What the Falanx is that? Mail200

Sounds hawt. Hoepfully its not fluff and they can get the funding they need.
 
This is certainly good news. Hopefully they will not go the way of BitBoys :p More competition is always good, but I wonder how anyone can get an integrated solution that rivals its add-in peers....must be one huge MB!
 
Gimme some game benchies and I'll have something to talk about on here. But it's exactly what Dell needs.

~Adam
 
ScourggeFX said:
This is certainly good news. Hopefully they will not go the way of BitBoys :p More competition is always good, but I wonder how anyone can get an integrated solution that rivals its add-in peers....must be one huge MB!
Click to expand...

Check the slides for unit size. TINY.
 
Here is a forum-only tidbit. I don't remember if it was in the slides or not.

Mali200 TDP = .5milliwat per MHz.
 
Yes but what stages of add-in graphics cards will we be on then (obviously I'm looking at this as desktop which isn't entirely fair but so what? :p). What sounds promising now may be old news when it is finally released.
 
tornadotsunamilife said:
Yes but what stages of add-in graphics cards will we be on then (obviously I'm looking at this as desktop which isn't entirely fair but so what? :p). What sounds promising now may be old news when it is finally released.
Click to expand...

I agree. With many great product announcements, only a few make it to market before they are obsolete and not so great anymore. Although I don't like a rushed product, I do like to see it released before it's usefulness is gone.

Does look like a nice product. OEM's will love it, for sure.
 
Menelmarar said:
you can't be serious? by my calculation thats about a metric fuckton of Watts.

edit, 0.5mW * 100MHz = 15kW !!! :eek: :confused:
Click to expand...

Isn't it 50Watts?

Still that seems high - though the actual grunt per clock may make mhz less necessary... time will tell!
 
lol @ math above.

0.5mW/MHz = 0.0005W/MHz

0.0005W/MHz * 100MHz = 0.05W, something not very likely unless ".5millwatt/MHz" is an idle figure.
 
pxc said:
lol @ math above.

0.5mW/MHz = 0.0005W/MHz

0.0005W/MHz * 100MHz = 0.05W, something not very likely unless ".5millwatt/MHz" is an idle figure.
Click to expand...


I am unsure as to how many units you would need to reproduce "on par" performance with today's GPUs though. Looking at the die size, 50 to 80 units would not be a stretch of the imagination...

Disclaimer: I am just thinking out loud on this and have no information as to the validity of this post. So don't repeat it and look like a jackass. ;)
 
So lemme get this straight , both Ati and Nvidia are blowing tons of wad on developing gfx cores and even a company the size of Via couldn't produce a competitive video card not to mention the others. Then a company comes along ,pulls some crap out of a hat and the investors get raped as usual. Call me when they have working samples that annihilates both ATi and Nvidia offerings for the same price , overpriced onboard chipset don't count since no manufacturer would support it and DX9 functionality don't count either see fx5200 etc . My 2 cents.
 
mashie said:
It is 0.05 Watts at 100MHz or 0.5 Watts at 1 GHz.

Edit: I type slow....
Click to expand...
yea im really dumb, im going on my 11th hour at work today.

Still, since my mind can't seem to grasp simple units. I don't know how I got 1*5 is 15 at the front of my calculation nor do I know why I calculated in mW but looked at W at the result. Still....if I can pull every bit of logic out of my brain.

0.0005W * 100,000,000Hz = 50,000W seems logical right?

multiplying by the 0.0005 should take 4 zero's out of 100MHz, leaving you with 10kHz, then times 5 is 50kW.
 
don't even bother calling me a dumbass, I know I am....Kyle said 0.5mW / Mhz ...not per hz....ow :( I'm loosing it.
 
Menelmarar said:
yea im really dumb, im going on my 11th hour at work today.

Still, since my mind can't seem to grasp simple units. I don't know how I got 1*5 is 15 at the front of my calculation nor do I know why I calculated in mW but looked at W at the result. Still....if I can pull every bit of logic out of my brain.

