JEDEC Releases GDDR7 Memory Spec: Enhanced Graphics Memory Introduces Lightning-Fast PAM3 Signaling & On-Die ECC

JEDEC, the organization responsible for establishing memory standards, has released the official specifications for GDDR7 DRAM. This new iteration of the graphics RAM standard brings significant improvements in memory capacity and bandwidth. The switch to PAM3 Signaling on the memory bus is primarily responsible for the boost in memory bandwidth. Additionally, GDDR7 introduces more channels per DRAM chip, new interface training patterns, and on-die ECC to enhance memory reliability.

Mian Quddus, Chairman of JEDEC Board of Directors, described GDDR7 as a significant advancement in high-speed memory design. The shift to PAM3 signaling opens up new possibilities for extending the performance of GDDR devices and driving the evolution of graphics and high-performance applications.

GDDR7 has been in development for several years, with initial disclosures about the memory technology made by JEDEC members approximately a year ago. The final version of GDDR7 is expected to launch in 2024, aligning with previous announcements from memory manufacturers.

The most notable technical change in GDDR7 is the transition from two-bit NRZ encoding to three-bit PAM3 encoding on the memory bus. This change allows GDDR7 to transmit 50% more data than GDDR6 at an identical clockspeed. Consequently, GDDR7 can support higher data transfer rates, which is crucial for achieving faster performance with each new generation.

The first generation of GDDR7 is projected to operate at data rates around 32 Gbps per pin, with the potential for even higher rates up to 48 Gbps per pin. JEDEC’s press release highlights that GDDR7 memory chips can reach up to 192 GB/s per device, emphasizing the significant increase in bandwidth compared to PAM3 signaling alone.

Delving into the specifications, JEDEC has divided a single 32-bit GDDR memory chip into four 8-bit channels, expanding on the two 16-bit channels offered by GDDR6. While this separateion may not be noticeable to end-users, it has a substantial impact on the chip’s internal operations. By maintaining the 256-bit per channel prefetch of GDDR5 and GDDR6, GDDR7 adopts a 32n prefetch design, effectively doubling the amount of prefetched data per cycle of the DRAM cells.

The change in the number of channels per memory chip also affects multi-channel “clamshell” mode for higher capacity memory configurations. GDDR7 accesses two channels, referred to as two-channel mode, reading channels A and C from each chip. This mode remains consistent with how clamshell mode worked in GDDR6, indicating that memory capacity improvements rely on increasing chip densities rather than employing new techniques.

GDDR7 officially supports 64Gbit DRAM devices, twice the maximum capacity of GDDR6/GDDR6X. Non-power-of-two capacities such as 24Gbit and 48Gbit are also supported. The addition of larger memory chips pushes the maximum memory capacity of high-end video cards with a 384-bit memory bus to as high as 192GB, which is particularly beneficial for datacenter operators utilizing large language AI models. However, it’s worth noting that current memory cards still predominantly use 16Gbit chips, and 64Gbit chips are not expected to be available in the near future.

In terms of reliability, GDDR7 introduces on-die ECC capabilities similar to DDR5. This inclusion is not surprising given the increasing difficulty of producing flawless memory chips at higher densities. On-die ECC allows memory manufacturers to ensure reliable operation in the face of inevitable errors. The GDDR7 specification mandates a minimum of 16 bits of parity data per 256 bits of user data, enabling the correction of 1-bit errors and detection of 2-bit errors. Additionally, GDDR7 incorporates command address parity with command blocking (CAPARBLK) to enhance the integrity of the command address bus.

While the impact of on-die ECC on consumer video cards may be limited, its implications for workstation and server video cards remain to be seen. These segments have utilized soft ECC on top of unprotected memory for several generations, and it is likely to continue with GDDR7. The use of soft ECC offers more flexibility compared to the CPU space.

GDDR7 also introduces various reliability-related features to support PAM3 operation. These include core independent LFSR training patterns with eye masking and error counters. LFSR training patterns test and adjust the interface, eye masking evaluates signal quality, and error counters track errors during training.

The announcement of GDDR7 has garnered support from key players in the industry, including AMD, nVidia, and memory manufacturers such as Micron, Samsung, and SKhynix. While no specific products have been announced yet, Samsung and Micron have previously stated their intentions to ship GDDR7 memory this year. Therefore, we can anticipate the arrival of new memory and GPUs later in the year.

