• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

NVIDIA Introduces Industry's First Debugger and Profiler For GPU Computing

btarunr

Editor & Senior Moderator
Staff member
Joined
Oct 9, 2007
Messages
34,496 (9.18/day)
Likes
17,519
Location
Hyderabad, India
#1
The CUDA architecture continues to blaze a trail as the leading platform for developing and running GPU Computing applications, with support for C, OpenCL, DirectX Compute, Fortran and other languages and APIs. The latest CUDA 2.2 Beta contains a host of significant new features, including:

Hardware debugger for the GPU
Linux developers can now use a debugger on CUDA-enabled GPUs that offers both the familiar interface of the popular open-source GDB debugger and the ability to debug kernels as they execute on the GPU. The GPU-side debugger has all the features that developers expect from GDB, including the ability to have breakpoints, watch variables, inspect state, etc., as well as additional functions for CUDA-specific features.

Visual Profiler v2.2 for the GPU
The most common step in tuning application performance is profiling the application and then modifying the code. The CUDA Visual Profiler is a graphical tool that enables profiling C applications running on the GPU. This latest release of the CUDA Visual Profiler supports full measurement of memory bandwidth within a kernel, giving developers visibility into one of the most performance-critical areas of CUDA.

Full support for Microsoft Windows Server 2003/2008
Tesla C1060 and S1070 are now fully supported under both Microsoft Windows Server 2003 and 2008, offering developers and high-performance computing users more flexibility in their choice of operating system. CUDA 2.2 runs on Windows, MacOSX and major LINUX distributions.

Additional Features Coming with CUDA 2.2
  • Improved OpenGL interop performance
  • Texture from pitch linear memory
  • Zero-copy support for direct access to system memory
  • Pinned shared system memory allows compute kernels to share system memory
  • Asynchronous memcopy on Vista

For a full list of new features in CUDA 2.2 and to get early access, see discussion on the CUDA forums.

Source: NVIDIA
 

Mussels

Moderprator
Staff member
Joined
Oct 6, 2004
Messages
46,204 (9.52/day)
Likes
13,692
Location
Australalalalalaia.
System Name Daddy Long Legs
Processor Ryzen R7 1700, 3.9GHz 1.375v
Motherboard MSI X370 Gaming PRO carbon
Cooling Fractal Celsius S24 (Silent fans, meh pump)
Memory 16GB 2133 generic @ 2800
Video Card(s) MSI GTX 1080 Gaming X (BIOS modded to Gaming Z - faster and solved black screen bugs!)
Storage 1TB Intel SSD Pro 6000p (60TB USB3 storage)
Display(s) Samsung 4K 40" HDTV (UA40KU6000WXXY) / 27" Qnix 2K 110Hz
Case Fractal Design R5. So much room, so quiet...
Audio Device(s) Pioneer VSX-519V + Yamaha YHT-270 / sennheiser HD595/518 + Corsair Void RGB
Power Supply Corsair HX 750i (Platinum, fan off til 300W)
Mouse Logitech G403 + KKmoon desk-sized mousepad
Keyboard Corsair K65 Rapidfire
Software Windows 10 pro x64 (all systems)
Benchmark Scores Laptops: i7-4510U + 840M 2GB (touchscreen) 275GB SSD + 16GB i7-2630QM + GT 540M + 8GB
#2
wow. nvidia really is serious about cuda.
 
Joined
Mar 20, 2009
Messages
69 (0.02/day)
Likes
1
Processor i7970
Motherboard Gigabyte X58A-UD3R
Cooling Coolmaster V6
Memory 12gb DDR3-2000
Video Card(s) 2 GTX660 SLI
Storage 2 velociraptor 300gb, 1 Intel X25-M 80gb
Display(s) 3x Samsung 23' LCD
Audio Device(s) SoundBlaster X-Fi Platnium
Power Supply Corsair HX1000W
Software Windows 8 Enterprise X64
#3
They have to be if they want to beat AMD and Intel when it comes to ATi Stream and Havok
 
Joined
Aug 30, 2006
Messages
6,377 (1.53/day)
Likes
984
System Name ICE-QUAD // ICE-CRUNCH
Processor Q6600 // 2x Xeon 5472
Memory 2GB DDR // 8GB FB-DIMM
Video Card(s) HD3850-AGP // FireGL 3400
Display(s) 2 x Samsung 204Ts = 3200x1200
Audio Device(s) Audigy 2
Software Windows Server 2003 R2 as a Workstation now migrated to W10 with regrets.
#4
The race for CUDA (for graphics and for math) is just like the format wars of HDDVD and BlueRay, or Betamax and VHS, or ATRAC and MP3. It is who gets there first and is adopted first.

