreement just like the ones with AMD, IBM, and everyone else not named nVidia and VIA.

Intel and MS also sit on many of the same standards boards and don't need side agreements, nVidia doesn't. As was mentioned, SLI is a monopoly because no 3rd party is allowed to build NF200, BIOS Cookies and etc.......

nVidia uses it (or tries to use it) as SLI a leverage when it should be an Open Standard. Just like some of the Stuff Intel has championed like PCI-E, AGP, USB and yada yada! Intel did give nVidia an open shot and nVidia didn't return the favor. As aways nVidia wants all incoming tech FOC and wants all THEIR out going tech paid for. That ain't going work without repercussions.

Quote:

---

Originally Posted by saaya

probably, but why would you do that?

---

you'd be able to upgrade v-ram! :rofl:

Quote:

---

Originally Posted by saaya

you want to cool several chips as good as you can, so you can clock them high and max them out performance wise. then what uber genius would come up with the idea to put it all on a 40x40mm package? :P

---

a understandable POV on XS, from someone involved in designing enthusiast motherboards :p:

but would you really care whether you buy one high end heatpipe cooler for your CPU and one for your GPU, or whether you buy one that costs twice as much for your GPU/CPU, if you'd get the same temperatures in the end?

Quote:

---

Originally Posted by saaya

thats what ive been saying for a long time, the whole point of FUSION doesnt add up... it makes sense for the mainstream and igp, but people always thought integrating gpus would mean better performance... well no...

both cpus and gpus are tdp limited, thats the whole point of overclocking, removing the tdp limit with better cooling and then beeing able to clock higher or bump voltages to clock even higher.

so we have two parts that are tdp limited, then how is putting them next to each other or merge them going to improve performance? :D

---

TDP has become outrageous on the latest GPUs, but at the same time the operating temperatures (for CPUs and GPUs) have expanded (up and down) and our heatsinks have become quite powerful. a tru120e can do nearly 200W, and it's not even that big - you could easily make it twice as deep and 50% taller and it'd handle 600W if cooling power scaled perfectly with size, 500W would definitely work. that'd cool a stock nehalem + GTX295!

back when CPUs didn't need heatsinks i'm sure it would have been hard to see today's coolers coming.. so i wouldn't say a future CPU/GPU combo would be impossible, or even inconvenient, due to thermals. maybe AMD will start selling heatpipe coolers again (i really wish intel would stick a small heatpipe in the middle of their stock coolers)

Quote:

---

Originally Posted by saaya

what benefits do you get from putting them on the same package or same piece of silicon?
more bandwidth...
was there a notable boost from pciE 1.1 to 2.0?
nope... then isnt it obvious that increasing the bandwidth between cpu and gpu isnt a limiting factor?

---

true, although i imagine the longer term processing core could be something like 50% x86, 50% GPU/CUDA (not those instruction sets, but... those functions compared to today's cores, and the manufacturer would make chips with varying ratios of CPU or GPU function to cater to different needs in the market,). or, to put it another way: a lot like larrabee ..
x86 with larrabee specific instructions..

Quote:

---

Originally Posted by saaya

the cpu memory bandwidth is a joke compared to gpus, so again, thats limiting, not speeding things up! unless you give them seperate memory interfaces which means loads of pins and defeats the purpose of putting them on the same package. or you let the cpu use the gpu mem, but while it has massive bandiwdth the latency is terrible, and the cpu would suffer and not benefit from that. so AGAIN, it doesnt make sense...

---

definitely technical challenges, much harder than solving the TDP problem i expect.

shared memory controller: it'd have to be very wide.
if the CPU was going to have to suffer slightly due to memory latency it could be compensated for by adding extra cache

Quote:

---

Originally Posted by saaya

the only good point is to save money if you put two mainstream or entry level parts on one package rather than two, or the same silicon die. and to everybody who thinks this is a revolution, amd geode gx cpus have had integrated gpus for a decade, and they are far from the only cpus with integrated gpus... its just uncommon outside of the ce segment, thats all...

---

the way i see it is... it's going to take time but it's not impossible. consider the i5 westmere MCM - here's a MCM a crazed billionaire could get made today, long before Intel or AMD or anyone else design and manufacture 'the ultimate gaming/multimedia core' with CPU/GPGPU/GPU logic all mixed in together:

GTX285 has just under 160GB/s theoretical memory bandwidth. triple channel DDR3-2000 has just under 50GB/s. DDR3 is lower latency than GDDR3, so let's say two triple-channel DDR3 memory controllers could feed a GTX285 pretty well, ~100GB/s bandwidth.

so -
32nm westmere quad w/o memory controller <-QPI-> 45nm x58gtx285 with 384bit MC
MCM on a single package, 6 DIMMs beside the socket
with a TRU120Ex2
delicious :ROTF:

^ of course that's not gonna happen.

[package]
phenom 2 with quad-channel DDR3-2000 for ~64GB/s
<-hypertransport 3->
HD4860 (4850 has 64GB/s, 4870 has 115GB/s)
[/package]
<-hypertransport 3->
790fx & sb750

hehe
[/dream]

Looks to me like an attempt to create a monopoly by intel. It would be bad.

SLI is an optional tech, while CPU with chipset support is a non replaceable-crucial yet universal part in modern computing. And thats already uncompetitive, and now they dont want other company to make product derivative components for their CPU.
Wonder what would happen if someday they able to cramp everything from graphic to usb interface into their cpu....

-tam2-

this