Overall, GDDR

Background Information

---

About AMD: AMD, a large player in the semiconductor industry is known for its powerful processors and graphic solutions, AMD has consistently pushed the boundaries of performance, efficiency, and user experience. With a customer-centric approach, the company has cultivated a reputation for delivering high-performance solutions that cater to the needs of gamers, professionals, and general users. AMD's Ryzen series of processors have redefined the landscape of desktop and laptop computing, offering impressive multi-core performance and competitive pricing that has challenged the dominance of its competitors. Complementing its processor expertise, AMD's Radeon graphics cards have also earned accolades for their efficiency and exceptional graphical capabilities, making them a favored choice among gamers and content creators. The company's commitment to innovation and technology continues to shape the client computing landscape, providing users with powerful tools to fuel their digital endeavors.

  

---

About nVidia: NVIDIA has firmly established itself as a leader in the realm of client computing, continuously pushing the boundaries of innovation in graphics and AI technologies. With a deep commitment to enhancing user experiences, NVIDIA's client computing business focuses on delivering solutions that power everything from gaming and creative workloads to enterprise applications. for its GeForce graphics cards, the company has redefined high-performance gaming, setting industry standards for realistic visuals, fluid frame rates, and immersive experiences. Complementing its gaming expertise, NVIDIA's Quadro and NVIDIA RTX graphics cards cater to professionals in design, content creation, and scientific fields, enabling real-time ray tracing and AI-driven workflows that elevate productivity and creativity to unprecedented heights. By seamlessly integrating graphics, AI, and software, NVIDIA continues to shape the landscape of client computing, fostering innovation and immersive interactions in a rapidly evolving digital world.

  

---

About Samsung: Samsung, a South Korean multinational conglomerate, has established itself as a global leader in various industries, including electronics, technology, and more. Founded in 1938, Samsung's influence spans from smartphones and consumer electronics to semiconductors and home appliances. With a commitment to innovation, Samsung has contributed products like the Galaxy series of smartphones, QLED TVs, and SSDs that have revolutionized the way we live and work.

  

Technology Explained

---

CPU: The Central Processing Unit (CPU) is the brain of a computer, responsible for executing instructions and performing calculations. It is the most important component of a computer system, as it is responsible for controlling all other components. CPUs are used in a wide range of applications, from desktop computers to mobile devices, gaming consoles, and even supercomputers. CPUs are used to process data, execute instructions, and control the flow of information within a computer system. They are also used to control the input and output of data, as well as to store and retrieve data from memory. CPUs are essential for the functioning of any computer system, and their applications in the computer industry are vast.

---

DDR5: DDR5 (Double Data Rate 5) is the next generation of memory technology for the computer industry. It is a modern day improvement on earlier DDR technologies, with faster speeds, greater bandwidth and higher capacities. DDR5 enables higher resolution, seamless gaming experiences and faster data transfer rates, making it an ideal choice for high-performance computing and 4K gaming. With its greater RAM compatibility, DDR5 provides faster buffering times and raised clock speeds, giving users an improved overall work system. DDR5 is also optimized for multi-tasking, allowing users to multitask without experiencing a significant drop in performance, increasing the productivity of digital tasks. As an ever-evolving technology, DDR5 is paving the way for the computer industry into a new and powerful era.

---

GDDR6: GDDR6 stands for Graphics Double Data Rate 6th generation memory. It is a high performance memory used in graphics cards and graphics processing units (GPUs), specifically targeting gaming, AI and deep learning-related applications. GDDR6 achieves higher bandwidth than previous generations, allowing faster and smoother gaming experience for users. It is also more power efficient, resulting in lower energy consumption overall. The improved power efficiency makes it adaptable to today's needs of thinner laptops and ultra-high definition gaming laptops. Additionally, GDDR6 is used in storage solutions and advanced data center applications to help streamline large amounts of data at lightning-fast speeds.

---

GDDR7: GDDR7 (Graphics Double Data Rate 7) is the seventh generation of graphics double data rate (GDDR) memory. It is a type of dynamic random-access memory (DRAM) that is specifically designed for use in graphics cards. GDDR7 memory offers a number of advantages over previous generations of GDDR memory. GDDR7 is a significant improvement over previous generations of GDDR memory. It offers faster speeds up to 32 gigabits per second (Gbps) per pin, lower power consumption, and improved error correction. This makes it ideal for use in high-performance graphics cards and other applications that require high bandwidth and low latency.

---

PAM3 Signaling: PAM3 (Pulse Amplitude Modulation with 3 levels) signaling is a technique used in communication systems to encode digital information for transmission over a physical medium, such as electrical wires or optical fibers. It's a modulation scheme that uses three different signal levels to represent multiple bits of data within a single symbol interval. PAM3 takes advantage of three distinct voltage levels in the signal to encode more information per signal pulse. This allows for higher data transfer rates without requiring a corresponding increase in the frequency of the signal.