Samsung Announces Aquabolt 2nd Gen HBM2 With 50% Performance Boost

samsung aquabolt hbm12
Not only has Samsung ramped up production of its sought-after 8GB HBM2 chip, today the company has announced Aquabolt, its significantly faster 2nd generation HBM2. Aquabolt boasts the fastest data transmissions speeds, delivering 2.4Gbps data transfer per pin at 1.2V.

Today’s announcement marks a leap forward in HBM2, a field in which Samsung was already dominating.  The company’s 1st generation 8GB HBM2 package reached speeds of 1.6Gbps pin speed at 1.2V, with a 256GB/s data bandwidth. The new 2nd generation Aquabolt delivers a 2.4Gbps pin speed at 1.2V with a 307GB/s bandwidth -- that 50% performance boost is the big takeaway here. By comparison, Aquabolt can achieve data transfer speeds nearly 10x faster than the 32GBps data bandwidth of older 8-gigabit GDDR5 chips. Imagine a 1.2TBps bandwidth when four of these 8GB HBM2 packages are installed to systems or future GPUs? Glorious!
samsung aquabolt hbm 2

To reach these new plateaus, Samsung has researched new technologies for TSV design (Through Silicon Via) and thermal control. To boost thermal control, the chip-maker increased the number of thermal bumps between the HBM2 dies. A single 8GB HBM2 package consists of eight 8Gb HBM2 dies, which are vertically interconnected using over 5,000 TSVs per die. The company ensures that the large stack of TSVs will not cause collateral clock skew, which can be a concern. By minimizing the skew to a nominal level the company says they also significantly increased chip performance. Rounding out the overall chip improvements, there is an additional protective layer at the bottom to improve the package’s overall durability.

Introduced just over a year ago, HBM2 memory can already be found in some of the most bleeding-edge tech devices such as NVIDIA’s new Titan V and the AMD Radeon RX Vega. Yet the need for high-performance HBM2 reaches far beyond the gaming sector. The company is positioning itself as the go-to provider for IT customers worldwide and OEMs in supercomputing, AI and general graphics processing professionals.