|
|
#1 |
|
Just wondering
Join Date: May 2002
Location: Germany
Posts: 1,377
|
Today, intel showed the first detailed die-fotograph of Larrabee - some pictures are to be seen at PC Games Hardware.
http://www.pcgameshardware.de/aid,68...fikkarte/News/ Some of the functional units can be distinguished also - pleases take your guesses as to their respecitive number now. Some photoshopped versions of the pics are soon to follow in the image gallery: http://www.pcgameshardware.de/aid,68...&show=original I really hope I#ve not been living under a rock the last couple of months and this is really something new.
__________________
English is not my native tongue. So please, before being too nitpicky about my choice of words, please consider the possiblity that I did not mean to say what you might have read into them and inquire if not sure. TwitterDid you know? The Future is Fusion! |
|
|
|
|
|
#2 |
|
Regular
|
Awesome, great wake-up surprise.
Looks like 32 cores and 8 texture units. Jawed |
|
|
|
|
|
#3 |
|
Member
Join Date: May 2002
Location: Austria
Posts: 688
|
I don't know much about how hw structures look on a die shot but this seemed fairly obvious:
|
|
|
|
|
|
#5 |
|
Member
|
|
|
|
|
|
|
#6 |
|
Regular
|
What looks like the GDDR interface is interesting as there's so much of it. If the die is 30x20 (which I think is a bit small) then that means the DDR interface is about 54mm.
The 256-bit GDDR5 on RV770 is about 36mm. The 512-bit GDDR3 on GT200 is about 75mm. Larrabee has a 384-bit bus? GDDR3? Jawed Last edited by Jawed; 12-May-2009 at 19:41. Reason: G80->GT200 - GT200=75 - RV770=36 |
|
|
|
|
|
#7 |
|
Senior Member
|
Can anyone spot the PCIe interface pads somewhere?
__________________
Baby, It's Cold Outside
|
|
|
|
|
|
#8 |
|
Senior Member
Join Date: Feb 2002
Posts: 2,813
|
Is Larrabee going to be competitive with 8 texture units ?
|
|
|
|
|
|
#9 |
|
Regular
|
|
|
|
|
|
|
#10 |
|
Regular
|
Assuming they're quads, that's 32. The clock is the real killer question.
Larrabee should have the best overdraw reduction (other GPUs' early Z rejection is coarse-grained and so conservative). Dunno, at 1GHz it doesn't seem likely. Jawed |
|
|
|
|
|
#11 |
|
Member
Join Date: May 2002
Location: Austria
Posts: 688
|
|
|
|
|
|
|
#12 |
|
Regular
|
|
|
|
|
|
|
#13 |
|
Member
Join Date: Sep 2004
Location: Eastbourne, England, UK.
Posts: 468
|
I bet this is a mainstream part. They are saving the high end to surprise us. nearer launch.
__________________
____________________________________ Why buy one, when you can have two at twice the price? |
|
|
|
|
|
#14 |
|
Senior Member
|
![]() Red - 32 cores Green - 8 structures (?) Blue - 4 structures (?)
__________________
Baby, It's Cold Outside
|
|
|
|
|
|
#15 | |
|
ಠ_ಠ
|
Quote:
__________________
"He's more gamerscore than man now." -Referring to Wavey --- <Burga>stop whining <DeanoC>no --- <zomgfarhan>i can get it up just fine thx <KidBeta>beta was useful last night |
|
|
|
|
|
|
#16 |
|
Member
Join Date: May 2002
Location: Austria
Posts: 688
|
One more, because photoshop is fun.
![]() If the green blocks are texture units, then what are the blue ones? They are strangely asymmetrical. |
|
|
|
|
|
#17 |
|
Regular
|
Blue ones for bus<->memory?
One thing I'm wondering is if Intel has built the DDR interface in two parts, GDDR3, say 256-bit and GDDR5, 128-bit. Rather than trying to make the interfacing multifunctional. Or, maybe it's 256-bit GDDRx + 128-bit "communicate with another Larrabee"? Jawed |
|
|
|
|
|
#18 |
|
Senior Member
|
The x86 cores are very Atom-like in shape. The L2 array can definitely be spotted in the one end of the tile, and in the opposite end there is also some uniform array of structures (registers ?).
__________________
Baby, It's Cold Outside
|
|
|
|
|
|
#19 |
|
Senior Member
Join Date: Sep 2003
Location: Well within 3d
Posts: 2,535
|
The cores are more svelt than my guesstimated lower bound of ~18mm2.
It's what, 15 or so? I unfortunately don't have time to do more than glance. At 32 cores, Larrabee would have two short-linked rings of 16. Other statements said that the bus was routed over the L2. The odd part is that the caches don't really point to a very even distribution. The four blue sections would fit with the desire to space memory controller ring stops evenly, and they could also be used to serve as the connection points that connect the two rings.
__________________
Dreaming of a .065 micron etch-a-sketch. |
|
|
|
|
|
#20 |
|
Regular
|
12mm˛?
|
|
|
|
|
|
#21 |
|
Senior Member
|
Now, I wonder what will heat up this thing more -- FurMark or some weird port of the IBT (Linpak) test?
__________________
Baby, It's Cold Outside
|
|
|
|
|
|
#22 |
|
Senior Member
|
![]() Silverthorne die (Atom) in a LRB disguise -- minus half the L2 and the AGTL+ interface. Imagine the SIMD unit for yourself!
__________________
Baby, It's Cold Outside
|
|
|
|
|
|
#23 |
|
Member
Join Date: Nov 2006
Location: San Antonio, TX
Posts: 903
|
Can anyone guess a die size on this one?
__________________
ATi Beta Tester AMD Panelist Member |
|
|
|
|
|
#24 |
|
ಠ_ಠ
|
__________________
"He's more gamerscore than man now." -Referring to Wavey --- <Burga>stop whining <DeanoC>no --- <zomgfarhan>i can get it up just fine thx <KidBeta>beta was useful last night |
|
|
|
|
|
#25 | |
|
Member
Join Date: Nov 2006
Location: San Antonio, TX
Posts: 903
|
Quote:
__________________
ATi Beta Tester AMD Panelist Member |
|
|
|
|
![]() |
| Bookmarks |
| Tags |
| analysis, intel |
| Thread Tools | |
| Display Modes | |
|
|