⋮    ⋮  

NVIDIA GeForce RTX 40 ‘Ada Lovelace’ GPU Rumors: Graphics Card Lineup Detailed, Q3 Launch, Possible GDDR7 Memory Support

Submit

Moore's Law is Dead has posted some meaty information regarding NVIDIA's GeForce RTX 40 'Ada Lovelace' GPU lineup. The new rumors include all the way from launch and performance details to the full lineup and what we can expect from each tier that the green team is going to offer later this year.

NVIDIA Next-Gen GeForce RTX 40 'Ada Lovelace' GPU Lineup, Rumored For Q3 Launch To Tackle AMD's RDNA 3 MCM GPU Lineup

Most of the information that MLID has shared is something that we have come to know over the last couple of months and more recently in the NVIDIA leak which revealed not only the GPU code names but also the exact core configuration that we will be getting in the next-generation lineup.

NVIDIA To Finally Release GeForce GTX 1630, The Successor To GT 1030: Faster Than GTX 1050 Ti In Performance

NVIDIA GeForce RTX 40 'Ada Lovelace' GPU family rumors. (Image Credits: Moore's Law is Dead)

Power Consumption Maxing Out at 600W For Ada Lovelace Flagship?

Starting with the details, it is clear that NVIDIA is pushing the power figures for its entire GeForce RTX 40 'Ada Lovelace' GPU lineup compared to their initial targets. Hence it is why we have been hearing some absurd rumors of power figures going up to 850W but those are just internal test cases and MLID states that the nominal power consumption of the flagship AD102 GPU die will land somewhere between 500-600W and the upper limit is doable on air-cooling thanks to AIBs already invested in triple and quad-slot cooling solutions with hefty heatsinks. Anything above is going to require more serious cooling solutions and we can see liquid-cooled or hybrid models featured on the ultra-premium GPU variants.

NVIDIA is already investing development around the new PCIe Gen 5 connector that offers up to 600W power input per connector. The delayed GeForce RTX 3090 Ti is one example where the card is expected to rock at a TGP of 450W and will be the first desktop graphics card to utilize such a connector interface. The next-gen cards are also expected to utilize the same PCIe standard.

Several PSU makers have already started releasing their brand new Gen 5 power supplies which would include the necessary connectors to support the next-gen GPUs but they only feature one primary Gen 5 connector which means that if NVIDIA was to use a second 16-pin port, users will have to use a 2x 8-pin to 1x 16-pin adapter which will ship with most of these PSUs.

GDDR6X Is Nice But Ada Might Also Be Ready For The Next-Gen GDDR7 Memory Standard

NVIDIA Kepler GPUs Get Security Updated In GeForce 473.47 WHQL Driver

One interesting thing mentioned by MLID is that his sources seem to be reporting the inclusion of GDDR7 memory controllers within Ada Lovelace GPUs. Now NVIDIA is yet to offer the full GDDR6X speeds (24 Gbps) on any of its graphics cards and even the upcoming RTX 3090 Ti is only expected to utilize just 21 Gbps G6X modules so there's still some room left for the memory standard. It is likely that this could be more of a future-proofing move by NVIDIA where they could initially offer Ada Lovelace Rev.1 on GDDR6X memory and move to GDDR7 with an Ada Lovelace Rev.2 refresh (RTX 40 SUPER?) later on.

The GDDR7 memory is expected to offer up to 32 Gbps transfer speeds along with a real-time error protection technology. A GDDR7 memory sub-system across a 256-bit wide bus interface at 32 Gbps transfer speeds will offer 1 TB/s of total bandwidth. That's 1.5 TB/s with a 384-bit bus interface and up to 2 TB/s on a 512-bit system. This is just insane amounts of bandwidth for a GDDR standard.

GDDR Memory Specifications:

FeatureGDDR5GDDR5XGDDR6GDDR6XGDDR6+GDDR7
DensityFrom 512Mb to 8Gb8Gb8Gb, 16Gb8Gb, 16Gb8Gb, 16Gb8Gb, 16Gb, 32Gb?
VDD and VDDQEither 1.5V or 1.35V1.35VEither 1.35V or 1.25VEither 1.35V or 1.25VTBDTBD
VPPN/A1.8V1.8V1.8V1.8VTBD
Data ratesUp to 8 Gb/sUp to 12Gb/sUp to 16 Gb/s19 Gb/s, 21 Gb/s,
>21 Gb/s
24 Gb/s32 Gb/s
PackageBGA-170
14mm x 12mm 0.8mm ball pitch
BGA-190
14mm x 12mm 0.65mm ball pitch
BGA-180
14mm x 12mm 0.75mm ball pitch
BGA-180
14mm x 12mm 0.75mm ball pitch
BGA-180
14mm x 12mm 0.75mm ball pitch?
TBD
I/O widthx32/x16x32/x162 ch x16/x82 ch x16/x82 ch x16/x8TBD

NVIDIA GeForce RTX 40 'Ada Lovelace' GPU / Graphics Card Configurations

According to the leak, NVIDIA will have at least six GPUs within its Ada Lovelace lineup. These will include AD102, AD103, AD104, AD106, AD107, and AD10B. The first five SKUs will be designed for the desktop and mobility segments and featured in both GeForce RTX 40 and RTX Workstation solutions.