To win the CUDA race, nV *must* get their hardware out there NOW and programmers using their tools, so as to win the GPU-coding library wars.

One way to do that is to offer easy to use free developer software and good documentation. Another idea is to HALVE PRICES on enthusiasts boards :D, with special discounts to TPU users :D LOL

They need to do a competitor trade-up program. Hand in your X1, x2 or X3 generation ATI card for a half price nV GTX260 series. That would help their domination.
 
Joined
Sep 25, 2007
Messages
5,822 (1.54/day)
Likes
618
Processor Core I7 3770K@4.3Ghz
Motherboard AsRock Z77 Extreme
Cooling Cooler Master Seidon 120M
Memory 12Gb G.Skill Sniper
Video Card(s) MSI GTX 1070
Storage Sandisk SSD + 1TB Seagate Barracuda 7200
Display(s) IPS Asus 26inch
Case Antec 300
Audio Device(s) Xonar DG
Power Supply EVGA Supernova 650 G2
Software Windows 10/Windows 7
#5
I think Nvidia might be screwed with CUDA, ATI's STREAM has made alot of progress really fast, as soon as ATI gets some developers to utilize it nvidia will really have to get off their toes and I think thats starting to happen now
 
Joined
Aug 30, 2006
Messages
6,377 (1.53/day)
Likes
984
System Name ICE-QUAD // ICE-CRUNCH
Processor Q6600 // 2x Xeon 5472
Memory 2GB DDR // 8GB FB-DIMM
Video Card(s) HD3850-AGP // FireGL 3400
Display(s) 2 x Samsung 204Ts = 3200x1200
Audio Device(s) Audigy 2
Software Windows Server 2003 R2 as a Workstation now migrated to W10 with regrets.
#6
It might be too late, but nV need to get nV CUDA products into the universities and software houses that write and use library code. They also need to make a CUDA warehouse website to open source a lot more stuff. They are doing something, but not enough. If a developer wants to "try out" some CUDA for proof of concept, or to speed up certain code, or whatever, the support, libraries and cookbooks need to be there.

nV should do a "Developer month". Register as a developer, download the tools, access the forum, and should get a CUDA-COUPON for big discounts on the hardware.

And the developer forum should be actively supported with seasoned developers helping with code snippets for n00bies.
 

iStink

New Member
Joined
Sep 22, 2008
Messages
648 (0.19/day)
Likes
49
System Name The beast
Processor E6600@ 3.4
Motherboard P5W DH Deluxe
Cooling Tuniq Tower 120
Memory 3GB G.Sill DDR2-800
Video Card(s) 8800GT @ 700/975
Storage (1)80GB With XP, (1) 320GB storage
Display(s) 205BW & 710N
Case XClio Windtunnel (I love it)
Audio Device(s) X-Fi Platinum
Power Supply OCZ gamerXStream 700W
#7
I admire their efforts. It will be interesting to see if it pays off.
 
Joined
Mar 15, 2009
Messages
132 (0.04/day)
Likes
11
#8
little note havok is OpenCL based for gpu acceleration, and as i understand it, current nvidia beta's support opencl.....hence we SHOULD beable to accelerate HavoK with any cuda compatable card.
 
Joined
Aug 30, 2006
Messages
6,377 (1.53/day)
Likes
984
System Name ICE-QUAD // ICE-CRUNCH
Processor Q6600 // 2x Xeon 5472
Memory 2GB DDR // 8GB FB-DIMM
Video Card(s) HD3850-AGP // FireGL 3400
Display(s) 2 x Samsung 204Ts = 3200x1200
Audio Device(s) Audigy 2
Software Windows Server 2003 R2 as a Workstation now migrated to W10 with regrets.
#9