• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

NVIDIA Unveils Tesla V100s Compute Accelerator

btarunr

Editor & Senior Moderator
Staff member
Joined
Oct 9, 2007
Messages
47,853 (7.39/day)
Location
Dublin, Ireland
System Name RBMK-1000
Processor AMD Ryzen 7 5700G
Motherboard Gigabyte B550 AORUS Elite V2
Cooling DeepCool Gammax L240 V2
Memory 2x 16GB DDR4-3200
Video Card(s) Galax RTX 4070 Ti EX
Storage Samsung 990 1TB
Display(s) BenQ 1440p 60 Hz 27-inch
Case Corsair Carbide 100R
Audio Device(s) ASUS SupremeFX S1220A
Power Supply Cooler Master MWE Gold 650W
Mouse ASUS ROG Strix Impact
Keyboard Gamdias Hermes E2
Software Windows 11 Pro
NVIDIA updated its compute accelerator product stack with the new Tesla V100s. Available only in the PCIe add-in card (AIC) form-factor for now, the V100s is positioned above the V100 PCIe, and is equipped with faster memory, besides a few silicon-level changes (possibly higher clock-speeds), to facilitate significant increases in throughput. To begin with, the V100s is equipped with 32 GB of HBM2 memory across a 4096-bit memory interface, with higher 553 MHz (1106 MHz effective) memory clock, compared to the 876 MHz memory clock of the V100. This yields a memory bandwidth of roughly 1,134 GB/s compared to 900 GB/s of the V100 PCIe.

NVIDIA did not detail changes to the GPU's core clock-speed, but mentioned the performance throughput numbers on offer: 8.2 TFLOP/s double-precision floating-point performance versus 7 TFLOP/s on the original V100 PCIe; 16.4 TFLOP/s single-precision compared to 14 TFLOP/s on the V100 PCIe; and 130 TFLOP/s deep-learning ops versus 112 TFLOP/s on the V100 PCIe. Company-rated power figures remain unchanged at 250 W typical board power. The company didn't reveal pricing.



View at TechPowerUp Main Site
 
Just bought a 16 GPU Tesla V100 Server. Now i Need to sell it to upgrade it. Way to go nvidia.
Thats what you get for paying 300k EURO on Nvidia Server.
The middle Finger.
 
Just bought a 16 GPU Tesla V100 Server. Now i Need to sell it to upgrade it. Way to go nvidia.
Thats what you get for paying 300k EURO on Nvidia Server.
The middle Finger.

I see these type of comments all the time.
I'm sorry but... what did you expect, an email from them warning you about an upcoming product telling you "please, don't purchase this stock from us, we are launching a newer technology, like we always do on an annual basis"

Really?
 
Last edited:
Just bought a 16 GPU Tesla V100 Server. Now i Need to sell it to upgrade it. Way to go nvidia.
Thats what you get for paying 300k EURO on Nvidia Server.
The middle Finger.

According to people like you there must be no progress. It's getting tiresome really.

Do you remember the 90s and early zeros? CPUs got twice as fast each year and a half. You must have been constantly mad at the time. If you didn't upgrade your PC each 2-3 years it began to run at a snail's pace. Meanwhile recently you could easily get away with running your PC for up to 10 years without any issues. Sandy Bridge systems are still perfectly capable of web browsing and office work - just make sure you have more than 4 gigs of RAM. You didn't have this luxury just 20 years ago.
 
Last edited:
Just bought a 16 GPU Tesla V100 Server. Now i Need to sell it to upgrade it. Way to go nvidia.
Thats what you get for paying 300k EURO on Nvidia Server.
The middle Finger.
Thank god amd announce their competing cards with a years notice.
 
Just bought a 16 GPU Tesla V100 Server. Now i Need to sell it to upgrade it. Way to go nvidia.
Thats what you get for paying 300k EURO on Nvidia Server.
The middle Finger.

My hunch is this is a tongue-in-cheek comment, but on the off chance you're serious...

Did you spec your purchase appropriately for the intended workload?

Then you have nothing to complain about. The hardware you purchased is just as fast today, despite the introduction of new hardware.
 
My hunch is this is a tongue-in-cheek comment, but on the off chance you're serious...

Did you spec your purchase appropriately for the intended workload?

Then you have nothing to complain about. The hardware you purchased is just as fast today, despite the introduction of new hardware.
I think you didn't get the point.
When they released the new hardware, it means 2 things:
1. old hardware becomes cheaper and the improved one takes its current price
2. based on 1. he could either have gotten the same performance CHEAPER or a GREATER performance for the same cost, either way, he got screwed.

Surely we are talking about a 15-20% increase, but still.
 
I think you didn't get the point.
When they released the new hardware, it means 2 things:
1. old hardware becomes cheaper and the improved one takes its current price
2. based on 1. he could either have gotten the same performance CHEAPER or a GREATER performance for the same cost, either way, he got screwed.

Surely we are talking about a 15-20% increase, but still.

This is a consumer attitude. In a professional setting, you spec hardware purchases for the intended workload PERIOD. Something better will always be available, sometimes even right around the corner. Capital budgets, project end dates, hardware refresh cycles, and a whole host of other factors determine when purchases are made, not vendor release dates.
 
Back
Top