Guru3D.com
  • HOME
  • NEWS
    • Channels
    • Archive
  • DOWNLOADS
    • New Downloads
    • Categories
    • Archive
  • GAME REVIEWS
  • ARTICLES
    • Rig of the Month
    • Join ROTM
    • PC Buyers Guide
    • Guru3D VGA Charts
    • Editorials
    • Dated content
  • HARDWARE REVIEWS
    • Videocards
    • Processors
    • Audio
    • Motherboards
    • Memory and Flash
    • SSD Storage
    • Chassis
    • Media Players
    • Power Supply
    • Laptop and Mobile
    • Smartphone
    • Networking
    • Keyboard Mouse
    • Cooling
    • Search articles
    • Knowledgebase
    • More Categories
  • FORUMS
  • NEWSLETTER
  • CONTACT

New Reviews
FSP Hydro PTM Pro (1200W PSU) review
ASUS ROG Radeon RX 6750 XT STRIX review
AMD FidelityFX Super Resolution 2.0 - preview
Sapphire Radeon RX 6650 XT Nitro+ review
Sapphire Radeon RX 6950 XT Sapphire Nitro+ Pure review
Sapphire Radeon RX 6750 XT Nitro+ review
MSI Radeon RX 6950 XT Gaming X TRIO review
MSI Radeon RX 6750 XT Gaming X TRIO review
MSI Radeon RX 6650 XT Gaming X review
Deepcool AS500 PLUS CPU Cooler Review

New Downloads
FurMark Download v1.30
Display Driver Uninstaller Download version 18.0.5.1
Download Samsung Magician v7.1.1.820
Intel ARC graphics Driver Download Version: 30.0.101.1732
HWiNFO Download v7.24
GeForce 512.77 WHQL driver download
Intel HD graphics Driver Download Version: 30.0.101.1960
AMD Radeon Software Adrenalin 22.5.1 WHQL driver download
3DMark Download v2.22.7359 + Time Spy
Prime95 download version 30.8 build 15


New Forum Topics
[3rd-Party Driver] Amernime Zone Radeon Insight 22.5.1 WHQL Driver Pack (Released) AMD FidelityFX Super Resolution 2.0 - Deathloop preview GeForce RTX 4090 is twice as fast as RTX 3090 and uses 450W (16128 Shader Cores) EK Announces Quantum CPU Water Blocks with Walnut Wood Tops 5700x & B550 TUF NVIDIA Releases Kepler GPU Security Update 473.47 WHQL Driver New Upcoming ATI/AMD GPU's Thread: Leaks, Hopes & Aftermarket GPU's NVIDIA GeForce 512.77 WHQL driver download & Discussion NVIDIA GeForce 473.47 Driver - Kepler Security Update AMD could be developing a 24-core Ryzen 9 7950X CPU with a TDP of 170 W.




Guru3D.com » News » Nvidia GP100 GPU architecture recap - full GPU has 3840 Shader processors

Nvidia GP100 GPU architecture recap - full GPU has 3840 Shader processors

by Hilbert Hagedoorn on: 04/12/2016 09:19 AM | source: | 47 comment(s)
Nvidia GP100 GPU architecture recap - full GPU has 3840 Shader processors

Last week Nvidia announced the GP100 GPU powering the Tesla P100 HPC module. More rumors are surfacing about GP104 as well. Since the full block diagrams for Gp100 are now available, we can also tell what a full GP100 looks like when fully enabled, in this post a little recap on the GP100 architecture and its positioning. 

This year several GPUs are going to be released from Nvidia, all based on their new Pascal architecture in a wide variety of segments in the market channels. For consumers the first wave of graphics cards it will be the GP104 GPU, these are empowering the high-end products like 'GTX 980' class products, the current rumor is that the new GTX 1070 and 1080 albeit with a bit of weird Full HD like naming, will use that chip. These should be announced during Computex time in June with availability in the summer (likely July). 

Then there is big Pascal, the big daddy Nvidia GPU developed under GPU codename GP100. This is the GPU that will empower (for the consumer side) the enthusiast class products e.g. the Titan etc. Make no mistake, this product will not launch anytime soon for consumers. Expect at the very best a launch late this year closer to the Christmas season, likely even later in Q1/Q2 2017 (we think). 

