NVIDIA Announces RTX A6000 48 GB Professional Graphics Card Accelerators

[Raevenlord]

NVIDIA today announced their RTX A6000 series of graphics cards, meant to perform as graphics accelerators for professional workloads. And the announcement marks a big departure for the company's marketing, as the Quadro moniker has apparently been dropped. The RTX A6000 includes all raytracing resources also present on consumer RTX graphics cards, and marks a product segmentation from the company's datacenter-geared A40. The RTXA6000 features a full-blown GA102 chip - meaning 10752 CUDA cores powering single-precision compute performance of up to 38.7 TFLOPs (3.1 TLFOPs higher than that of the GeForce RTX 3090). Besides offering NVIDIA's professional driver support and features, the RTX A6000 features 48 GB of GDDR6 (note the absence of the X) memory - ensuring everything and the kitchen sink can be stored in the cards' VRAM. GDDR6X doesn't currently offer the per-chip density of GDDR6 solution, hence why NVIDIA opted for the lower-performing, yet denser memory variant.

The RTX A6000 features a classic blower-type cooler, and presents a new low-profile NVLink bridge that enables two of them to work in tandem within the same system. NVIDIA vGPU virtualization technologies are supported as well; display outputs are taken care of by 4x DisplayPort connectors, marking the absence of HDMI solutions. The card is currently listed for preorder at a cool and collected $5,500, but with insufficient silicon to offer even to its highest-margin datacenter customers, it remains to be seen exactly how available these will be in the market.



View at TechPowerUp Main Site

 

[lexluthermiester]

This was predictable. Wafer yields must be improving.

 

[Fouquin]

Did this not launch back in October?

 

[lexluthermiester]

Did this not launch back in October?

Don't think so. Otherwise why would they be announcing it now?

 

[mouacyk]

Did this not launch back in October?

No, Anandtech posted it properly as an announcement in October but some sites incorrectly reposted it as a launch. NVidia's own announcement actually stated availability in mid-December:
NVIDIA Ampere GPUs for Professional Designers Bring Unparalleled Performance to Desktop, Data Center and Cloud | NVIDIA Blog

@Raevenlord Perhaps, this article needs to be titled Launch instead of Announcement?

 

[Fouquin]

No, Anandtech posted it properly as an announcement in October but some sites incorrectly reposted it as a launch. NVidia's own announcement actually stated availability in mid-December:
NVIDIA Ampere GPUs for Professional Designers Bring Unparalleled Performance to Desktop, Data Center and Cloud | NVIDIA Blog

@Raevenlord Perhaps, this article needs to be titled Launch instead of Announcement?

Ah yep, it was just the announcement. Got it.

Quadro No More? NVIDIA Announces Ampere-based RTX A6000 & A40 Video Cards For Pro Visualization

www.anandtech.com

Nvidia launches Ampere A40 & RTX A6000 cards for graphics professionals

Nvidia’s Ampere graphics cards for ProViz professionals drop Quadro brand

www.techradar.com

Some were indeed saying it was a launch in October. Very confusing.

 

[Nater]

To think I actually waited for the "A4000" instead of just scooping up an RTX 4000. And now the RTX 4000 is about +$200 vs when I was going to do a build 4-5 months ago.

 

[mouacyk]

@Nater Curious what you're using these accelerators for.

 

[Caring1]

@Raevenlord Perhaps, this article needs to be titled Launch instead of Announcement?

Small print says availability 2 - 4 weeks, so early January.
It's not a Launch yet.

I don't see how they can operate effectively being that close together and not get hot or overheat.
Blower fans are not designed for that style of cooling and require an air gap to individually draw air in.

 

[erek]

whats different in terms of features between the RTX A6000 and 3090?

 

[mouacyk]

48GB of GDDR6 vs 24GB of GDDR6X and full die vs partial.

 

[lexluthermiester]

whats different in terms of features between the RTX A6000 and 3090?

More compute cores, more VRAM.

48GB of GDDR6 vs 24GB of GDDR6X and full die vs partial.

Ninja'd... LOL!

 

[Vya Domus]

It's actually cheaper than Quadros of the past considering how much memory it has, I wonder why is that.

 

[erek]

is the RTX 3090 gimped in performance Double Precision or something? what are the pro features that are disabled?

 

[mouacyk]

It's actually cheaper than Quadros of the past considering how much memory it has, I wonder why is that.

Ignoring /s -- VideoCardz speculated that NVidia hoarded GDDR6, and with the shortage of GPUs needing GDDR6, it was a good opportunity to offload them. At full bus-width, the 16gbps speed should still pack plenty of bandwidth for any scientific use.

 

[lexluthermiester]

It's actually cheaper than Quadros of the past considering how much memory it has, I wonder why is that.

Reduced cost is the likely reason. NVidia can make them for less while still make a solid profit and they're passing on the saving. They kinda have to as AMD is on their heels with FirePro(?) offerings.

 

[mouacyk]

is the RTX 3090 gimped in performance Double Precision or something? what are the pro features that are disabled?

Some quick googling suggests that the 3090 is 1/64 fp64 while the A6000 is 1/32.

 

[DeathtoGnomes]

A6000

If someone said they just bought one of these I would have guessed it was an AMD card, but adding RTX in front of it raises an eyebrow.

So when is the A6080 launch?

 

[Lycanwolfen]

Hey Look SLI

 

[erek]

Hey Look SLI

excited?

 

[mechtech]

Probably better value than the 3090

 

[Solaris17]

So when is the A6080 launch?

The consumer edition of this card will be the new titan.

 

[enzolt]

The consumer edition of this card will be the new titan.

Yup! Personally waiting for full chip Titan Ampere!

 

[Crackong]

That fan must be drawing air from the 4th dimension in Pic No.2

 

[Chiven]

That fan must be drawing air from the 4th dimension in Pic No.2

It has cutout on the back