0.0005W * 100,000,000Hz = 50,000W seems logical right?

multiplying by the 0.0005 should take 4 zero's out of 100MHz, leaving you with 10kHz, then times 5 is 50kW.
Click to expand...

Kyle said 0.0005W per MHz which is equivalent to 5x10^-10 W per Hz.

5x10^-10W/Hz*100,000,000 Hz = 0.05 W
 
Taken from their website:

Falanx was incorporated in April 2001 and is a privately held corporation developing and marketing IP Graphics Cores for the mobile market for use in mobile phones, PDAs, set-top boxes, handheld gaming devices and infotainment systems. Falanx is licensing its family of IP Cores directly to equipment vendors and System-on-a-Chip (SoC) vendors. The Mali Graphics Solution™ does not have any third-party licenses.

Falanx's Mali Graphics Solution™ provides end-users with the very highest image quality without compromising performance, power consumption or system cost. The technology, a result of six years of research and development, is embedded into a scalable and easy-to-integrate architectural framework and is ready to accelerate existing and future APIs for 2D and 3D graphics. The Mali Graphics Solution™ with its patent pending algorithms, is capable of accelerating video encoders and decoders, e.g. MPEG4, H.264.
Click to expand...
 
A very rough estimate but take the Mali200 vs 7800GTX pixel performance (No idea of shader performance on either, or geometry performance on 7800GTX... so this is all with a grain of salt.)

Mali200 = 300Mpix @ 200mhz
7800GTX = 10.32Bpix @ 430mhz

Bump the speed by 2.15 so the Mali is on part MHZ wise with 7800 (Mali 645Mpix @ 430)
10320Mpix / 645Mpix = 16 cores. Completely rough estimate... and I seriousally doubt it would only take that many, but I couldn't imaging 50 to 80 cores or pipelines on one chip.

Course, we could always quote Bill Gates too, so anything is possible.
 
Kyle, this is their 3rd generation revision/core. Are there any benchmarks (or anything of the like) available?

And 4xAA for free? Are they being serious or just saying it doesn't impact performance much. PR talk is different than real world.
 
be it true or not...one thing is for certain...is great to see a completely new approach to gfx solutions and hopefully it will deliver...

Second, its also true what someone already said...if it does deliver, who will buy it first....my guess its it would be nVidia...however, maybe it wont be like that...

If you thought seeing a new core from one of "the usual suspects" was exciting...
Few things are as exciting in the HW world as seeing a new player enter the game...

Just as it was when Sony entered with the PS, and when Microsoft with the XBOX in consoles....
 
So this would be a reeeeeeally good thing for the mini-ITX and mATX makers out there right?
 
. Now we are a long way from seeing this technology on the market, if we ever see it on the market at all
Click to expand...

Is it just me, or does anybody else stop reading once they get to this line?


I hope this is not "BitBoys" all over again
Click to expand...

:confused: Bitboys does make GPU's for phones and handheld devices you know
 
SulSeeker said:
A very rough estimate but take the Mali200 vs 7800GTX pixel performance (No idea of shader performance on either, or geometry performance on 7800GTX... so this is all with a grain of salt.)

Mali200 = 300Mpix @ 200mhz
7800GTX = 10.32Bpix @ 430mhz

Bump the speed by 2.15 so the Mali is on part MHZ wise with 7800 (Mali 645Mpix @ 430)
10320Mpix / 645Mpix = 16 cores. Completely rough estimate... and I seriousally doubt it would only take that many, but I couldn't imaging 50 to 80 cores or pipelines on one chip.
Click to expand...
All right, now take this plus the scalability of the cores, the low power consumption, and the small die size. What about 32 Mali200 cores at 1Ghz, consuming 16 Watts, with a total die size of <100 square mm? Tempting? I think so... Even if they can't run at 1Ghz, drop them back to your speed figure, cut the power consumption by more than half, and have an array that can theoretically double the performance of a GTX in 1/3 the die space with a fraction of the power.
 