All Pascal products are based on a 16nm FinFet design and the GP100 in particular comes with stacked HBM2 (16GB in four stacks). The Pascal based GPU driving the unit holds 15 Billion transistors which is roughly double that of the current biggest Maxwell chip. Gp100 is huge at 600mm^2. The prognosis performance (according to Nvidia) is 5.3TFLOPS using 64-bit floating-point numbers and is rated at 10.6TFLOPS using 32-bit and 21.2TFLOPS using 16-bit. P100 has 4MB of L2 cache and 14MB of shared memory for just the register file. The following table provides a high-level comparison of Tesla P100 specifications compared to previous-generation Tesla GPU accelerators, however I added the GP100 as a fully enabled product:

Tesla Products Tesla K40 Tesla M40 Tesla P100 GP100
GPU GK110 (Kepler) GM200 (Maxwell) GP100 (Pascal) GP100 (Pascal)
SMs 15 24 56 60
TPCs 15 24 28 30
FP32 CUDA Cores / SM 192 128 64 64
FP32 CUDA Cores / GPU 2880 3072 3584 3840
FP64 CUDA Cores / SM 64 4 32 32
FP64 CUDA Cores / GPU 960 96 1792 1920
Base Clock 745 MHz 948 MHz 1328 MHz ~1328 MHz
GPU Boost Clock 810/875 MHz  1114 MHz 1480 MHz ~1480 MHz
Texture Units 240 192 224 240
Memory Interface 384-bit GDDR5 384-bit GDDR5 4096-bit HBM2 4096-bit HBM2
Memory Size Up to 12 GB Up to 24 GB 16 GB 16 GB
L2 Cache Size 1536 KB 3072 KB 4096 KB 4096 KB
Register File Size / SM 256 KB 256 KB 256 KB 256 KB
Register File Size / GPU 3840 KB 6144 KB 14336 KB 14336 KB
TDP 235 Watts 250 Watts 300 Watts ~300 Watts
Transistors 7.1 billion 8 billion 15.3 billion 15.3 billion
GPU Die Size 551 mm² 601 mm² 610 mm² 610 mm²
Manufacturing Process 28-nm 28-nm 16-nm 16-nm

As the block diagram now shows, the GP100 features six graphics processing clusters (GPCs). Just look at the diagram and count along with me - each GPC holds 10 streaming multiprocessors (SMs) and then each SM has 64 CUDA cores and four texture units. Do the math and you'll reach 640 shader processors per GPC and 3840 shader cores with 240 texture units in total.

  • 6 (GPC) x (10x64) = 3840 Shader processor units in total.

Meaning the GP100 used on the Tesla P100 is not fully enabled. Nvidia is known to out GPU that have disabled segments, it helps them selling different SKUs, the Tesla P100 holds a shader count of 3584 and thus has 56 SMs enabled (from the 60).

GP100’s SM incorporates 64 single-precision (FP32) CUDA Cores. In contrast, the Maxwell and Kepler SMs had 128 and 192 FP32 CUDA Cores, respectively. The GP100 SM is partitioned into two processing blocks, each having 32 single-precision CUDA Cores, an instruction buffer, a warp scheduler, and two dispatch units. While a GP100 SM has half the total number of CUDA Cores of a Maxwell SM, it maintains the same register file size and supports similar occupancy of warps and thread blocks.GP100’s SM has the same number of registers as Maxwell GM200 and Kepler GK110 SMs, but the entire GP100 GPU has far more SMs, and thus many more registers overall. This means threads across the GPU have access to more registers, and GP100 supports more threads, warps, and thread blocks in flight compared to prior GPU generations.

Since the graphics memory is on-die HBM2, the VRAM amount is fixed. That means that ALL GP100 products will get 16GB of memory. HBM2 will run a wide 4096-bit HBM2 (1024 bit per IC stack) memory interface running an effective bandwidth anywhere up-to a full 1 TB/s.

This is a big chip, very big at 600mm^2 hence it is interesting to see that 16nm can offer a lot in terms of clock frequency, The Tesla P100 is an enterprise part that ends up in servers, however this part already is clocked at 1328 MHz with Boost capabilities towards a frequency of 1480 MHz. Combined the TDP still remains to be under 300W. 



