NVIDIA is launching its brand new Ampere-based A10 and A30 Tensor Core GPUs using the GA102 GPU and GA100 GPU, respectively, aimed at data centers and are mostly geared towards virtualization platforms.
NVIDIA Ampere A10 24 GB GDDR6 & A30 24 GB HBM2 Tensor Core CPUs Launched
While both of these GPUs will be Ampere-based, the memory subsystem will be very different. The A10 will offer GDDR6 and the A30 going with the standard HBM2 memory standard for data centers. Here’s a detailed look at the specs.
NVIDIA A10 Ampere Tensor Core GPU
The NVIDIA A10 Tensor Core GPU features 72 SMs for a total of 9216 CUDA Cores and is powered by the GA102-890 SKU. The GPU features PCIe Gen 4.0 compliance and operates at a base clock of 885 MHz, and boosts up to 1695 MHz. In terms of memory, it features 24 GB GDDR6 VRAM, which operates at 12.5 Gb/s across a 384-bit wide bus interface while delivering a bandwidth of 600 GB/s.
It uses a champagne gold-colored shroud for its design, which comes in a single-slot, full-length form factor. There is no fan on it since this is a passively cooled card, and power is provided through a single 8-pin connector with the A10 Tensor Core GPU running off a single 8-pin connector meet its 150W TDP demand.
In terms of performance, according to WCCHTECH, “the NVIDIA A10 Tensor Core GPU offers up to 31.2 TF FP32, 62.5 TF TF32, 125 TF BFLOAT16, 250 TOPS INT8, 500 TOPS INT4 & twice the rates with sparsity.”
NVIDIA A30 Ampere Tensor Core GPU
The NVIDIA A30 Tensor Core GPU seems to make use of a rather cut-down variant of GA100 SKU that features a base clock of 930 MHz and a boost clock of up to 1440 MHz. The GPU operates at a speed of 1215 MHz across a 3072-bit wide bus interface as it is equipped with 24 GB of HBM2 VRAM. This means that we are looking at only three active HBM2 memory stacks that deliver up to 933 GB/s of memory bandwidth.
The NVIDIA A30 Tensor Core GPU features a dual-slot and full-length design and is powered by a single 8-pin connector too but has a higher rated TDP of 165W. “In terms of performance, the NVIDIA A30 Tensor Core GPU offers up to 5.2 TF FP64, 10.3 TF Peak FP64TF, 10.3 TF FP32, 82 TF TF32, 165 TF BFLOAT16, 330 TOPS INT8, 661 TOPS INT4 & twice the rates with sparsity.”
NVIDIA Tensor Core Ampere GPUs
According to a report by WCCFTECH, here are Inspur’s All-New GPU Servers Supporting A30, A10, and A100:
NF5468M6: “Ultra-flexible for AI workloads, supports 2x Intel 3rd Gen Intel Xeon Scalable processor and 8x NVIDIA A100/A40/A30 GPUs, 16x NVIDIA A10 GPUs, or 20x NVIDIA T4 GPUs; supports up to 12x 3.5-inch hard drives for large local storage in a 4U chassis”
NF5468A5: “Versatile AI server featuring 2x AMD Rome/Milan CPUs and 8x NVIDIA A100/A40/A30 GPUs; N+N redundancy design enables 8x 350W AI accelerators in full-speed operations for superior reliability.”
NF5280M6: “Purpose-built for all scenarios, with 2x Intel 3rd Gen Intel Xeon Scalable processor and 4x NVIDIA A100/A40/A30/A10 GPUs or 8x NVIDIA T4 Tensor Core GPUs in 2U chassis, capable of long-term stable operation at 45°C.”
Currently, more than 20 NVIDIA-Certified Systems are available from worldwide computer makers, and NVIDIA A30 and NVIDIA A10 GPUs will be available later this year from manufacturers.
At $3,595 per CPU socket, NVIDIA AI Enterprise is available as a perpetual license. At $899 annually per license, Enterprise Business Standard Support for NVIDIA AI Enterprise is available. Customers can apply for early access to NVIDIA AI Enterprise as they plan their upgrades to VMware vSphere 7 Update 2.