And this comes back to the fact that it doesn't matter how fast theoretically it is if it doesn't have the drivers to back it. XGI anyone?
 
1mW = 1 thousandth of a watt.
1 thousandth = 0.001 now 0.0001, that's 1 millionth
0.001*200 = .2W

Good stuff. Interesting design. Can't wait to see it.
 
kcthebrewer said:
So, if their technology is sound, who is going to buy it first ATI or nVidia? :p
Click to expand...

ding ding ding we have a winner...

Anyone remember a company called GigaPixel? 3dfx bought them, nvidia ended up with the IP and a tile renderer that could do FSAA "for free" back in 2000 was sat on to push out more geforce chips.

Their GP1 chip was pretty sweet, running Q3 at 10x7 with 4xfsaa and using only 100k gates... But since it didn't have "TnL" which was nVidia's "next big thing", it didn't fit in with their marketing and where they intended to push the market..

They talked about using it for mobile devices and consoles as well, but as far as I know, nothing ever came of it.

And of course, without some silicone, we're basically dealing with a BitBoys Oy solution (insert ATI Crossfire joke here).
 
meanmodda[H] said:
So lemme get this straight , both Ati and Nvidia are blowing tons of wad on developing gfx cores and even a company the size of Via couldn't produce a competitive video card not to mention the others. Then a company comes along ,pulls some crap out of a hat and the investors get raped as usual. Call me when they have working samples that annihilates both ATi and Nvidia offerings for the same price , overpriced onboard chipset don't count since no manufacturer would support it and DX9 functionality don't count either see fx5200 etc . My 2 cents.
Click to expand...

Well that settles it... There will never be another graphics company. :rolleyes:

I am unsure when "crap out of a hat" equated to 4 years of government funded research, but these guys are far from being an Infinium Labs IMO.
 
This sounds to good to be true. Just like BitBoys awsome concept and idea but nothing ever came out of it. I doubt that this will ever see the light of day. It does look quite impresive and it would be nice integrated solution but we will know in 2006.
 
The Diagram for scaling is nuts..

Just imagine what this can do as a physics engine AND a GPU at the same time!

Wow I hope this comes out.. I'd expect the mobo's to be atleast $150 - $200 more then the usual. But if I'm able to run 16xFSAA without a new GPU.. DAMN I AM SOLD!
 
If they are able to sell to mainstream mini mobo vendors like Shuttle or VIA, that would be quite a boon for HTPC, imo. Cant wait.
 
if they are trying to go up against nvidia and ati than they better be providing something better than them both. Better drivers? please.

Now if we could just get nvidia to do a PCI-Express Soundstorm 2 card, that would solve the other monopoly of a market, sound cards. *mumbles something like down with creative*
 
with the skyrocketing prices of video cards i think someone should be able to come in and take the 100-200 dollar price range, i'm tired of having to spend 400+ dollars for a graphics card.
 
Deviationer said:
if they are trying to go up against nvidia and ati than they better be providing something better than them both. Better drivers? please.
Click to expand...

Everyone has to start from somewhere. AMD didn't come out right away being a "great" cpu alternative. They might not be the best right out the gate, but they might be very competative price vs performance wise (say like it handles 1024x768 with 16x AA but anything higher is crap for only $100)
 
The solution is only scalable if I can put more in chips into the mix. By this I mean once the motherboard is design is done the amount of processing power is fixed. Is this all that different from current solutions available to Dell from ATI or NVIDIA now?

While I'm sure Dell would like to see more integrated graphic solutions based on newer parts from NVIDIA/ATI, it seems like this solution won't impact the PC market greatly. The reason you don't see a 7800 GTX on a motherboard is cost. Perhaps their colution will cost significantly less, but I'd expect to see this solution really work better in the area of cell phones and PDAs.
 
You must log in or register to reply here.
Back
Top