NVIDIA GeForce RTX 40 'Ada Lovelace' GPU Lineup rumors. (Image Credits: Moore's Law is Dead)

MLID only reports with confidence on the AD102 GPU rumors and other performance data are mostly his speculation which are provided below:

  • RTX 4090 (AD102) -> +80-110% Rasterization vs RTX 3090 (GA102) / Double RT Perf
  • RTX 4070 (AD103) -> +10-30% Rasterization vs RTX 3090 (GA102) / 2-3 Tier Better RT Perf
  • RTX 4060 (AD104) -> Same Rasterization as RTX 3080 (GA102) / 3090-Like RT Perf
  • RTX 4050 Ti (AD106) -> Same Rasterization as RTX 2080 Ti (TU102) / 3070-Like RT Perf
  • RTX 4050 (AD107) -> Same Rasterization as RTX 3060 (GA106) / 3060-Like RT Perf

So starting with the leaked SKUs, the top AD102 GPU which is likely going to power the next-gen GeForce RTX 4090, RTX 4080 Ti graphics cards will make use of 144 SMs, a 71% increase over the existing GA102 GPU and house a massive 18,432 CUDA core count. The interesting thing here is that the AD102 GPU is the only SKU that is getting over a 50% increase in SM count & considering what we have heard about the flagship chip, in terms of performance and power consumption, it looks very likely that NVIDIA is going all out with its top chip in the Ada Lovelace family.

The AD103 GPU will replace the GA103 GPU which was recently introduced on mobile and feature the same SM count as the GA102 GPU at 84. The AD104, AD106, and AD107 GPUs will feature 60, 36, and 24 SM units, respectively. Besides the AD103 GPU which is a 40% SM increase over GA103, every other GPU gets a 25-20% SM count increase over its predecessor. It's not as significant as the AD102 GPU but considering this is the mainstream segment, we are likely going to get RTX 3080 or similar performance out of an RTX 4060 Ti & RTX 3070 or higher performance out of the standard RTX 4060. The RTX 4050 should be close or on par with an RTX 3060 given the addition of IPC and clock improvements aside from architectural upgrades.

In addition to the SM counts, the Ada Lovelace GPUs will also feature increased L2 cache sizes. Starting with the AD102 GPU, the flagship would be outfitted with up to 96 MB of L2 cache, an insane 16x increase over the 6 MB L2 cache featured on GA102. The AD103 GPU will feature 64 MB, AD104 will feature 48 MB while both AD106/AD107 GPUs will feature 32 MB of L2 cache. As for the memory bus, the flagship AD102 GPU will feature a 384-bit bus interface, the AD103 GPU will get a 256-bit bus interface, AD104 will feature a 192-bit bus interface, while the AD106/AD107 GPUs will get a 128-bit bus interface.

NVIDIA Ada Lovelace 'GeForce RTX 40' GPU Configurations

GPU NameGPCs / TPCsSMs Per TPC / TotalCUDA CoresL2 CacheMemory Bus
AD10212 / 62 / 1441843296 MB384-bit
AD1037 / 62 / 841075264 MB256-bit
AD1045 / 62 / 60768048 MB192-bit
AD1063 / 62 / 36460832 MB128-bit
AD1073 / 42 / 24307232 MB128-bit
NVIDIA GeForce RTX 40 'Ada Lovelace' vs AMD Radeon RX 7000 'RDNA 3' GPUs. (Image Credits: Moore's Law is Dead)

NVIDIA Ada Lovelace & Ampere GPU Comparison

Ada Lovelace GPUSMsCUDA CoresGraphics Card SeriesMemory BusAmpere GPUSMsCUDA CoresTop SKUMemory BusSM Increase (% Over Ampere)
AD10214418432RTX 4090?384-bitGA1028410752RTX 3090 Ti384-bit+71%
AD1038410752RTX 4080?256-bitGA103S607680RTX 3080 Ti256-bit+40%
AD104607680RTX 4070?192-bitGA104486144RTX 3070 Ti256-bit+25%
AD106364608RTX 4060?128-bitGA106303840RTX 3060192-bit+20%
AD107243072RTX 4050?128-bitGA107202560RTX 3050128-bit+20%

The NVIDIA Ada Lovelace GPU family is expected to bring a generational jump similar to Maxwell to Pascal. It is expected to launch in the second half of 2022 but expect supply and pricing to be similar to current cards despite NVIDIA spending billions of dollars to acquire those good good TSMC 5nm wafers.

Which next-generation GPUs are you looking forward to the most?

Products mentioned in this post

GeForce RTX 3090
GeForce RTX 3090
USD 2780.99
RTX 2080
RTX 2080
USD 999.94
RTX 2080 Ti
RTX 2080 Ti
USD 1199.95

The links above are affiliate links. As an Amazon Associate, Wccftech.com may earn from qualifying purchases.

Submit