Nvidia GP100 GPU architecture recap - full GPU has 3840 Shader processors Nvidia GP100 GPU architecture recap - full GPU has 3840 Shader processors




« Nvidia Pascal Consumer card announced during Computex · Nvidia GP100 GPU architecture recap - full GPU has 3840 Shader processors · SteelSeries Apex M500 Gaming Mechanical Keyboard »

Related Stories

Download Nvidia GeForce 364.72 WHQL drivers - 03/28/2016 03:17 PM
You can download the new Nvidia GeForce 364.72 WHQL driver for Windows 7 - 8.1 and 10 in 32 and 64-bit flavors. This title adds VR support and game ready support for Quantum Break, Killer Instinct, ...

Nvidia GeForce X80 and X80 Ti Pascal Specs? - 03/17/2016 08:57 AM
In our forums somebody posted the spec list shown after the break. Indicating the Pascal graphics cards would be called GeForce X80 and X80Ti....

NVIDIA GameWorks SDK 3.1 Released - 03/15/2016 08:28 AM
NVIDIA today announced worldwide availability of the NVIDIA GameWorks software development kit (SDK) 3.1, which introduces three groundbreaking graphics techniques for shadows and lighting as well as ...

Download Nvidia GeForce 364.51 WHQL driver - 03/08/2016 09:23 PM
Nvidai updated to a new WHQL driver. This is a new driver following the issues Nvidia had with multi-monitor support. The latest version, 364.51 This driver ensures you will have an optimal experienc...

Download Nvidia GeForce 364.47 WHQL driver - 03/07/2016 04:38 PM
Nvidia just released a new GeForce 364.47 WHQL driver. It's aimed at an optimal experience with Tom Clancy’s The Division, Hitman, Need for Speed, Ashes of the Singularity, and Rise of th...


10 pages « 3 4 5 6 > »


PrMinisterGR
Senior Member



Posts: 7975
Joined: 2014-09-27

#5257045 Posted on: 04/12/2016 04:03 PM
It's not like Nvidia is marketing GP100 towards gaming. Everything they said about it was related to HPC and they did say that the Tesla variant won't hit OEM's till Q4, so yeah, basically a year for a consumer one.

It depends on your definition of excitement. It's becoming pretty obvious that both AMD's Polaris 10 and Nvidia's GP104 aren't going to be that much faster then what we have already. Just far more efficient.

They don't have to tease or "officially" market. They have created enough anticipation in enthusiasts that we drool all over it by ourselves, thank you. This is marketing 101 (literally it is, I have been taught this in a university course :P ).

As for the new cards, if they are at 8GB minimum and they have more efficient chips with the same amount of shaders at crazy frequencies, I'm ok with that. An 8GB super-tweaked Fiji @1.6GHz is seriously fine with me, and I would upgrade.

rl66
Senior Member



Posts: 3370
Joined: 2007-05-31

#5257047 Posted on: 04/12/2016 04:07 PM
GDDR5 packages are not pin compatible GDDR5X.


not as CPU that you change but like an USB to old iPhone plug for those who don't know it is really easy).

most GDDR5 controler work as is with GDDR5X (256bit vs 512bit, and the voltage are lower on the X)

BD2015
Member



Posts: 51
Joined: 2015-08-08

#5257054 Posted on: 04/12/2016 04:20 PM
I'm wondering about a name for the new Titan.

We already had Titan, Titan Black, Titan Z, Titan X.. What's next? Titan Y? Titan 4k? (Would be funny if they really sticked with naming GP104 - 1080) Titan XXL? Titan King?

Mannerheim
Member



Posts: 4885
Joined: 2004-01-24

#5257055 Posted on: 04/12/2016 04:21 PM
300W fail... unless watercooled.

Fox2232
Senior Member



Posts: 11809
Joined: 2012-07-20

#5257056 Posted on: 04/12/2016 04:26 PM
not as CPU that you change but like an USB to old iPhone plug for those who don't know it is really easy).

most GDDR5 controler work as is with GDDR5X (256bit vs 512bit, and the voltage are lower on the X)
You have it really wrong, while biggest increase on pins on GDDR5X comes from proper voltage/grounding delivery, signaling pins are different and on top of that there is one extra. They use different communication method + ability to use 1/2 of signaling each for 2 chips. Effectively reducing bandwidth to 1/2 per GDDR5X package, but doubling number of packages.
While it is not that hard to move from GDDR5 to GDDR5X, it is not like plugging GDDR5X into GDDR5 memory controller will do you any good It will not work.

10 pages « 3 4 5 6 > »


Post New Comment
Click here to post a comment for this news story on the message forum.


Guru3D.com © 2022